Does dark social count as direct traffic?

Usually yes. Dark social means links people share in private: chats, texts, direct messages and forwarded emails. If the click arrives with no tags and no referrer, GA4 files it under Direct. If the app passes its name, it can land in Organic Social. If GA4 already knows the visitor, it may wear their earlier campaign.

What one store's data shows

The usual answer goes one step further: most of your Direct traffic is dark social, so a big Direct row means people talk about you. It is a flattering theory. One store's export shows why no GA4 file can confirm it.

One store's anonymised GA4 export, 1 January 2024 to 21 August 2026. It holds shares of revenue only: no ad spend, no order counts.

On the Channels sheet, Direct holds 57.7% of revenue in every view. A link pasted from a group chat would sit in that row. So would a regular typing the address, a bookmark and an untagged newsletter. The row has no column that tells them apart.

On the Channels sheet, the touched view adds up to 110.4%. That total runs over the whole by design, because a journey that touched two channels counts in both. It runs over by little, so little of this one store's revenue came from journeys that mixed channels.

On the Journeys sheet, journeys with 1 touch hold 79.5% of revenue, and those buyers took 0.5 days to buy. One recorded visit, then an order within hours. That is what a friend's link looks like when it lands on a made-up mind. It is also what a loyal regular looks like.

On the Journeys sheet, journeys with 2 to 3 touches hold 12.2% of revenue and took 12.5 days to buy. A chat link could have opened some of those journeys too. If one did, the export shows a Direct first touch and nothing about the chat.

So this one store's file can tell you how much revenue arrived without a source. It cannot tell you how much of that was dark social. No GA4 export can, because the chat itself never reaches GA4.

Why does "Direct is mostly dark social" mislead?

Because Direct has many tenants, and dark social has more than one address.

Start with the tenants. Google defines Direct as arriving via a saved link or by entering your URL. Its examples include a link someone sends, which the reader copies and pastes into their browser. That is dark social by any definition. But the same bucket holds bookmarks, PDFs, untagged emails and other visits without a clear referral source. Where direct traffic really comes from maps the rest.

Now the addresses. One shared link can end up in four places:

The third row surprises people. In Google's own example, someone first arrives from Google and later comes back by typing your address. While those visits fall inside the key event lookback window, GA4 keeps them in the Google row. So a friend's link opened by an old customer can wear your search ad's label.

So the honest answer has two halves. Most dark social does land in Direct. But some of it hides in Organic Social, Referral or a campaign you paid for. And Direct holds plenty that is not dark social at all.

What can GA4 not tell you about dark social?

Who shared the link, and what the share was worth. GA4 identifies a user by their browser or device and your reporting identity setting. The friend who sent the link and the friend who bought are two users. If the sender came from your ad, that ad never meets the purchase in one path.

It also cannot give a shared first visit any credit once another channel follows. Google says its attribution models exclude direct visits from receiving attribution credit, unless the whole path was direct. So if a chat link brings someone in and they later return through a search ad, the ad gets the sale. The chat gets nothing, under every model on the menu.

That leaves one fair way to value talk: change something people talk about in some regions and not others, then compare total sales. That is a holdout test, and no channel label can stand in for one.

What to do this week

1. Find the Direct visits that look shared. In GA4, open Reports > Acquisition > Traffic acquisition, switch the dimension to Session source / medium and search for (direct). Add the secondary dimension Landing page + query string. Pass: Direct mostly lands on your home page, which fits typed addresses and bookmarks. Fail: it lands deep on product pages, which is what a pasted link looks like, so run step 2.
2. Send yourself one shared link. In the chat app your customers use most, send yourself a product link ending in a marker such as?share=test1. Open it on a phone that has never visited your store, so GA4 holds no earlier campaign for it. The next day, open Reports > Engagement > Landing page, search for share= and add the secondary dimension Session source / medium. Pass: you can see whether that app's shares land in Direct or under the app's name. Fail: the visit never shows, so check whether a data filter drops your own traffic, then retest on mobile data.
3. Tag the links you send yourself. Support replies, community chats and broadcast lists are your own dark social. Give each link utm_source with the app's name, utm_medium=social and a utm_campaign, since Google advises one unique utm_source for each distinct platform. Pass: next week those visits sit under Organic Social. Fail: they land in Unassigned, so the medium does not match GA4's social list.

Check the homework. Your GA4 Attribution paths export already holds the evidence. Causality Engine reads that one file and shows what each channel caused next to what last-click gave it, in 1 to 2 minutes, for €99 once (excluding VAT), refundable within 30 days. Check the homework

Sources, 1 October 2026: Default channel group (Google); User acquisition report (Google); GA4 default-channel-group sources and categories (Google); Understand (direct) / (none) traffic (Google); User acquisition report vs. Traffic acquisition report (Google); Scopes of traffic-source dimensions (Google); Get started with attribution (Google); Traffic acquisition report (Google); Landing page report (Google); URL builders: Collect campaign data with custom URLs (Google); Filter out internal traffic (Google)

Related answers

• All Direct traffic questions
• How to find dark social in GA4, step by step
• Does dark social count as direct for a home goods store?
• Direct traffic, defined
• Incrementality testing, explained