Skip to content
Vs Measured

Run the experiment, or read the history.

Measured is a serious incrementality platform. It runs geo-based holdout experiments: hold spend back in matched regions, keep it running elsewhere, and measure the difference in outcomes. That is a real experiment, and a well-designed geo test is the strongest evidence of causal lift a marketer can buy. We are not going to pretend otherwise.

The trade is cost, latency, and what you have to give up to get the answer. Measured is enterprise-priced and quoted via sales, each test takes roughly four to eight weeks, and the method requires deliberately withholding spend in some regions. You also learn about the window you tested, not the year you already spent.

Causality Engine answers the same question from the opposite direction: causal inference over the GA4 history you already have. No experiment to design, no spend to pause, no region to sacrifice. A per-channel causal read with confidence intervals in 5 to 10 minutes, for €99.

// Measured answers
hold spend back in matched geos, wait 4-8 weeks  → measured lift for the tested window
// Causality Engine answers
read the GA4 history you already have  → incremental ROAS + confidence interval, 5-10 min
The difference

Same question, different answer. Often the opposite answer.

 Causality EngineMeasured
Question answeredWhat did each channel cause?What did this channel cause, in this test?
MethodCausal inference on existing GA4 historyGeo holdout experiment (prospective)
Evidence strengthObservational, confidence-interval boundedExperimental, stronger when well powered
Spend paused to learnNoneYes, in the holdout regions
Time to first number5–10 minutes4–8 weeks per test
Price€99 per read · €299/mo ProEnterprise contract, quoted via sales
Covers past spendYes, any historical periodNo, only the window you test
By scenario

Which one is the right call for your situation.

Choose Measured when

you have the budget, the spend scale to power a geo test, and a measurement team that will run a formal programme. If a nine-figure media plan turns on the answer, buy the experiment. Observational inference is not a substitute for a properly powered randomised test, and anyone who tells you otherwise is selling something.

Choose Causality Engine when

you want an answer this week rather than next quarter, you cannot justify an enterprise contract to find out, or you need to know about the spend you have already made. It also works as the cheap first pass: run a €99 read to find which channels look non-incremental, then spend the geo-test budget confirming only those. That sequencing costs less and tests the right things.

Take it to your CFO

An enterprise contract quoted via sales, with a four-to-eight-week wait for the first answer, versus €99 on a card with the answer in ten minutes. If the €99 read and the geo test disagree, that disagreement is itself worth knowing, and you found it for €99.

FAQ

The questions buyers ask about us vs Measured.

Is a geo holdout test more accurate than causal inference on GA4 data?
A well-designed, adequately powered geo test is the stronger evidence, because randomisation removes confounders rather than adjusting for them. Causal inference on observational history is bounded by the assumptions in the model, which is why our output carries confidence intervals. The practical question is not which is theoretically better but which you can actually run: a test you cannot afford or power teaches you nothing.
Can I use Causality Engine and Measured together?
That is the sequencing we would recommend if you have both budgets. Use a €99 causal read across all channels to find the candidates that look non-incremental, then spend the geo-testing budget confirming those specific channels. Testing everything is expensive; testing the suspects is not.
Do I have to pause ad spend to use Causality Engine?
No. That is the main operational difference. Geo holdouts require withholding spend in some regions to create a control group. We infer the counterfactual from your existing GA4 history, so nothing has to be turned off and no revenue is deliberately foregone to produce the measurement.
How much does Measured cost compared to Causality Engine?
Measured is enterprise-priced and quoted via sales; it does not publish a price list. Causality Engine is €99 per causal read with no subscription, or €299/mo for Pro. The gap in commitment is why the two tend to serve different-sized brands.
Named comparison

Every causal vs MTA tool in one row each.

They track clicks. We measure cause. Four of these five tools are pixel or multi-touch-attribution dashboards on an annual contract; they reassign credit for conversions that already happened, and they need a script installed and maintained. Causality Engine measures incrementality, the sales a channel actually caused, from your existing GA4 and Shopify export, no pixel, with confidence intervals, for €99 a read. Different question, different bill, different install.

ToolCore methodCausal?Pixel / SDK?Time to valuePricing modelYr-1 fully loadedContract
Causality EngineCausal inference (Bayesian) on GA4 + ShopifyYes, nativeNone5–10 min€99 per read · €299/mo Pro€99 one-time · €3,588/yr ProCredit card, cancel anytime
Triple WhalePixel + MTA (Clicks & Deterministic Views)NoYes (Triple Pixel)Days (after pixel install and data)Subscription, GMV tiersGMV-priced, quote by store sizeNot published
NorthbeamML MTA + MMM + incrementality (2026 add-on)Partial (newer layer)Yes (script + UTMs)Weeks (onboarding)Starter $1,500/mo; Pro/Ent custom quote$1,500/mo Starter; $3,500/mo Professional; higher tiers quotedNot published
RockerboxMTA + MMM + incrementality (enterprise)Yes (native)Yes (integration)Weeks (config + experiments)Quote onlyQuote only, plus onboardingNot published
HyrosServer-side click tracking + MTANoYes (watcher scripts, heavy)2–12 weeks (setup)Subscription, demo-gatedQuote only, demo-gatedNot published
Wicked ReportsFirst-party click MTA (LTV focus)NoYes (pixel + CRM/UTM)Days to weeksRevenue-banded: $499 / $699 / $999 / $4,999+$499–$999/mo by revenue bandMonthly billing; term not published
CometlyMTA pixel + server-side + CAPINoYes (Comet Pixel + CAPI)Days (post pixel install)Usage-based on pageviews; no free trialQuote only, usage-basedMonthly or annual (annual −20%)
Billy GraceDeep-learning MTA + Unified Marketing MeasurementModeled, no CIsYes (own first-party pixel)Demo-gated onboardingSubscription, by ad spend€499–3,250/mo + €99 integrationsDemo-gated
EYKDataServer-side MTA on owned BigQuery (correlation-based)Quote-only top tierYes (GTM + BigQuery + Stape)Weeks (stack stand-up)Subscription, by sales€399–€529+/mo (Advanced = quote)Subscription
Funnel.ioData integration / ETL platformAdd-on at quote-only top tierConnectors / warehouseWeeks (pipeline build)Base plan + paid Measure add-on$300–600 base + $2,250/mo Measure (≥$500k spend)Billed annually
SupermetricsMarketing data pipeline / connectorsNo (claims unverified)ConnectorsHours per sourcePer-source subscriptionFrom $44/mo billed annually; Growth from $177/moAnnual
ElevarShopify server-side conversion tracking + data layerNo (not attribution)Yes (data layer install)1–2 weeks (setup)Subscription, by orders$225–3,000+/mo + $1k–4.5k installMonthly
FairingPost-purchase survey (zero-party)No (self-reported)Survey widgetMinutes (install)Subscription, by ordersPlans by order volume; quote on fairing.coMonthly
Google AdsSelf-attributing network, data-driven attributionNo (window-bound, self-mediated)NativeBuilt inNative to ad platformFree (with the spend)Native, no contract
Meta Ads ManagerSelf-attributing, 7-day click + 1-day view defaultNo (view-through inflates)Native + Meta Pixel/CAPIBuilt inNative to ad platformFree (with the spend)Native, no contract
TikTok Ads ManagerSelf-attributing, 7-day click + 1-day viewNo (passive-feed view-through)Native + TikTok PixelBuilt inNative to ad platformFree (with the spend)Native, no contract
Shopify Marketing ReportsLast-click (last non-direct), rules-basedNo (erases assists)NativeBuilt inIncluded with ShopifyFree (with the plan)Native, no contract
GA4Data-driven attribution, click/visit-basedNo (ignores impressions, no control group)Native GA4 tagBuilt inFreeFreeNative, no contract

Competitor figures are our estimates of a fully loaded monthly cost in EUR, built from each vendor's published pricing (USD where quoted) on the date shown under this table. A tilde or "est." marks an estimate, "reported" marks a figure from a third-party report, and quote-only vendors are marked as such. Prices change often: verify with the vendor before relying on them. Not vendor quotes to us.

Run the side-by-side

Run the side-by-side. Watch the model work on a sample store first, no signup. Then upload your GA4 export and get incremental ROAS with confidence intervals in 5–10 minutes, €99. Refundable if the first read doesn't pay for itself in one reallocation decision. Keep your current tool; use us as the independent causal second opinion.

We're the only one in this table without an ad budget, or a dashboard subscription, to protect.