# What is dark social? Canonical: https://postmesh.app/glossary/dark-social By PostMesh team. Published 2026-10-31; updated 2026-10-31. Dark social is content sharing that happens in private channels, like DMs, messaging apps, texts and email, where analytics can't tell where a visit came from. Alexis Madrigal coined the term in The Atlantic in October 2012. Those visits often show up as direct traffic. ## What dark social means Dark social is sharing you can't see. When someone copies a link to your post or product and pastes it into a group chat, a text, a DM or an email, the people who click it arrive at your site, but often without any sign of where they came from. In your analytics they look like they typed your address in by hand. Dark doesn't mean sinister. It means unlit: the sharing is real, possibly a big part of how people pass your content around, but it happens where your measurement tools can't look. Public sharing, like a repost on X or a share to an Instagram story, leaves a trail. Private sharing mostly doesn't. **Public social sharing** - Reposts, shares, quote posts - Visible on the network - Counted in post analytics - Visits usually arrive with a referrer **Dark social sharing** - Links in DMs, group chats, texts, email - Invisible to everyone outside the chat - Missed by social listening tools - Visits often look like direct traffic ## Where the term dark social comes from The term comes from an article by Alexis C. Madrigal in The Atlantic, [Dark Social: We Have the Whole History of the Web Wrong](https://www.theatlantic.com/technology/archive/2012/10/dark-social-we-have-the-whole-history-of-the-web-wrong/263523/), published on October 12, 2012. Madrigal pushed back on the popular story that the web only became social with sites like Facebook and Twitter. His own web in the 1990s, he wrote, was full of instant messenger, chat rooms, forums and email, all used to share links. His argument was that public sharing on social networks is the visible tip of a much larger iceberg, impressive mainly because it's easy to measure, while most sharing happens through private channels like email and IM. He cited new data from many media sites showing that 69% of social referrals came from dark social and 20% from Facebook. - 69%: Share of social referrals from dark social, in data Madrigal cited in 2012 - 20%: Share from Facebook in the same data - Oct 12, 2012: Date Madrigal's article coined the term in The Atlantic *Source: The Atlantic, October 12, 2012. These are 2012 figures from media sites, not a current benchmark.* Those numbers are more than a decade old and came from media sites, so don't quote them as today's share for your business. The idea has aged better than the figures: people still pass links to each other privately, and analytics still struggles to see it, which is why marketers keep using the term. ## Why dark social shows up as direct traffic Madrigal explained the mechanics in the article. When someone follows a link from a site like Facebook, a small piece of data comes along that tells the destination site where the visit came from. Many private channels don't pass that data along, so the visit arrives with no source. Analytics tools put sourceless visits in a catch-all bucket. In Google Analytics 4, the Direct channel is defined as visits whose source is (direct) and whose medium is (not set) or (none), which Google describes as covering saved links and addresses typed into the browser. Dark social lands there too. If your direct traffic is high for pages nobody would type from memory, like a long blog URL, a lot of it is probably links being passed around privately. ## Where dark social happens - Messaging apps: WhatsApp, iMessage, Messenger, Telegram, Signal, WeChat and others. - [DMs](/glossary/dm) inside social networks, when someone sends a post or a link to a friend. - Group chats for friends, families, teams and communities. - Email, including newsletters forwarded to colleagues. - Work chat tools, where links get shared in channels. Some illustrative examples of dark social in everyday life: - "Sending you this recipe, we should make it Friday." - "Saw this tool in a newsletter, forwarding it to the team." - "Look at this jacket" in a family group chat, with a product link. - "lol this is so you" with a link to a Reel, sent in a DM. ## How to measure more of it You can't see inside private chats, and you shouldn't try. But you can make more of the sharing traceable when it lands on your site: 1. **Tag every link you share**: Add UTM parameters to the links in your posts, bio and newsletters. Google's guidance is to always include utm_source, utm_medium and utm_campaign. When someone copies a tagged link into a chat, the tags travel with it. The [UTM link builder](/tools/utm-builder) makes them. 2. **Make sharing buttons tag links too**: If your site has share buttons for messaging apps or email, have each one add its own tags, such as a source of whatsapp or email. 3. **Watch direct traffic to deep pages**: Direct visits to your home page are often typed. Direct visits to a long article or product URL are a good proxy for dark social. 4. **Ask people**: Add an optional question to sign-up or checkout forms: how did you hear about us? A text box catches answers like a friend sent it, which no tool can. 5. **Compare spikes with your posts**: If direct traffic to a page jumps right after you post about it, the post probably started private sharing. Your [link in bio](/glossary/link-in-bio) is a good place to start, since it's the link people most often copy. Give each network's bio its own tagged link so you can tell them apart, and use a separate campaign tag for newsletters, since a forwarded newsletter is a textbook case of dark social. ## Why dark social matters It changes how you judge your posts. A post with modest likes might be sent around privately far more than its public numbers suggest, especially for content people share with specific friends: recipes, gift ideas, job posts, practical guides. Shares and saves, where a network reports them, can be a better signal for this kind of post than likes. It also limits [social listening](/glossary/social-listening). Listening tools only see public posts, so conversations in group chats and DMs are missing from every sentiment chart. Treat public data as a sample of the conversation rather than the whole, and don't assume a quiet feed means nobody's talking. And it's a reminder that [reach and impressions](/glossary/reach-vs-impressions) on the network count only what the network can see. ### What it does to your numbers Dark social skews several metrics at once. Your [click-through rate](/glossary/ctr) on a post only counts clicks the network saw on that post; if someone copies the link into a chat and ten friends click it there, those visits don't count toward the post at all. Your [engagement rate](/glossary/engagement-rate) misses private sends unless the network reports shares. And your website analytics will credit some of your social posts' impact to direct traffic. None of this means the numbers are useless. It means they undercount, and probably undercount some kinds of content more than others. If a post is the kind of thing people would send to one specific friend, like a recipe, a gift idea or a job opening, assume its public numbers miss part of its reach. Keep your measurement consistent so the gaps are at least the same from month to month, and use the steps above to bring some of that sharing back into view. ## How people use the term - "Direct traffic to the guide tripled. Classic dark social." - "Half our sign-ups say a friend sent them the link. That's dark social we can't see in GA." - "Let's add WhatsApp share buttons with UTM tags so dark social shows up." - "Low likes, but people keep DMing it to each other." ## When the term is misused Dark social isn't the same as the dark web, and it isn't secret or shady marketing. It also isn't a channel you can buy. Some people stretch it to mean any untracked traffic, but Madrigal's point was specifically about private, person-to-person sharing. If you're explaining it to someone outside marketing, private sharing or links sent in messages is clearer than the jargon. ## Questions ### Who coined the term dark social? Alexis C. Madrigal, in an article in The Atlantic published on October 12, 2012. ### What are examples of dark social? Copying a link into a WhatsApp group, texting a friend a product page, forwarding a newsletter or sending a post in a DM. ### Why does dark social show up as direct traffic? Private channels often don't pass along where a visit came from, so analytics can't assign a source and file the visit as direct. ### Can you track dark social? Not inside private chats. You can make more of it visible by tagging your links with UTM parameters, tagging share buttons and asking people how they found you. ### Is dark social bad? No. It's often a sign people find your content worth sending to someone specific. It just makes measurement harder. ### Is dark social the same as the dark web? No. The dark web is a part of the internet that needs special software to reach. Dark social is ordinary private sharing. ## Related terms - [Social listening](https://postmesh.app/glossary/social-listening) - [DM](https://postmesh.app/glossary/dm) - [Reach vs impressions](https://postmesh.app/glossary/reach-vs-impressions) - [Link in bio](https://postmesh.app/glossary/link-in-bio) - [CTR](https://postmesh.app/glossary/ctr) ## Sources - [The Atlantic: Dark Social: We Have the Whole History of the Web Wrong, by Alexis C. Madrigal (October 12, 2012)](https://www.theatlantic.com/technology/archive/2012/10/dark-social-we-have-the-whole-history-of-the-web-wrong/263523/) - [Google Analytics Help: Default channel group](https://support.google.com/analytics/answer/9756891) - [Google Analytics Help: URL builders: Collect campaign data with custom URLs](https://support.google.com/analytics/answer/10917952)