Attribution
368 articles on attribution, newest first.
Do Google Ads and Meta double count sales?
Usually, yes. Google Ads and Meta each count a sale their own ads touched, by their own rules, and neither checks the other's claims. One Shopify order can sit in both dashboards, so their total often tops your real sales. Only Shopify's order count counts each sale once.
Do Google Ads and Meta double count food and drink sales?
If you sell food or drink, usually yes, and most often on reorders. A regular sees your Meta ad, then clicks your Google brand ad to reorder, and both platforms claim that one order. Split each platform's purchases into new and returning customers before you judge either.
How to check if Google Ads and Meta double count sales
Pull one closed month from Shopify, Meta and Google Ads, then set the two platforms' purchase claims against Shopify's orders. If the sum is higher, rule out duplicates inside each platform, then use GA4's paths to see which journeys both ads touched.
Do Meta view-through conversions count?
Yes in Meta's reports: with the usual default, a purchase within 1 day of someone seeing your ad counts as a result, no click needed. GA4 never counts it, because a view brings no visit. Whether the view caused the sale is a separate question.
Do Meta view-through conversions count for food brands?
If you sell food or drink, Meta usually counts a purchase within 1 day of someone seeing your ad, no click needed. When buyers reorder often, that window fills with regulars who were buying anyway, so compare it with first purchases before you trust it.
How to check Meta view-through conversions, step by step
Open Compare attribution settings in Ads Manager and put 1-day view beside 7-day click, with Meta's incremental model next to them. Then look for the same buyers in GA4 and in Shopify's order journeys. If views carry a campaign, test it before you scale it.
Does Meta overreport conversions?
Usually yes, if you mean sales the ads caused. Meta counts every purchase inside its window after a click, view or engagement, including buyers who were coming anyway. The credit runs highest on past customers and quick buyers, and a doubled pixel can inflate the count itself.
How to check if Meta overreports conversions, step by step
Split Meta's purchases three ways in Ads Manager: by attribution window, by first versus all conversions, and by audience segment. Then set each slice against Shopify's first-time and returning customers and confirm each order reaches Meta once. Test the slice that leans on views or past buyers.
How do I know if Google Ads is profitable?
Usually when the gross profit on the sales Google Ads adds beats what you pay Google. Its columns show revenue, so take product costs off, ask which buyers were coming anyway, and wait for late conversions before you judge a month.
How to check if Google Ads is profitable, step by step
Find your margin per order in Shopify, give Merchant Center your product costs and add Gross profit to your Google Ads columns. Subtract ad cost per campaign, leave out unfinished days and refunds, then test the biggest campaign in a few regions.
How do I know if my Meta ads are working?
You usually know when three checks agree. Meta's return beats the break-even ROAS your margin sets. Total Shopify sales move with Meta spend. Sales dip where people stop seeing the ads. Ads Manager alone only shows what Meta counted.
How to tell if your Meta ads are working, step by step
Work out break-even ROAS from your Shopify margin and confirm each ad set has left learning. Read Purchase ROAS under one attribution setting. Then set weekly Meta spend against Shopify sales, and test with a control group before you add budget.
How do I measure Meta ads incrementality?
Usually with an experiment that keeps some people or regions away from your Meta ads, then compares total sales. Use Meta's Conversion Lift if a campaign qualifies, or a regional holdout if not. Ads Manager, GA4 and Meta's incremental column show credit or predictions, not cause.
How do I measure Meta ads incrementality for fashion?
If you sell clothes, use a holdout: Meta's Conversion Lift if a campaign qualifies, a regional test if not. Run it in full-price weeks, test catalog ads apart from prospecting, and cut the sales lift by your return rate before you judge it.
How to measure Meta ads incrementality, step by step
Write down the return your margin needs. Read Meta's claim next to its incremental estimate. Then test with a holdout: Conversion Lift if a campaign qualifies, a regional holdout if not. Judge the result against your bar and re-test when budgets or seasons change.
How do I measure Performance Max?
Measure Performance Max in two layers. Google Ads' channel performance and search terms reports show where it served and what it claims. An uplift experiment or a geo Conversion Lift study shows what it added. The first is a claim; the second is a test.
How do I measure Performance Max for sports nutrition?
If you sell sports nutrition, measure Performance Max on new customers, not total conversions. Reorders can fill its report with buyers who would have come back anyway. Split new and returning buyers, keep brand searches out of its count, and test before you raise the budget.
How to measure Performance Max, step by step
Check the purchase action's windows, then read Performance Max's channel performance and search terms reports. Split new and returning buyers, find its visits under Cross-network in GA4 and download the paths. Then run an uplift experiment to see what it adds.
How do I split budget between Google and Meta?
Usually by job, not by reported ROAS. Google search and Shopping mostly catch people already looking for what you sell, while Meta mostly finds people who are not looking yet. Fund each until a test shows its extra sales no longer clear your break-even, then move money in small steps.
How to split budget between Google and Meta, step by step
Usually, start by putting Google and Meta on the same footing: one break-even ROAS, both counting rules written down and brand search split out. Then compare them on GA4's ruler, check where each has room, and let a regional test settle the split.
How do I track Google Shopping sales?
Google Ads counts purchases after a Shopping ad click and shows them by product. GA4 files the visits under Paid Shopping, under Cross-network for Performance Max, or under Organic Shopping for free listings. No report says which sales the ads added.
How to track Google Shopping sales by product, step by step
Link Merchant Center, Google Ads and GA4, check the purchase action and both auto-tagging switches, then read sales per product in the Products tab. Cart data shows what actually sold, and GA4 splits the visits across Paid Shopping, Cross-network and Organic Shopping.
How do I track Meta ads in GA4?
Add utm_source, utm_medium and utm_campaign to every Meta ad. Use a source on GA4's social list, such as facebook or instagram, and a paid medium, and GA4 files the clicks under Paid Social. Link Meta under Product links to add cost. Ad views never reach GA4.
How to track Meta ads in GA4, step by step
Pick one paid medium and a social source GA4 recognises, then add them in each Meta ad's URL parameters field. Confirm in Traffic acquisition that the clicks land in Paid Social. Then link Meta under Product links to bring cost, clicks and impressions into GA4.
How does ad attribution work?
Ad attribution ties a sale back to an ad. A click ID, cookie or pixel marks the visit, and the purchase fires a conversion event. The platform credits its ad if the sale falls inside its window. Each platform does this on its own data, so totals rarely agree.
How to trace ad attribution for one order, step by step
Pick one recent order and follow it through each tool. Check the mark the ad left, then the order's journey in Shopify, then the model and window in GA4, Google Ads and Meta. Adding up what each tool claims shows how one sale gets counted more than once.
How does Google Ads attribution work?
Google Ads credits a conversion to its own ads when the buyer clicked one inside the conversion window, 30 days by default. Data-driven attribution, the default for most actions, splits that credit across Google ad interactions only.
How does Google Ads attribution work for electronics?
If you sell consumer electronics, Google Ads usually credits a sale to its ads clicked in the 30 days before purchase. Data-driven attribution shares it out among Google clicks. Long research and searches for the exact model decide which click looks like the hero.
How to check Google Ads attribution, step by step
Open each purchase conversion action and read its attribution model, conversion windows and count. Then compare models in the attribution reports and line up click-date and conversion-date columns before you hold Google Ads against GA4 or Shopify.
How does Meta ads attribution work?
Meta usually credits a purchase to an ad clicked within 7 days or seen within 1 day. Likes, shares and saves count separately, as engage-through. Meta checks only its own ads, so its report cannot show what an ad caused.
How does Meta ads attribution work for sports nutrition?
If you sell sports nutrition, Meta usually credits a sale to an ad clicked within 7 days or seen within 1 day. Reorders make that generous: a regular who sees your ad on restock day can count as a Meta sale.
How to check Meta ads attribution, step by step
Confirm Meta receives each purchase once, read every ad set's attribution setting, tag each ad link and see how GA4 files the visits. Then the gap between Ads Manager and GA4 usually comes down to views and engagements only Meta sees.
Is 7-day click attribution accurate?
Accurate as a record of timing: a purchase came within 7 days of a link click on your Meta ad. Not as a measure of cause. It credits buyers already on their way, misses slow ones and never looks at your other channels.
How to test 7-day click attribution, step by step
Count repeat purchases with First and All conversions, and see how much of the count needs days 2 to 7. Rank campaigns under Meta's incremental model and count returning buyers in Shopify. Then book a lift test for the campaign whose profit depends on the window.
Is branded search worth paying for?
Often only in part. Brand campaigns look brilliant because people searching your name have mostly decided. They earn their keep when competitors or resellers bid on your name, or the ad shows what your free listing cannot. A regional pause test settles it.
Is branded search worth paying for in fashion?
If you sell fashion, branded search is more often worth some spend. Stockists and resellers may bid on your name, and drops send shoppers searching for one item. Check who shows up on your name before you pay full price or pause.
How to test if branded search pays off, step by step
Put brand searches in one campaign, check who else bids on your name, and record paid and organic brand clicks with Shopify orders. Then pause brand ads in a few regions. If total orders there hold, the ads were buying clicks you already had.
How to calculate time decay attribution, step by step
GA4 and Google Ads no longer offer time decay, so you build it in a sheet. Export your paths from GA4 and give each touch a weight that halves for every half-life before the sale. Split each path's revenue by those weights, and test more than one half-life.
Is Google Ads stealing my organic sales?
Usually not the sales themselves. An ad above your own top free listing can buy clicks you would have had. GA4 then files those sales under Paid Search instead of Organic Search. Judge all your Google channels together, not the split between them.
How to spot Google Ads stealing organic sales in GA4
Check which attribution model GA4 uses, put every Google channel in one monthly table and mark the dates your Google Ads changed. Then read brand clicks and position in Search Console. A flat total with growing Paid Search points to ads taking sales organic would have brought.
Is multi-touch attribution dead?
Not dead, but smaller. Google retired first click, linear, time decay and position-based in 2023, so in GA4 and Google Ads multi-touch now means data-driven. Consent and device switching leave it shorter paths. It still splits credit; it never tested what an ad caused.
Is multi-touch attribution dead for health and beauty?
For a health and beauty store, multi-touch attribution is alive but half blind. Creator videos persuade without clicks, consent banners cut paths, and reorders return by habit. It still splits the visits GA4 records, so judge creator and social spend with a holdout test.
How to check if multi-touch attribution still works for you
Check which model you still have, how much revenue sits on multi-touch paths and how far data-driven moves each channel. Then see how far Shopify's Any click overshoots your sales, read Meta's incremental column, and send the channel the models argue about to a holdout.
Last click vs data-driven attribution, which is better?
Usually data-driven, if your store has enough journeys with more than one click. It spreads credit across recorded touches, while last click gives it all to the final one. If most sales follow a single visit, the two agree. Neither tells you what a channel caused.
Last click or data-driven for a food and beverage store?
If you sell food and drink, the two models usually agree on quick reorders, because a one-visit order has nothing to split. Data-driven earns its place on first orders that take several visits. Neither can credit tastings, shelves or renewals charged without a visit.
How to test data-driven against last click before switching
Compare the two on purchases only in GA4, then price them in Google Ads' Model comparison report. Switch one conversion action, reset its Target ROAS by the same percentage change, and wait out the lag before you judge the result.
Should I bid on my own brand name?
Usually yes when someone else shows up on searches for your name: a rival, a reseller or an affiliate. If your free listing sits alone at the top, bid little or test a pause. Either way, keep budget off customers looking for a login, an order or a return.
Should a health and beauty brand bid on its own name?
If you sell health and beauty, bidding on your own name usually pays when retailers, marketplaces or rival brands show up on your product names. It pays less on plain reorder searches you already win.
How to bid on your own brand name, step by step
Check who shows up on your name in Ad preview first. Then build a separate Search campaign with exact and phrase brand keywords. Block support searches, keep other campaigns off your name and cap bids with a limit worked out from your margin.
Should I trust Meta ads conversions?
Trust Meta's conversions as Meta's claim about which sales its ads touched, not as sales the ads caused. They usually hold up for comparing ads under one attribution setting, and mislead when added to other platforms' numbers or used to set budget.
How to audit Meta ads conversions, step by step
Start in Events Manager with coverage, matching and diagnostics for Purchase. Then set Ads Manager's claim against what Meta received and against your Shopify orders. Read the modeled and incremental results, and settle the rest with a lift test or a geo holdout.
Should I turn off my Meta ads?
Usually not all at once. Switch off ad sets that lose money even on Meta's own count. Keep the rest until a region test shows what happens to total sales without them, because a full stop mostly measures the calendar.
How to test turning off Meta ads, step by step
Pick regions that move like the rest, take Meta out of them in every ad set and cut budgets by their share. Then compare Shopify sales by billing location with each group's own baseline, for longer than your slowest buyers take.
What does Paid Search mean in GA4?
Paid Search is GA4's channel label for visits from ads on search engines. Google Ads clicks from Search campaigns land there, and so do tagged links that pair a search engine with a paid medium such as cpc. Shopping and Performance Max ads usually land in other rows.
How to audit Paid Search in GA4, step by step
Filter GA4's Traffic acquisition report to Paid Search. Then add Session source / medium, Session source platform and Session campaign to see what fills the row. Check the channel rules in Admin, the Google Ads link and auto-tagging, and test that the click ID survives redirects.
What is causal attribution?
Causal attribution gives each channel credit only for sales that would not have happened without it. It is usually estimated with a holdout test, a geo experiment or a mix model. Last click and data-driven only split up sales that already happened.
How to measure causal attribution, step by step
Pick one channel, pair your regions by past sales and switch the channel off in one region of each pair. The difference in sales growth between the groups is what the channel caused. Divide it by GA4's last-click credit to see how far to trust that number.
How to diagnose a Meta ROAS drop, step by step
Compare two full weeks in Ads Manager. Split ROAS into four parts: cost per thousand impressions, link click rate, purchase rate and value per purchase. The part that moved most usually names the cause. Check Shopify sales and Events Manager first.
Why do attribution models give different answers?
Because each model is a different rule for splitting one sale among the touches before it, and each tool sees different touches over different windows. If most sales follow a single visit, models mostly agree. The arguments sit in the longer journeys.
What is engage-through attribution on Meta?
Engage-through is how Meta credits a purchase made within 1 day after someone engaged with an ad but did not click its link. Likes, comments, shares, saves and 5-second video plays count. It replaced engaged-view in March 2026 and shows timing, not cause.
What is engage-through attribution for home goods stores?
If you sell home goods, engage-through attribution credits a Meta ad when someone liked, saved, shared or watched it and bought within 1 day. If your shoppers save a sofa and buy it a week later, the save earns nothing, so it mostly catches quick add-ons.
How to check engage-through attribution in Meta Ads Manager
Find the ad sets that use engage-through, then use Compare attribution settings to put 1-day Engagement beside click and view results. Learn how Results adds them up, check the video rule, and look for the same buyers under Direct in GA4 and on Shopify.
What is first click attribution?
First click attribution gives all the credit for a sale to the first channel the buyer interacted with. It shows who introduced the buyer, not who closed the sale. GA4 and Google Ads dropped it in 2023; Shopify's marketing reports still offer it.
What is first click attribution for a pet products store?
First click gives all the credit to the first channel a buyer interacted with. If you sell pet food or supplies, reorders matter. Shopify resets the first interaction after each order, while GA4's User acquisition keeps the original channel.
How to see first click attribution in Shopify and GA4
GA4 dropped first click in November 2023. Use Shopify's Attribution menu, or rebuild it from GA4. User acquisition sorts revenue by the channel that first brought each user, and the paths export lets you credit each path's first touch.
What is Google Ads data-driven attribution?
Data-driven attribution is the default model in Google Ads. It splits each conversion's credit across the Google ad interactions before it, based on how converting and non-converting paths differ in your account. It cannot see Meta, email or organic visits.
How to use Google Ads data-driven attribution, step by step
Open the Switch to DDA tab to see each purchase action's model and eligibility, then switch it there or in its settings. Add the current model columns and move Target CPA or Target ROAS by the same percentage the results moved. Then wait out your average days to conversion.
What is last click attribution?
Last click attribution gives all the credit for a sale to the last channel clicked before buying. GA4's Paid and organic last click skips direct visits; Shopify's Last click counts them. It is fair for one-visit buyers and hides earlier touches on longer journeys.
How to see last click attribution in GA4, step by step
Open GA4's Traffic acquisition report, which always uses last click, then compare Paid and organic last click with data-driven under Advertising, Attribution, Attribution models. Check Google Ads' Model comparison and Shopify's Attribution menu the same way.
What is linear attribution?
Linear attribution splits the credit for a sale equally across every recorded touch on the path, so three clicks get a third each. GA4 and Google Ads no longer offer it, while Shopify's marketing reports still do. It rewards channels that show up often, not the ones that caused the sale.
How to run linear attribution in Shopify and GA4
In Shopify, add Linear to a marketing report's Attribution menu. GA4 dropped linear in November 2023. There you export the Attribution paths report, split each path's revenue evenly across its touchpoints and add the shares up by channel.
What is marketing attribution?
Marketing attribution is how you split the credit for a sale among the marketing a buyer touched first: ads, emails, searches and visits. A rule or a model does the splitting, inside each tool's own window. It shows which channels were on the path, not whether the sale needed them.
What is marketing attribution for a sports nutrition store?
For a sports nutrition store, marketing attribution is the same credit split as anywhere, with one twist: products run out and get bought again. A refill typed in within GA4's lookback window can be credited to the ad that won the first order. So read first orders and reorders apart.
How to set up marketing attribution, step by step
Start in GA4: check that purchase is a key event, choose the attribution model and lookback window, and link Google Ads with auto-tagging on. Then tag your links, write down Meta's attribution setting and Shopify's model, and read GA4's attribution reports with those rules in mind.
What is multi-touch attribution?
Multi-touch attribution splits the credit for a sale across several touchpoints on the buyer's path, by a fixed rule or a data-driven model. It only changes the answer on journeys with more than one touch, and more than one channel.
How to read multi-touch attribution in GA4, step by step
Check that GA4's reporting model is data-driven, open the Attribution paths report and keep only paths longer than one touch. Then read how credit spreads over early, mid and late touchpoints, and compare Google Ads' assists and Shopify's Linear model.
What is Paid Shopping in GA4?
Paid Shopping is the GA4 channel for visits from paid shopping ads, usually Google Ads Shopping campaign clicks. Tagged links with a paid medium and a shopping-site source or shop in the campaign name land there too. Performance Max lands in Cross-network.
What is Paid Shopping in GA4 for a pet store?
For a pet store, Paid Shopping usually holds clicks from Google Shopping campaigns for food, litter and gear. Paid links with shop in the campaign name land there too. Reorders through Shopping ads land there too, so the row can look like more new demand than it is.
How to clean up Paid Shopping in GA4, step by step
Filter Traffic acquisition to Paid Shopping and add source platform, source / medium and campaign beside it. Every line should be a shopping ad you paid for. If Meta ads turn up, a shop campaign name usually sent them: fix the tags and reorder a custom channel group.
What is position-based attribution?
Position-based attribution, or U-shaped, gives 40% of a sale's credit to the first touch and 40% to the last, and shares 20% across the middle. GA4 and Google Ads dropped it in 2023 and Shopify does not offer it, so you rebuild it from a paths export.
Position-based attribution from a GA4 export, step by step
GA4 and Google Ads dropped position-based attribution, and Shopify does not offer it. Build it from GA4's Attribution paths export: 40% of each path's revenue to the first and last touch, 20% across the middle, totalled by channel.
What is the best attribution model for ecommerce?
Usually there is no single best model. In GA4, data-driven is the sensible default and last click the check. But every model only splits credit among touches GA4 recorded, so if most revenue comes from one-touch journeys, the choice barely changes the answer.
What is the best attribution model for a home goods store?
If you sell home goods, the best model is usually GA4's data-driven model with the longest lookback window, checked against last click. Furniture and decor buyers often compare and come back before paying, so last click tends to overpay whichever channel closes.
How to pick the best ecommerce attribution model in GA4
Check which model GA4, Google Ads and Shopify each use. Compare data-driven with last click in GA4's Attribution models report and size your multi-touch revenue. Then see whether the model flips a channel across break-even ROAS.
What is the default Meta attribution setting?
Usually 7-day click, 1-day view and 1-day engagement. Meta credits a purchase to an ad clicked within 7 days, or seen or engaged with within 1 day. Its API reference leaves engagement out. The setting also steers who sees your ads.
What is the default Meta attribution setting for beauty?
Usually 7-day click, 1-day view and 1-day engagement, the same as for any store. If you sell health and beauty, the 1-day parts matter most. Quick buys after a scroll get credit. A save today and a purchase next week get none for the save.
How to find your Meta attribution setting, step by step
Add the Attribution setting column and open each ad set's attribution selections. Compare the windows side by side, then hold them against GA4's lookback window and your buyers' days to purchase. Meta names 7-day click, 1-day view and 1-day engagement as its default.
What is time decay attribution?
Time decay attribution splits a sale's credit across the touches before it and gives the most to the latest ones. Google's old default halved a touch's credit for every 7 days before the sale. GA4 and Google Ads no longer offer it, so you calculate it yourself.
What is time decay attribution for an electronics store?
Time decay gives most of a sale's credit to the touches nearest the purchase. If you sell electronics, that usually means the deal email, the price check or the brand search on buying day. The reviews that sold the model weeks earlier get a sliver.
Which attribution model should a Shopify store use?
Usually last non-direct click for weekly channel reports, because it names the marketing that brought a buyer back. Keep first click as a second view, and leave Meta and Google Ads on their own settings for bidding. No model shows what a channel caused.
How to choose a Shopify attribution model, step by step
Open a Shopify marketing report with the Attribution menu and compare last click, last non-direct click and first click. Then check GA4 and Meta against the same rule, pick one house model per job and label every report with it.
Why did my Meta ROAS drop?
Usually one of three things: each buyer cost more, fewer people bought, or Meta counted fewer of the purchases that still happened. The first two cost money; the third only changes the report. Check Shopify's sales for the same days before you change any ad.
Why did Meta ROAS drop for an electronics store?
If you sell consumer electronics, a Meta ROAS drop often starts outside Meta. A rival cut the price of the same model, a new version made buyers wait, or a best seller ran short. Thin margins make each drop hurt more, so check prices, launches and stock first.
Why do attribution models disagree for a fashion store?
Usually because a fashion purchase mixes steps each rule weighs differently: a creator post nobody clicks, several browsing visits and a sale email that closes. Last click credits the email, Meta may credit an ad view and data-driven splits the visits between.
How to trace why attribution models disagree, step by step
Pick one common multi-touch path in GA4's paths report. Run it through every rule: GA4's models, Google paid channels last click, Shopify's five models and Meta's windows. Note each tool's window and scope. The channel whose credit swings most is your next test.
Why does Google Ads show more conversions than GA4?
Google Ads only counts sales that followed its own ads, dated to the click, and may include modeled conversions. GA4 shares each sale across every channel and dates it to the purchase. Counting settings and time zones widen the gap.
How to reconcile Google Ads and GA4, step by step
Pick one closed week in one time zone. Keep one primary purchase action, and compare Google Ads conversions by conversion time with GA4's Paid Search and Cross-network purchases. Then read GA4's credit rules and paths. What remains is usually credit.
Why does Meta report more sales than GA4?
Meta counts a sale when the buyer clicked a link, engaged with or only saw an ad inside its window. GA4 only sees visits, so it gives the sale to the channel behind them, often not Meta. Both follow their own rules.
How to reconcile Meta and GA4 sales, step by step
You will not make Meta and GA4 agree, but you can explain the gap. Use one week and one time zone. Split Meta's purchases by link click, engagement and view, check that GA4 files Meta clicks as Paid Social, and set both against Shopify orders.
Why don't my Meta ads show in GA4?
Usually because GA4 never got a source it could read. Untagged clicks land in Organic Social or Direct, refused cookies and ad blockers hide visits, and a phone-to-laptop purchase looks like a new user. Part of the gap is permanent.
Why don't Meta ads show in GA4 for a pet store?
If you sell pet products, Meta ads usually vanish from GA4 in three places. Reorders return by email, subscription or a typed address. Phone taps become laptop orders. Gear purchases take days of comparing.
How to find missing Meta ad clicks in GA4, step by step
Count Meta's outbound clicks for one closed week, then GA4's sessions from every Meta source. Test-click your ad in debug mode to see if the tags survive. Then sort the rows by channel and check your cookie banner and landing domain.
Why is my Google Shopping ROAS so high?
Usually because Shopping ads meet buyers at the last step: they saw the photo and price before clicking. Google Ads then counts the sale for up to 30 days, shares credit only among Google ads and keeps returns in the value. High often means late, not more.
Why is Google Shopping ROAS so high for home goods?
If you sell home goods, Shopping ROAS often runs high because big baskets meet decided buyers. People measure, compare and come back to search the exact item, then click the ad with the photo and price. Google's value leaves out delivery costs and keeps returns until you correct it.
How to audit a high Google Shopping ROAS, step by step
Work through Google Ads in order: what feeds the Conversions column, how each sale is valued, and which searches and products carry the value. Then compare Shopify and GA4. Fix brand searches, duplicates and returns before you raise the budget.
Attribution models: buyer questions answered
14 buyer questions about attribution model, each answered from the platforms' own help pages: a short answer, a step-by-step walkthrough with menu paths, and for some questions a version for one store type.
Meta ads attribution: buyer questions answered
14 buyer questions about Meta ads attribution, each answered from the platforms' own help pages: a short answer, a step-by-step walkthrough with menu paths, and for some questions a version for one store type.
Google Ads attribution: buyer questions answered
14 buyer questions about Google Ads attribution, each answered from the platforms' own help pages: a short answer, a step-by-step walkthrough with menu paths, and for some questions a version for one store type.
A Shopify measurement stack: what each tool answers
Shopify answers what sold, GA4 which visits led to key events, ad platforms what they claim, a holdout whether it would have happened anyway, and a causal read what each channel caused. Drop any tool with no question.
Attribution platforms compared on published facts
Published facts let you compare platforms on method, what you install and what they charge, and they can't say which estimate is right. Each price below is dated, in alphabetical order, with no ranking.
Cross-channel attribution: how credit gets double counted
Every platform reports the orders it can tie to its own ads inside its own window, so the same order shows up in several dashboards. Reconcile one full week of platform claims against Shopify orders and the any click model.
Campaign attribution: from UTM tags to a budget decision
Fix the tags first, then read the Unassigned share, then ask whether the campaign's order count is big enough to act on. GA4 treats UTM values as case sensitive and sorts channels by fixed rules.
Attribution analytics: five questions your data must answer
The five questions are which channels start journeys, how long buyers take, how many touches they need, how many are new, and how much traffic is unassigned. GA4's attribution paths report and Shopify's customer reports answer them from data you already have.
Attribution software: seven criteria for comparing options
Put validation first: ask whether a vendor's numbers have been checked against randomised experiments. Then compare input data, method, price model, privacy, exports and time to first answer, and send the gaps to the vendor as written questions.
Single-touch vs multi-touch attribution on the same orders
Single-touch rules answer who closed or who started; multi-touch rules spread credit over the touches they see. None shows what a channel caused, so a budget call needs a test.
Do you need attribution software? A sizing test
Not until the free rules disagree about a decision bigger than a test, and your data can carry the method. Three gates on your own export tell you where you stand.
Where attribution tools get data: pixels, APIs, exports
Attribution tools get data by browser pixel, server-to-server call, platform API or analytics file, and each route loses something first. Comparing one week of orders across Shopify, GA4 and the tool shows your own gap.
Multi-channel attribution on Shopify: a setup walk-through
Four steps, in order: a purchase event with a unique transaction ID, a consent setting that decides who is measured, UTM tags on every link, and GA4's channel rules. Reconcile one closed week of Shopify orders to GA4 after each.
MMM vs MTA: which question each one answers
MTA splits credit for orders along the paths of buyers you can see; MMM estimates how sales respond to channel spend from weekly totals. They need different data, and both need an experiment to check them.
Multi-touch attribution models explained with one order
Each rule splits the same sale differently: last click gives one touch everything, linear splits it evenly, time decay favours the latest touches. GA4 still offers data-driven and two last-click models.
Attribution in 2026: what changed for measurement software
Meta announced it would drop two view windows from its API and narrow click-through to link clicks, GA4 added per-conversion settings and longer windows, and Google announced it would retire Chrome's Attribution Reporting API. Each changes what a tool can see, not what buyers did.
Measuring marketing ROI with multi-touch attribution
Not by itself. It splits credit for orders; return also needs margin, returns and the share of those orders the ads caused. A holdout gives that share, and one line of arithmetic turns it into profit.
Attribution models compared: rules, data-driven, and MMM
GA4 offers three models: data-driven, paid and organic last click, and Google paid channels last click. First click, linear, time decay and position-based are gone. A marketing mix model reads weekly totals instead of paths.
Evaluate an attribution platform in a week: a scorecard
Test four things in a week: totals that reconcile to Shopify, one number rebuilt from order rows, a known channel change, and a comparison with an experiment. Score each pass, partial or fail before you commit.
From attributed revenue to profit: attribution and ROI
Profit is attributed revenue after returns, product cost, shipping, fees and ad spend. Break-even ROAS is order value divided by contribution per order, and the attribution model moves profit between channels without changing the total.
Multi-touch attribution in 2026: what it can and cannot see
Multi-touch tools credit only what they can see: Safari, Apple's tracking prompt and declined consent each remove visits, and credit along a path is not cause. Measure your own gap between Shopify orders and GA4 purchases.
Ecommerce attribution tools: a shortlist by job
Pick by the job: tracking paths, running experiments, modelling spend, or reading what each channel caused from history. Start with what GA4 and Shopify already report, then add one tool whose method matches the decision.
Attribution software: four types and what each can see
Four types exist: rule-based tools credit logged touchpoints, data-driven models compare converting and non-converting paths, mix models read aggregated spend, and experiments measure one randomised change. Pick the type that can see your question.
How accurate is campaign attribution? What experiments show
Less accurate than a dashboard implies: in published Facebook experiments, observational methods often missed the randomized result. A paired-region holdout on one campaign shows how far off your own number is.
Multi-touch attribution software: a buyer's guide
Judge a multi-touch tool on what it sees, how it treats visitors who decline consent, what its numbers were compared with, and whether you can export the rows. Put five questions to the vendor in writing before any demo.
How to choose an attribution tool: a decision guide
Start from the decision and the data you hold. To cut or move a channel you need an experiment; a mix model needs years of weekly history; platform lift tests have entry thresholds. Otherwise start with GA4 and Shopify.
Do UTM parameters show which ad made the sale?
UTM parameters show which ad sent the last click before an order, not which ad caused it. They miss buyers who saw an ad and came back another way, and on a small budget each ad collects too few sales to tell a winner from luck. Tag the angle, judge totals, and test before you scale.
Q4 attribution checkup: what to check before Black Friday
Run the Q4 attribution checkup in October, not in Cyber Week. Black Friday 2026 is Friday 27 November. Tag the promotion, put paid, email and affiliates on one UTM scheme, learn how many days your multi-channel buyers take, price the discount against break-even ROAS, and leave tracking alone the week before.
How often to check marketing attribution: once a quarter
Once a quarter, before the next budget is set, with four checks out of turn: after a tracking or site change, when you add or cut a channel, before peak season, and when a platform's ROAS and your revenue disagree. Why the interval is a quarter, and a seven-step checkup that runs on GA4 and free tools.
First order at a loss? Judge it on payback, not ROAS
A first order priced to lose money will always fail on first-order ROAS. Count the loss as acquisition cost, track how many subscribers reach the second charge, split them by the channel of their first order, and find the month each cohort pays back.
The five-minute story ad that never gets the last click
A story ad that sends viewers to an advertorial opens the path and rarely closes it, so last-click pays the price ad that brings buyers back. Read the whole path in GA4's Attribution paths export, tag the advertorial's links, and test the story ads with a regional holdout.
The trust check that hands your sale to search
Shoppers wary of scam stores may check a new brand before buying, then come back through search or direct. Last-click attribution credits that return trip and undercounts the ad that started it.
How long do your multi-channel buyers take?
In Causality Engine's own GA4 property, sign-ups from people who saw two or more channels took 25.49 days, against 1.35 days for one channel with Direct left out. Your store will differ, so measure your own lag in GA4 and fit your windows and tests to it.
Comment a keyword, lose the attribution trail
When a viewer comments a keyword and gets the link by DM, GA4 only knows what the link tells it. Tag every DM link with its own campaign, then judge each keyword on orders and revenue, not on comments or clicks.
Scaling, proof and the cloned page: 12 answers
Twelve answers grouped into the three problems the last week of ecommerce discussion keeps circling: what scaling does to the read, what makes a number survive finance, and how to measure a page an AI rebuilt in minutes.
LTV, creative volume, and what breaks measurement
Trading margin for lifetime value and shipping creative in volume are both sound strategies. Both also remove the contrast that a causal read depends on, and both are fixable.
The real cost of causal attribution software
Licence fees are the smallest line in the total. For a small ecommerce team the cost that matters is integration work, analyst hours, and the decisions delayed while you wait.
Judging an attribution vendor on transparency
Nine questions that separate a documented method from a black box, and what a straight answer to each one sounds like when a vendor actually gives you one.
How to defend an attribution finding
What a skeptical finance stakeholder actually asks about an attribution result, and the four things the report has to carry so the answers are already in it.
AI era ecommerce attribution: 24 answers
Twenty-four answers grouped into three questions the last month of ecommerce discussion keeps circling: who sent the visitor, what did the agent change, and why did that product sell.
The cost of leaving AI traffic out of ROAS
When a channel goes uncounted, its revenue is credited to whatever sat closest to the sale. That inflates the neighbours and distorts every budget decision built on them.
Where AI orders hide in attribution reports
Assistant-driven orders do not vanish. They get filed under something else entirely. Five places they surface, and the signature that gives each one away.
The AI referral baseline window is closing
A clean before reading for assistant referrals only exists while the channel is still small. Once it is material, the comparison period has already gone.
What DTC teams get wrong about AI traffic
Four reasoning errors that show up repeatedly when ecommerce teams first try to measure assistant referrals, and what the correction looks like for each.
Own your AI referral data before a vendor does
Assistant referral data either lives in an export you control or in a vendor dashboard you cannot audit. That difference decides whether you can ever check the number.
Testing if AI referrals are incremental
Three designs that separate assistant referrals creating demand from those merely harvesting demand you already had, ordered by cost and by strength of evidence.
How to see AI referrals in a GA4 export
A concrete method for isolating assistant referrals in a GA4 export, and an honest account of what the number you get can and cannot support.
AI referrals broke your attribution quietly
AI assistants send buyers with no UTM and often no referrer at all. Those orders land in direct, which is already the bucket that absorbs everything nobody can explain.
Telling the team a channel is wasting spend
The person running it will hear a verdict on their work, because that is what it sounds like. Four framings that keep it about the budget instead.
The cost of a method nobody can explain
An unexplained number survives until the first serious challenge, then takes the whole measurement programme down with it. Three costs of that dependency.
What your team should know before agents read the data
Once a model can summarise the numbers, its summary becomes the version people quote. Three things worth establishing before that happens.
Ask for the methodology before you buy, not after
Leverage is highest before the invoice and effectively gone after it. Four things worth requesting in writing while the asking still costs the vendor something.
A method you can explain is a method you own
If understanding your own numbers depends on a vendor being reachable, you are renting comprehension along with the software.
Plain-English methodology is part of the product
A number you cannot interrogate is a number you cannot defend in a budget meeting. The documentation is a feature, not a courtesy.
What marketing attribution debt actually is
Attribution debt is the accumulated gap between what your reporting claims and what your marketing caused. Like any debt, it compounds through what is built on it.
How to stop attribution debt accumulating
Paying it down once is straightforward. Keeping it down takes four habits, none needing tooling, all of which fit inside a meeting you already hold.
Explaining attribution debt to the rest of the business
Finance understands debt intuitively. Borrowing that vocabulary makes a measurement problem legible to exactly the people who decide whether to fix it.
What an AI agent needs from attribution data
An agent inherits whatever certainty your data implies. Four fields that stop a model turning an unresolved estimate into a confident recommendation.
What an agent can actually do with attribution data
Four tasks a model genuinely handles well once the data carries its uncertainty, and two it should not be given no matter how good the data gets.
The cost of a finding that does not land
A correct analysis nobody acts on costs more than the hours. It spends the credibility the next finding will need, and the next one usually matters more.
Defending a finding with ten minutes on the agenda
Ten minutes forces hard choices about what to leave out. What survives the cut, what goes in the appendix, and the two sentences that have to open it.
Your own record is what defends the finding
One report is a claim. A dated series of reports plus the decisions taken on them is a track record, and scepticism responds to those very differently.
Who to convince first about an attribution finding
Bringing a difficult finding straight to the full meeting is usually a mistake. The order of private conversations that gets it accepted instead of argued.
How to answer the objection you cannot answer
Sometimes the sceptic is right and your data cannot settle the question. Three responses that keep your credibility, and the one that ends it.
The recurring channel read a team comes to rely on
A report becomes load-bearing when people plan around it. Three properties that get you there, and the single change that undoes it in one cycle.
The bar a non-technical founder should set
Not being technical is not a reason to accept an unexplained number. Four things any vendor should be able to say plainly, and what it means when they cannot.
Holding your own in an agency attribution call
Five questions that work no matter how technical the room gets, and the one kind of answer that means the conversation has stopped being useful.
What continuous channel monitoring is actually for
Continuous measurement is not about getting the number sooner. It is about building a series long enough that this month means something against last month.
Alert fatigue and what it does to a marketing team
Alert fatigue does not just waste attention. It teaches a team that measurement is noise, and that lesson outlives the tool that taught it.
Own your alert rules, not the vendor defaults
Vendor defaults are calibrated for a median customer who is not you. Three settings worth overriding on day one, and what each one is quietly assuming.
What a refund guarantee signals about a vendor
A guarantee is a costly signal. What a vendor's willingness to carry the risk of being wrong tells you about its confidence, and what its absence tells you.
What you keep when you take the refund
Getting your money back is not the same as getting your position back. Three things to settle before paying, so that leaving costs only the fee.
Attribution refund windows expire quietly
Refund windows are missed by drift rather than decision. Three ways evaluations quietly overrun, and the two-line calendar habit that prevents each of them.
What an attribution channel does to a team
Putting measurement in a shared channel changes how people behave long before it changes any budget. Three effects, one of which is worth actively preventing.
What a real refund guarantee in analytics means
A guarantee that only covers the first fourteen days of an annual contract is a cancellation policy with confident wording. Four properties that tell them apart.
Attribution in Slack: what to post and what to skip
Analytics channels get muted for predictable reasons. Four rules about what earns a post, and the single category of update that guarantees people stop reading.
Attribution belongs where the decision happens
The distance between where a number is produced and where a decision is made is where most measurement work dies. Closing it is a design problem, not a tooling one.
Showing the reported vs causal ROAS gap to a team
Present a ROAS gap badly and every channel owner hears a performance review. Four framings that keep the conversation about the instrument rather than the person.
The Shopify channel lift you are not measuring
A channel missing from your reporting is not a channel with no effect. Three ways channels vanish from the numbers, and what each disappearance costs you.
Multi-touch attribution coverage is eroding now
Path-based attribution degrades quietly, because the model keeps reporting on whatever it can still see. Twenty minutes to measure how much of your own it has lost.
The cost of keeping multi-touch attribution in the EU
Running a path-based model on a European customer base carries four costs. Three are quiet, one arrives all at once, and none of them appear on the invoice.
Explaining a move away from multi-touch attribution
Someone in the room approved the tool you are replacing. Three arguments that work, one that never does, and how to structure it so nobody has to be wrong.
What a privacy-first attribution standard looks like
Privacy-first is either a property of the method or a sticker on the marketing site. Four tests that tell you which one you are being sold.
What you can answer without multi-touch attribution
Losing path-level data does not lose you the budget question. Five things you can still answer, two you genuinely cannot, and a way to tell them apart.
What a pixel tells you that a GA4 export does not
We sell an export-based read, so here is the uncomfortable list: five things a pixel genuinely gives you that an export does not, and what each one is worth.
Explaining pixel-free attribution to your developer
The developer review is where measurement tools die. Four questions engineering always asks, the honest answers, and the one objection worth taking seriously.
The Shopify lift number your agency will accept
The argument with your agency is never really about the number. It is about who chose the method. Four things to agree in advance so the result survives contact.
The attribution report a CFO will subscribe to
Nobody in finance is going to log into your attribution tool. The report that gets read is the one that arrives, fits on a page, and says how sure it is.
Own the pipe: attribution API keys and exports
A number you cannot export is a number you are renting. What to check about keys, exports and retention before your attribution history becomes someone else's asset.
Attribution that adds nothing to your storefront
There are two ways to measure a channel: collect more data about shoppers, or work harder on the data you already have. Only one of them adds code to your checkout.
What an attribution API returns that dashboards hide
Dashboards are edited. The API response is not. Four fields that tend to exist in the payload and never make it onto the screen, and what each one tells you.
What a good attribution reporting API returns
Automating a bad number just produces the bad number faster. Four fields separate an attribution API worth wiring into your reporting from one that is a schedule around folklore.
Omnichannel Attribution With Klaviyo in the Stack
Klaviyo counts orders from recipients who engaged inside its lookback. Meta and Google count the same orders under their own rules. Nobody is wrong, and the totals still cannot all be true.
Which Customer Journey Attribution Model Should You Use?
Five attribution models, one dataset, five answers. The choice between them is a policy decision about how to share credit, and no amount of picking well converts an allocation into evidence about cause.
Omnichannel Attribution: One Sale, Every Channel Claiming It
In an omnichannel stack every platform reports the same sale under its own rule and its own window. The sum exceeds the orders, and the size of the excess is the most useful number in your reporting.
Customer Journey Attribution and What It Cannot Tell You
Customer journey attribution divides credit across the touchpoints a system observed. It is a bookkeeping rule chosen by whoever configured it, and no rule turns an observed sequence into evidence of what caused the sale.
Meta Wants 50 Events in 7 Days. Cyber Week Is Four
Meta documents that an ad set's learning period stabilises with a minimum of 50 events over a 7-day period, and that edits can restart it. Cyber Week is four days long. The arithmetic gives you a launch date of 20 November and a rule about not touching anything.
Attribution Tools With Automated GA4 Ingestion and an API
'GA4 integration' can mean an OAuth connection, a scheduled export, or a pixel that ignores GA4 entirely, and 'API access' can mean anything from a CSV download to a keyed endpoint. The four questions that make the claims comparable.
Causal vs Rule-Based vs Data-Driven Attribution in GA4
GA4 has offered two kinds of attribution since 2023: rule-based last click and data-driven. Both distribute credit over observed journeys. Causal attribution answers a different question, and the difference is a counterfactual, an interval and a coverage rate.
How to Diagnose Over-Attribution in Meta and Google Ads
Every ad platform counts orders it touched inside a window it defines, and none subtracts what another platform also claimed. Four checks, using orders as the denominator, size the over-attribution per platform and tell you which channel's number is the most inflated.
How to Settle Attribution Debates With Your Agency
Agencies and platforms are both paid on the same reported number, and the number grades itself. The debate ends when the brand owns the scoreboard: orders from the store, coverage, the claim ratio, and one holdout the agency helps design but cannot reinterpret.
Klaviyo Will Claim Your Black Friday. How Much to Believe
In Cyber Week nearly every buyer opened or clicked a message inside the attribution window, so the email platform's share looks enormous. It is a claim under a rule, not a cause. The cheapest holdout in your stack says how much is real.
The Black Friday Post-Mortem You Cannot Run in January
By January the data behind Cyber Week is a model of itself. Meta's windows have closed, GA4's default retention has expired the events, and nobody remembers what was live on the Sunday. The post-mortem window is the week of 7 December.
Black Friday 2026: Four Measurement Deadlines Come First
Seventy-nine days out, three of the four dates that decide whether you can measure Black Friday are already inside the next seven weeks. The calendar, with the arithmetic behind each date.
Meta's 7-Day Click Window Will Claim Black Friday Sales
Under the default setting, a click on 20 November is credited with a purchase on Black Friday, and a view on 26 November with a purchase on the 27th. The window, the March 2026 definition change, and the number to compute instead.
You Don't Own Your Attribution Data. Here's What That Costs You.
Ecommerce sellers have learned to own the storefront and rent the discovery. Measurement is the asset they still rent without noticing - and the bill comes due the day a platform changes its mind.
Selling on Amazon and Your Own Store: The Attribution Blind Spot
Your ads sell on Amazon and your store, but each only sees half the picture. That gap is where budget leaks.
TikTok Shop's Native Checkout and the Credit Your Ads Never See
When the sale never leaves the app, the app writes the whole story. Your other channels get edited out.
When the Feed Finds the Buyer: Who Gets Credit for the Sale?
In a feed-driven sale, credit goes to whoever showed up last. That is a receipt, not attribution.
From Search Intent to Feed Discovery: What It Did to Your Attribution
The way people discover products flipped from search to feed. Your attribution model did not get the memo.
The Attribution Library: Every Guide, Organized
Every attribution guide on Causality Engine in one place: platform playbooks, model explainers, tracking and privacy deep-dives, and fixes: organized by topic.
ChatGPT Ads Are Live, and GA4 Cannot See Them: The 2026 Measurement Gap
ChatGPT Ads went self-serve on May 5, 2026, with a pixel and a Conversions API, but there is no query-level reporting and GA4 has no channel for the clicks. Here is how to measure the channel anyway.
Consent Mode v2 Is Fully Enforced: What EU Brands Lost and How to Measure Anyway
Google now fully enforces Consent Mode v2 across the EEA, and most EU stores directly measure a minority of their traffic. Here is what each measurement layer still sees, and why a causal read on consented data beats a modeled guess.
Post-Purchase Surveys Meet Causal Reads: The Triangulation Playbook
Post-purchase surveys now capture AI referrals, but self-reported answers are biased in predictable ways. This is the three-way triangulation framework that turns survey responses into budget decisions.
Why Your Direct Traffic Grew: Finding AI Referrals Hiding in GA4
AI-assistant sessions often arrive with no referrer and quietly inflate Direct in GA4. This self-audit recovers them, then connects them to revenue causally.
Meta's March 2026 Attribution Overhaul: Why Your Reported Conversions Dropped
On March 3, 2026, Meta restricted click-through attribution to link clicks only and replaced view-through with a one-day engage-through window. Reported conversions dropped because the definitions changed. Here is the post-mortem.
Every 2026 Attribution Change That Moved Your Numbers: The Changelog
A running, reverse-chronological changelog of every platform attribution change in 2026 that moved reported numbers for ecommerce brands, with what changed, who it affects, and what to do. Last updated: July 2026.
SKAN for DTC in 2026: A Decision Tree for App Spend Under €50K a Month
Most SKAN advice is written for app-first companies with enterprise budgets. This decision tree shows when postbacks alone are enough, when an MMP pays off, and when to fix creative first.
Who Gets Credit When an AI Agent Buys? An Attribution Framework for Agentic Commerce
ChatGPT Instant Checkout died within six months, but agent-mediated buying is returning through Google and Shopify infrastructure. Here is a taxonomy for what attribution can and cannot claim when an agent touches the order.
Attribution for the Quarter You Already Ran
Attribution for the Quarter You Already Ran. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution When Your Pixel Broke Three Months Ago
Attribution When Your Pixel Broke Three Months Ago. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution Without a Data Team - the Pragmatic Path
Attribution Without a Data Team - the Pragmatic Path. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution Without Engineering Help - the Marketer's Guide
Attribution Without Engineering Help - the Marketer's Guide. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
An Attribution Proof-of-Concept for €99 - the Fastest Way to Test a Tool
An Attribution Proof-of-Concept for €99 - the Fastest Way to Test a Tool. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
How to Test an Attribution Tool Without Installing Anything
How to Test an Attribution Tool Without Installing Anything. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
How Every iOS and Cookie Update Resets Your Attribution Clock
How Every iOS and Cookie Update Resets Your Attribution Clock. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Attribution Pixels Keep Breaking - and How to Stop Caring
Why Attribution Pixels Keep Breaking - and How to Stop Caring. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
The 2-to-12-Week Attribution Setup Trap
The 2-to-12-Week Attribution Setup Trap. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution in 1 to 2 Minutes - Not 5 to 10 Weeks
Attribution in 1 to 2 Minutes - Not 5 to 10 Weeks. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution for a Campaign That Already Ended
Attribution for a Campaign That Already Ended. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Retroactive Attribution Analysis From Your GA4 Export
Retroactive Attribution Analysis From Your GA4 Export. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution for Historical Data - When the Pixel Wasn't Even Installed
Attribution for Historical Data - When the Pixel Wasn't Even Installed. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution Without Touching Your Shopify Theme
Attribution Without Touching Your Shopify Theme. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution Without an SDK or Storefront Code
Attribution Without an SDK or Storefront Code. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Attribution Without a Pixel - the 2026 Handbook
Attribution Without a Pixel - the 2026 Handbook. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Attribution Setup Takes Months - and How to Skip It Entirely
Why Attribution Setup Takes Months - and How to Skip It Entirely. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
One Dashboard, One Number: Causal Attribution on Your GA4 Export
One Dashboard, One Number: Causal Attribution on Your GA4 Export. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
A Single Source of Truth for Marketing Attribution - Is It Actually Possible?
A Single Source of Truth for Marketing Attribution - Is It Actually Possible?. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
The Correlation vs Causation Problem in Marketing Attribution
The Correlation vs Causation Problem in Marketing Attribution. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Cannibalisation vs Incremental Revenue - Why MTA Tools Can't Tell the Difference
Cannibalisation vs Incremental Revenue - Why MTA Tools Can't Tell the Difference. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Your Google Ads ROAS Is Different From Your MTA Tool
Why Your Google Ads ROAS Is Different From Your MTA Tool. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Your TikTok ROAS Doesn't Match Anything Else You're Looking At
Why Your TikTok ROAS Doesn't Match Anything Else You're Looking At. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Your Meta ROAS Is Inflated - and Roughly By How Much
Why Your Meta ROAS Is Inflated - and Roughly By How Much. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Self-Attribution Bias: Why Meta, Google and TikTok Overclaim
Self-Attribution Bias: Why Meta, Google and TikTok Overclaim. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Cookie Deprecation and Your Attribution Tool - the Survival List
Cookie Deprecation and Your Attribution Tool - the Survival List. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
iOS 14 and Why Your Attribution Broke - the Survivor's Guide
iOS 14 and Why Your Attribution Broke - the Survivor's Guide. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
The UTM Tagging Problem - Why Your Attribution Is Already Wrong
The UTM Tagging Problem - Why Your Attribution Is Already Wrong. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Multi-Touch Attribution vs Last Click - Why the Numbers Disagree
Multi-Touch Attribution vs Last Click - Why the Numbers Disagree. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Every Attribution Dashboard Tells You a Different Story
Why Every Attribution Dashboard Tells You a Different Story. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Fixing Attribution Mismatch Between Platforms - the Playbook
Fixing Attribution Mismatch Between Platforms - the Playbook. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Your Attribution Dashboards Disagree - Which One Should You Trust?
Your Attribution Dashboards Disagree - Which One Should You Trust?. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Shopify Doesn't Match Google Ads - and What to Do About It
Why Shopify Doesn't Match Google Ads - and What to Do About It. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Meta Doesn't Match GA4 - the Self-Attribution Problem
Why Meta Doesn't Match GA4 - the Self-Attribution Problem. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why GA4 Doesn't Match Triple Whale - and Which One You Should Actually Trust
Why GA4 Doesn't Match Triple Whale - and Which One You Should Actually Trust. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Northbeam Doesn't Match Meta - Explained Plainly
Why Northbeam Doesn't Match Meta - Explained Plainly. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
Why Triple Whale Doesn't Match Shopify - the Real Reason
Why Triple Whale Doesn't Match Shopify - the Real Reason. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
The Attribution Maturity Model: From Google Analytics to Causal Intelligence
Stop guessing with Google Analytics. The Attribution Maturity Model shows how causal inference measures real impact rather than just clicks.
LLMs Make Aggregation Errors: Why SUM, AVG, and COUNT Go Wrong
An average of ratios is not a ratio of totals, and daily distinct counts do not add up to a period count. A peer-reviewed error study and Google's own guidance show where aggregates go wrong, and a short query checks yours.
How to Test an LLM on Your Own Attribution Data
Write questions whose answers you can compute from your own GA4 and Shopify exports, ask the LLM each one three times, and grade its results against your answers. Add two causal questions and grade them on whether the tool admits the export cannot settle them.
Real-Time Attribution in a Cookieless World: Is It Still Possible?
Real-time attribution isn’t dead: it’s just broken.
LLM Confidence vs. Accuracy: Is a Sure Answer a Right One?
Not reliably: in published evaluations, models asked to state a confidence were overconfident, and people trusted longer explanations more. Tally stated confidence against answers you can recompute.
Attribution Windows in the Cookieless Era: Why 7 Days Is Not Enough
They were always a compromise, but in the cookieless era, they’re a delusion. The problem isn’t the window: it’s the entire framework.
Building Your Attribution Stack Without Cookies: A Technical Blueprint
The 3rd-party cookie deprecation isn’t a future threat; it’s a present reality. If your stack still relies on cookies, you’re measuring ghosts.
Google Ads Measurement After Cookie Deprecation: A Survival Guide
Google Ads attribution is broken without cookies.
Zero-Party Data Attribution: Letting Customers Tell You How They Found You
Cookieless attribution doesn’t mean guesswork.
YouTube View-Through Attribution: Measuring What Happens After the Watch
YouTube view-through attribution is broken. Causal inference unlocks true incremental sales from video.
LLM Analytics Vendor Lock-In: When Your AI Provider Changes the Model
Providers retire models on published schedules: at least 60 days at Anthropic and at least 6 months at OpenAI for released models, "such as 2 weeks" for OpenAI previews. Pin the model ID and keep a question set to re-test.
Can ChatGPT Replace Your Attribution Platform?
A chat helps with drafting and explaining. For the number that moves budget, OpenAI's own documentation, two 2026 data-agent benchmarks and a five-step test on your own data show what to check.
The Unified Measurement Framework: Combining MMM, MTA, and Incrementality
Stop pitting MMM, MTA, and incrementality against each other. The unified measurement framework uses causal inference to merge all three: no cookies required.
What Is Cookieless Attribution? The 2026 Marketer's Guide
Cookieless attribution isn’t just a workaround: it’s the future.
AI Chat in Attribution Tools: What to Ask the Vendor
A chat box changes how you ask, not how the number is produced. What a 2026 benchmark and vendor documentation show, and five tests to run before you trust an attribution tool's AI chat.
LLM SQL in BigQuery, Snowflake and Redshift: What Differs
A May 2026 preprint found most dialect failures were valid SQL with a different meaning, not syntax errors. Integer division is the clearest case, and one query tells you how your warehouse behaves.
Can an LLM Tell If an Attribution Result Is Significant?
Only if you hand it two groups to compare, and you should still recompute. A published benchmark found LLMs primarily err on choosing a valid test; four counts and one formula let you check the verdict in a spreadsheet.
What the Spider 2.0 SQL Benchmark Says About Marketing Data
It grades SQL against gold answers on enterprise databases, including Google Analytics ones, and its scores have moved a long way since the first release. It does not grade what a channel caused, which no gold answer in your export can.
Ten Questions to Ask Any Attribution Tool, LLM or Not
Ask which rule produced the number, then reconcile its totals to Shopify and GA4, rerun it, and ask what test would change its mind. Studies of ad experiments show last-click credit and observational methods often miss the experimental result.
Third-Party Cookies Are Dead. Your Attribution Should Be Too.
Third-party cookies are deprecated, and your attribution model based on them should be too. Traditional attribution, reliant on stitching together fragmented user journeys across the web, is crumbling.
TikTok Attribution: How Ads Manager Counts and How to Check
TikTok's Ads Manager counts purchases inside a window and dates them by the impression. What its help pages say, what experiments show, and five steps to check the number against Shopify.
Prompt Engineering for Attribution: The Myth of the Perfect Question
A better prompt changes how a question becomes a query, not the answer key attribution lacks. Ask one question five ways on your own export and see which numbers survive.
Propensity Score Matching for Attribution: Comparing Apples to Apples
Propensity score matching (PSM) claims to fix attribution bias. But does it actually work? We dissect PSM, expose its flaws, and offer a real solution: causal inference.
Does Your LLM Know Your Schema? A GA4 and BigQuery Check
A GA4 export keeps its width inside repeated records, and Google says the parameter set is unique to each implementation. Two studies measure schema difficulty for models, and a dry run checks a query before it touches your data.
Server-Side Tracking vs. Cookies: Which Actually Measures Revenue?
Server-side tracking captures every transaction, but only when paired with causal inference. Without it, you’re trading one broken system for another.
Probabilistic vs. Deterministic Attribution: Which Survives Without Cookies?
Probabilistic attribution crumbles without cookies. Deterministic holds shape but starves.
The Security Risk of Feeding Marketing Data to LLMs
The risk is what the model can see and do, and prompt injection has been measured. Give an assistant a view of aggregates, a read-only identity and a canary test, not your warehouse.
Marketing Mix Modeling Explained: The Cookieless Attribution Method CFOs Love
Marketing mix modeling (MMM) is a top-down, cookieless attribution method that uses statistical regression to quantify the impact of various marketing activities on sales.
Natural Language to SQL for Marketing: The Gap Between Demo and Reality
The Spider 2.0 leaderboard now lists far higher scores than its own introduction quotes, while an audit found annotation errors in 76 of 121 checked examples. Your store's definitions matter more than a score, and you can check them yourself.
Meta Ads Attribution Without Cookies: What the Platform Won't Tell You
Meta’s cookieless attribution is a black box.
LLM-Written Joins: How Orders Go Missing or Count Twice
An inner join discards rows with no match, and a join to a child table repeats the parent. Three counts before and after the join show which happened to your orders, and three papers show how model-written joins fail.
Podcast Attribution: How to Measure What No Cookie Can Track
They just don’t show up in your dashboard. While marketers obsess over click-through rates and last-touch models, podcast listeners convert offline, in-app, or days later: places no cookie can follow.
Can an LLM Read the Time Series in Your Attribution Data?
Two published tests of LLMs on time series disagree, and neither is about attribution. Check the dates first: GA4 puts each event on a day in the property's time zone, and credit can change for up to 12 days.
Where LLMs Actually Help in Marketing, With a Check for Each
LLMs help where the output is a draft you can check without asking the model again. What measured experiments show, and a check for each of five marketing tasks.
LLMs and Duplicate Conversions: A GA4 transaction_id Check
GA4 de-duplicates purchases that share a transaction ID, only when the ID is present, unique and from a web stream. Google's documentation sets the rules, and a short BigQuery check shows whether your export repeats orders.
Can You Trust a Chart an LLM Made From Your Data?
Only as far as you can trace each bar to rows in your export: a chart benchmark found LLM charts that ran but did not match the request. Compare each bar with a pivot of the same export before sharing.
LLM Bias in Marketing Data: Does Wording Steer the Answer?
Yes, it can. Studies report that assistants tend to agree with the person asking and that answers shift with option order and prompt format. Flip your framing and reorder your data to test yours.
LLMs, Time Zones and Currencies: Why Totals Differ
In the GA4 BigQuery export, event_date follows the reporting time zone, event_timestamp is UTC and purchase values stay in local currency. Documentation and a peer-reviewed bug study show why totals differ, and two queries test yours.
Is an LLM Too Slow for Attribution? Check the Data Lag First
Usually not: the model is one of four clocks and the one you can set, with response times from seconds to minutes. GA4 data takes 24 to 48 hours and credit can change for up to 12 days, so time all four.
LLM Hallucinations in Marketing Data: When Your AI Invents Revenue
Yes, an LLM can state a revenue figure that appears nowhere in your data. Recompute any total it gives you from your export, and reconcile it with Shopify and GA4, before it reaches a budget.
LLM Attribution and Compliance: GDPR, CCPA, and Your Data
Sending order or customer rows to an LLM provider is processing on your behalf, and GDPR Article 28 requires a contract with specific terms. List the columns you sent; a channel-by-day table has no per-person row.
iOS 14.5 Killed Your Attribution. Here's What to Do About It.
iOS 14.5 didn't just tweak your attribution model; it detonated it. The old methods of tracking and 'attributing' value are now about as reliable as a psychic reading.
Should You Use an LLM to Analyze Your Attribution Data?
Yes for drafting and explaining, not for taking the number on trust. What SQL and data-agent benchmarks measure in 2026, what they don't show about your export, and a five-step check.
Influencer Marketing Attribution: Beyond Discount Codes and UTM Links
Measuring influencer spend properly means measuring incrementality. Discount codes and UTM links record who clicked, not who would have bought anyway. Here is how Triple Whale, Northbeam, Hyros, Haus, Measured and Causality Engine compare on price and fit.
LinkedIn Ads Attribution for B2B: Measuring the Unmeasurable
LinkedIn's reports can't see deals that close offline months later, so check them against closed-won deals in your CRM. To learn what the ads add, pause them in a slice of your target markets for 30 days and compare conversion rates.
Does an LLM Leave an Audit Trail for Attribution Numbers?
Not by default: the trail that holds up is the one you write. Log the prompt, the pinned model ID, the date, the SQL and the result, and test that you can reproduce the number.
The Future of AI in Marketing Analytics: Assistants, Not Analysts
Assistants draft and explain; a person, or a method you can check, owns any number that moves budget. That is an opinion about roles, not a forecast: nothing here predicts adoption or which tools win.
Incrementality Testing: The Only Way to Know If Your Ads Actually Work
Stop guessing if your ads work. Incrementality testing reveals true ad impact with causal inference, not cookies. Learn how to measure lift accurately in a cookieless world.
GDPR-Compliant Attribution: Measuring Without Violating Privacy
GDPR compliance and accurate attribution don't have to be enemies. Discover how causal inference delivers privacy-compliant tracking and incrementality.
The Future of Marketing Measurement: 5 Predictions for 2027
The future of marketing measurement isn't about better attribution; it's about replacing attribution altogether. Buckle up; the next few years will be a wild ride.
GPT-4 and Enterprise SQL: What the Benchmark Numbers Mean
They describe one model snapshot, inside one wrapper, on one benchmark, on one date: the same method scores far higher on an older SQL benchmark than on the newer enterprise one. Find your model's exact string on the benchmark's own page before you quote a score.
Identity Resolution Without Cookies: Stitching the Customer Journey
Cookies are dead. Identity resolution without cookies requires causal inference, not guesswork.
Browser Fingerprinting Is Not the Answer to Cookieless Attribution
Browser fingerprinting won't save your attribution. It's brittle, unethical, and inaccurate. Learn why causal inference is the only future-proof solution.
Your Data Analyst vs. an LLM: Who Should Own Attribution?
A named person should own attribution, and that person can use an LLM. The definitions, the reconciliation to Shopify and the sign-off stay human; here is what four benchmarks measure and a test to run on your own data.
Data Clean Rooms for Attribution: Promise vs. Reality
Data clean rooms promised privacy-safe attribution. The reality? They're expensive, complex, and still rely on flawed correlation. Causality Engine offers a better way.
Econometric Attribution: What Regression Can and Cannot Show
A regression gives a causal answer only if spend is as good as random once your controls are in. What that assumes, what experiments found when they checked, and how to test your own model.
First-Party Data Attribution: How to Build a Strategy That Actually Works
First-party data attribution fails without causal inference.
Marketing Measurement in a Cookieless World: What Changes and What Doesn't
Cookies are crumbling. But the fundamentals of marketing measurement? They're not. Learn what's changing and what absolutely *isn't* in cookieless marketing.
The Cost of Wrong Attribution: What Happens When You Trust an LLM with Budget Decisions
It costs whatever you moved on the strength of the number. The net on each extra euro is incremental ROAS times margin, minus one; here it is worked with your own numbers, with the field evidence on how far off the number can be.
Cookieless Attribution for SaaS: Measuring Long Sales Cycles Without Cookies
You absolutely can achieve SaaS attribution without cookies. Stop relying on broken attribution models and embrace causal inference.
Cookieless Attribution for Supplement Brands: Measuring Health and Wellness Marketing
Cookieless attribution for supplement brands is not optional. If you’re still relying on last-click or multi-touch attribution, you’re flying blind. Causal inference measures what your ads caused.
Cookieless Attribution for Travel Brands: Measuring the Dreaming-to-Booking Journey
Without cookies, travel brands lose track of much of the dreaming-to-booking journey. Causal inference measures what each channel caused instead of guessing.
Cross-Channel Attribution Without Cookies: Connecting the Full Journey
If you’re still relying on third-party cookies or last-click models to stitch together customer journeys, you’re not measuring marketing: you’re guessing.
Can LLMs Tell Correlation From Causation in Attribution?
Not from your data alone. One published test found seventeen LLMs near random at inferring causation from correlation statements, and observational methods often failed to match ad experiments. A holdout test is how you check.
Cookieless Attribution for Fashion Brands: From Inspiration to Purchase
Fashion attribution without cookies is here. Learn how causal inference tracks customers from first inspiration to final purchase, even with privacy changes.
Cookieless Attribution for E-Commerce: A Practical Guide
Ecommerce tracking without cookies is here. Stop guessing with last-click. Use causal inference to measure real impact, not just clicks.
Cookieless Attribution for Mobile Gaming: Measuring Installs and In-App Revenue
Mobile gaming attribution is broken. Cookies are dead, SKAdNetwork is a blunt instrument.
Cookieless Attribution ROI: How to Calculate the Cost of Bad Measurement
Bad attribution makes you misallocate ad spend. Learn how to calculate the cost of cookieless measurement failures and fix it with causal inference.
Cookieless Attribution for Fintech: Compliant Measurement in Regulated Markets
Fintech attribution is broken. Cookies are crumbling. Compliance is key. Learn how cookieless causal inference delivers accurate, privacy-safe measurement for financial services marketing.
Cookieless Attribution for Food and Beverage: Measuring Taste Without Tracking
Ditch broken cookies. Learn how food beverage brands are using cookieless attribution to measure real-world impact and refine marketing ROI. Causal inference > guesswork.
Cookieless Attribution for Luxury Brands: Measuring Aspiration, Not Just Clicks
The death of third-party cookies isn't a crisis; it's an opportunity. For luxury brands, clinging to outdated attribution models based on clicks is like judging a masterpiece by the number of brushstrokes.
Cookie Consent Rates and the Attribution Gap: What Your Data Misses
Cookie consent rates are falling, and click-based attribution misses the visitors who opt out. Learn how causal inference closes the attribution data gap.
Contextual Targeting and Attribution: The Privacy-First Performance Model
Contextual targeting is back, but traditional attribution can't measure it. Learn how causal inference unlocks privacy-first performance measurement.
Will Your Marketing Data Fit in an LLM Context Window?
Multiply your row count by your tokens per row and compare it with the window on your provider's page, which is 1M tokens for some current models. Then test retrieval, because fitting is not finding.
Cookieless Attribution for DTC Brands: Scaling Without Tracking
DTC brands can scale without cookies.
Cookieless Attribution for Beauty Brands: Measuring the Full Funnel
Beauty is a high-consideration category. The average skincare buyer visits 7.2 touchpoints before purchasing, according to McKinsey.
The CMO's Guide to the Cookieless Transition: Budget, Timeline, and Quick Wins
CMOs face a $200B attribution crisis. This guide delivers a 90-day cookieless transition plan with budgets, timelines, and causal inference-powered quick wins to protect revenue.
Connected TV Attribution: Measuring the Living Room Without Cookies
Connected TV attribution is broken in a cookieless world.
Cohort-Based Attribution: What Google's Privacy Sandbox Means for Marketers
Google’s Privacy Sandbox kills cookies. Cohort-based attribution claims to fix it: but fails.
The CMO's Guide to AI Attribution Hype: What to Believe and What to Ignore
Believe a claim that names its baseline, its test and its date, and ignore a headline accuracy figure. Five questions to put to any AI attribution pitch, with what a pass and a fail look like.
Causal Attribution vs. Tracking: Why the Best Attribution Doesn't Need Cookies
The future of attribution is cookieless. Tracking provides a fragmented view of user activity.
GA4 or an LLM for Attribution: Which One Can You Audit?
Neither is fully open. Google documents GA4's three attribution models and lets you switch between them; an LLM answer is only as traceable as the record you keep. Re-derive one number from your own rows to audit either.
How Browser Privacy Updates Broke Marketing Attribution
Browser privacy updates didn’t just break marketing attribution. They set it on fire, poured gasoline, and handed marketers a blindfold.
Would You Present LLM-Generated Attribution Numbers to Your Board?
Only if someone else can reproduce each number and the slide says whether it was credited or measured. Give each figure a query, a run date, a model ID and a reconciliation to Shopify before the meeting.
Amazon Ads Attribution: The Walled Garden Problem
Amazon Ads attribution is broken. The walled garden approach gives you incomplete data and inflated metrics. Causality Engine unlocks true incrementality.
Bayesian Attribution Models: The Math Your Attribution Actually Needs
Bayesian attribution models offer a probabilistic approach, unlike deterministic models. Learn how Bayesian methods enhance accuracy and provide a more complete view.
Attribution Requires Causal Models, Not Language Models
Orders alone can't show what a channel caused. What the published evidence says about observational methods, where an LLM helps, and a region holdout you can run from a Shopify export.
Is the Future of Attribution Conversational or Causal?
A chat box is an interface, and the number it reads out has to be an estimate of what each channel caused. What published benchmarks show about language models and causation, and checks for your own export.
Linear Attribution Model Explained: Why It Lies About Your Ad Spend
The linear attribution model distorts your marketing data. Learn why this common model is flawed and what Dutch ecommerce brands should use instead.
Geo-Lift Testing for Ecommerce: A Practical Guide to Measuring Incrementality
Stop guessing your ad impact. This practical guide to geo-lift testing shows ecommerce brands how to measure true incrementality and make smarter budget decisions.
Why Your Attribution Platform Reports 4.5x ROAS but Revenue Says Otherwise
Your attribution platform shows a high ROAS, but revenue is flat. Discover the truth behind why your marketing attribution platform is failing you and how to fix it.
Incrementality Testing: The Only Way to Know If Your Ads Actually Work
Stop guessing your ad performance. Incrementality testing is the only way to measure the true causal impact of your marketing. Learn how.
The ROAS Trap: Why a 6x Campaign Might Be Worth Less Than a 2x Campaign
Escape the ROAS trap. Learn why a high ROAS can be a misleading metric and how to measure the true incremental value of your marketing campaigns.
Why Multi-Touch Attribution Models Fail Ecommerce Brands
Discover why the multi-touch attribution model is failing your ecommerce brand and what to do instead. Learn how causal inference provides a better solution.
Incremental Revenue vs. Attributed Revenue: The Gap Costing You Millions
Stop chasing attributed revenue. Learn the difference between incremental revenue and attributed revenue and why focusing on incrementality is the key to profitable growth.
How to Run a Holdout Test on Meta Ads Without Killing Your Revenue
Learn how to run a holdout test on Meta Ads to measure the true incremental value of your campaigns without sacrificing revenue. A guide for Dutch Shopify brands.
Marketing Attribution is Dead. Here's What Replaced It.
Traditional marketing attribution is obsolete: GA4 has already dropped four of its rule-based models. Here's what replaced it for Dutch ecommerce brands, and the free tests to start with.
Cross-Channel Attribution: Stop Platforms From Stealing Credit
Stop letting platforms take credit for sales they did not generate. Learn how true cross-channel attribution reveals which channels actually drive growth.
The Death of Attribution: Why Causal Attribution Is the Replacement
Click-based attribution can't tell you what your ads caused. Why causal measurement replaces it, and how to start with one holdout test.
Why Your CFO Does Not Trust Your ROAS Numbers (and How to Fix It)
Your CFO doesn't trust your ROAS reporting. Learn why and how to fix it with causal inference to improve ROAS credibility and marketing ROI reporting.
Attribution After iOS 14.5: What Actually Works in 2026
iOS 14.5 broke marketing attribution. Discover what actually works for attribution after iOS 14.5 and how to measure marketing ROI in a post-cookie world.
How to Calculate True ROAS When Every Platform Inflates Its Numbers
How to Calculate True ROAS When Every Platform Inflates Its Numbers Learn more about how to calculate true roas when every platform inflates its numbers and how it impacts your marketing attribution strategy.
How View-Through Attribution Distorts ROAS
View-through attribution gives false credit to ads that were never clicked, distorting your ROAS. Learn why it's a dangerous metric for Dutch ecommerce brands.
First-Touch vs. Last-touch Attribution: Why Both Are Wrong for Ecommerce
First-touch and last-touch attribution models provide a distorted view of your marketing. Discover why these single-touch models are wrong and what to use instead.
Data-Driven Attribution vs. Causal Attribution: The Critical Difference
Data-driven attribution models are correlation engines, not truth machines. Discover the critical difference causal attribution makes for your marketing budget.
Blended ROAS Is a Lie: Here Is What to Track Instead
Blended ROAS is a misleading metric that hides the true performance of your marketing. Discover what to track instead for real growth.
What Is ROAS and Why It Is the Most Dangerous Metric in Marketing
What Is ROAS and Why It Is the Most Dangerous Metric in Marketing Learn more about what is roas and why it is the most dangerous metric in marketing and how it impacts your marketing attribution strategy.
Your platforms guess.
We run the math.
Upload a GA4 export and see what each channel caused, next to last-click, in 1–2 minutes. The read is yours to keep.
Free, in your browser: your file is not uploaded. The full read is €99, refundable within 30 days. Prices exclude VAT.
Or book a 30-min call.