What Is Dark Social? How to Measure WhatsApp and DM Shares

What is dark social?

Dark social is the sharing of a link through messaging apps, direct messages (DMs), or email, where the resulting visit reaches your analytics tool without a visible source. Because the visit carries no source information, it usually hides inside "direct" traffic, and the business never notices the share.

The term does not describe anything sinister. Instead, it describes the area that measurement tools cannot see. When a friend sends you a product link on WhatsApp and you tap it, your analytics tool often has no idea where that tap came from.

This guide stays on one topic: noticing this invisible sharing and measuring it roughly. Ads, SEO, and other platform features also appear, but only as context. In each section we tell you what to do step by step, and we are honest about the limits of every method.

Why does analytics count dark social as direct traffic?

When a visitor taps a link, the browser can tell the site which page the visitor came from (the referrer). Messaging apps and some email clients often do not pass this information along. Copying a link and pasting it into a browser also wipes the trail.

The Google Analytics help page describes the Direct channel as traffic from users who arrive through a saved link or by typing your address. The source reads "(direct)" and the medium reads "(none)" or "(not set)". So any visit with an unknown origin lands in that bucket. In other words, dark social hides inside direct traffic.

One caution applies here. App behavior changes with versions and settings, so check your own setup. So it would be wrong to claim that a specific app always or never passes the source.

Which channels does dark social cover?

Dark social is not one app. What the channels share is that people send the link privately or inside small groups, not in a public feed. As a result, the chance of losing source information is high in all of them.

Typical examples include:

• One-to-one and group chats in WhatsApp, Telegram, and similar messaging apps.
• The DM inboxes of platforms such as Instagram, Facebook, and LinkedIn.
• Workplace chat tools and email, both inside and outside a company.
• A link copied and pasted into a document or a note.
• Addresses sent by SMS.

Availability and usage vary by country, app version, and account type. So the healthiest approach is to watch which channel your own audience prefers. When you ask customers directly, the picture gets much sharper.

Email deserves its own note. For example, if you tag a newsletter link, you see the source. But when a reader forwards the newsletter to a colleague, the second visit often arrives untagged. This chain sharing is the quietest form of dark social, and you only spot it by asking questions or watching deep pages.

In-app browsers can also affect measurement. A link opened inside a messaging app's own browser may behave differently with sources and cookies. Because this changes with the app and version, we avoid fixed rules. Still, testing your own site on several devices is the most reliable path.

How does dark social differ from organic social and direct traffic?

People confuse these three, since they sit side by side in a report. However, their origins, measurability, and next actions differ. The table below summarizes the distinction that helps when you decide. It is a general frame based on our field experience, and it does not behave identically in every account.

The second row matters most. Because source information is missing, dark social never shows up as its own channel. You separate it through indirect signs.

In practice, you use the split like this. First, for organic social, you tie traffic to the post that created it. Then, for dark social, you read the link, the content, and the customer's answer together. Finally, for direct traffic you invest in brand awareness, and for dark social you invest in share-worthy content.

Why does dark social matter for your business?

In practice, people often make buying decisions by talking to those they trust. Also, a link from a friend carries a stronger trust signal than an ad. So visitors who arrive this way tend to have real intent.

Here is the problem: if reports hide this effort, you decide wrongly. For example, demand from customers who recommend you shows up as "direct", and you move budget to the wrong channel. In short, you may cut a channel that you simply could not measure.

The goal of measuring is not to find a perfect number. Instead, the goal is to improve decision quality and to grow word of mouth on purpose. Businesses that do this know which content circulates in chats, and they produce more of it.

There is also a cost angle. You do not spend ad budget to win a visitor who arrived through a recommendation. But if you cannot see this demand, you may drop the content that feeds it. For instance, imagine deleting a price guide that circulates in chats. That move could lower demand without ever appearing in a report.

Can you measure dark social exactly?

No, and it helps to accept that from the start. That said, no method can catch every share without peeking into private chats. The goal is a rough picture and a trend over time.

In practice, you combine four tools: tagged share links, channel-specific short links, campaign codes, and a source question that you ask customers directly. However, each tool alone is incomplete. Together, even their disagreements help, because they show which channel the report still misses.

Treat the numbers as a directional signal, not as hard truth. The honest answer to "what percent is dark social?" is "we don't know exactly, but here is how we read the trend." This approach also builds trust when you present to management.

Be patient during the first weeks. You are testing the setup itself. Do the tags work? Are the codes recorded? Does anyone skip the source question? Until you can answer these, drawing conclusions from the numbers is premature. Verify the setup first, then read the trend.

How do you add UTM tags to share buttons?

If your site has a "Share on WhatsApp" or "Send by email" button, tagging the button link is the easiest measurement step. UTM parameters are the tags you add to the end of a link to describe its source. For details, see our UTM parameters guide.

Google Analytics help explains how to add parameters to custom URLs. It also recommends using utm_source, utm_medium, and utm_campaign together. It also stresses lowercase and consistent naming, since upper and lower case count as different values.

Instead of typing tags by hand, you can use our UTM builder. Ask your developer to make the share button read the page address automatically and join it with the tag.

Which rules should you follow when tagging share links?

The shorter and tidier your tags, the easier the report is to read. First, pick one naming scheme for share buttons and share it with the whole team. That way, the same channel never splits your report with different spellings.

• Write the channel as the source and the type of share as the medium.
• Name the campaign after the topic of the content.
• Do not add UTM tags to links inside your own site, because they can break the session source.
• Remember that someone can copy a tagged link into another channel.
• Record all tags in one shared table.

Do not underestimate the last item, because it explains most surprises. If someone copies a tagged link, it keeps traveling in another channel. So you read results as a rough estimate, not as a fact. Even so, a tagged link tells you far more than an untagged one.

Here is a small sample scheme. For instance, use "whatsapp" as the source, "share_button" as the medium, and a short content name as the campaign. For email, use "email" and "share_button". Because the scheme is simple, a new teammate can apply it on day one.

How do you use channel-specific short links?

Creating a separate short link for each channel keeps tags readable. It also makes the address easier to remember when you give it out verbally or in print. A short address on your own domain adds a feeling of trust.

For example, imagine one campaign page. If you open three short links for a dealer, a newsletter, and a flyer, each points to the same page but carries a different tag. The page stays single, and the source information separates.

Also, keep redirect chains short. Each extra redirect adds speed and error risk. Before you publish a short link, test that it opens the right target on both phone and desktop. Also keep a table of your links, so the question of who created which link, for which channel, and when always has an answer.

How do you track conversations that start from a WhatsApp link?

WhatsApp Help describes the click to chat link. It opens a chat with a phone number and can include a prefilled message. You can use this link on your site, in your profile, or in your email.

The trick for measurement is to place a short source code inside the prefilled message. For instance, you write "Hi, I'm writing from your website, code: WEB1". When the message reaches you, you see the source. If you choose a different code for each surface, you can count which surface brings conversations.

Our WhatsApp link generator helps you build this link. However, availability may differ by country and account type. Check the current details on the official help page.

Record the source once the conversation arrives. Then the person who reads the message can write the code into a table or the customer card. At month end, you compare how many conversations, quotes, and sales each code produced. This simple table often paints a more realistic picture than many dashboards.

How do you ask "How did you hear about us?"

Direct questions often catch what technical tags cannot. Add one source question to your form, your checkout, or your first phone call. Keep it short and allow free text.

If you offer only preset choices, people tick the first reasonable option. So include choices such as "Recommendation from a friend" and "Link sent in a message", and leave an open field as well.

Read the answers together once a month. The story in the customer's own words matters more than the count. Sentences like "my partner sent it" or "it came up in our group chat" carry information that no tag will catch. Share these sentences with the team and reflect them in your content plan.

Timing matters too, because it changes the answer rate. Asking right before a sale interrupts the flow, while asking right after a sale raises the answer rate. If you work by phone, add a one-line field to the call notes. However, do not push a customer who prefers not to answer.

How do you set up campaign codes and vanity URLs?

A campaign code makes it easy for customers to report the source themselves. For an offer you share by message, pick a short, memorable code. When the customer says it at checkout, in a form, or in a conversation, you record the source.

A vanity URL also does a similar job. For example, you print a short address on a flyer that belongs only to that campaign. Visits to that address count as direct hits from the flyer, since no other path leads there.

If you use QR codes, our QR code generator can encode a tagged address. In practice, this method works especially well on offline surfaces. If scanning fails, see our QR code not scanning guide.

So do not make the code complicated. Customers can slip when they say it aloud. Choose codes that are easy to read, avoid confusing letters, and give each code a single purpose.

How do you look for dark social in GA4?

In general terms, Google Analytics 4 collects visits with an unknown source in the Direct channel. Visits with a missing or wrongly tagged source appear under a different heading. For that, read our GA4 unassigned traffic guide.

Do not mix the two ideas. In short, direct means the source is absent, while unassigned means the source could not be classified. The fixes differ.

Menu names in reports change often, so we do not give an exact click path. Look for the relevant section in the acquisition reports. First compare the channel groups. Then check which landing pages the direct traffic hits. This second step is the basis of the clue we describe next.

Also, do not compare a single day. Work with weekly averages, because sharing behavior differs between weekdays and weekends. Also mark known events such as campaigns, newsletters, or press mentions on a calendar. That lowers the chance of misreading a spike.

How do you read direct traffic by landing page?

People usually type your home page address into the address bar. Typing the address of a long blog post or a specific product page from memory is unlikely. So when direct traffic lands on deep pages, the chance of a shared link behind it rises.

This is not proof. Instead, it is a clue based on field experience. Still, it is a strong sign. List the landing pages of direct traffic other than the home page and watch which ones suddenly rise.

A sudden rise can mean that the page started circulating in a chat or group. So in that case, it makes sense to speed up and clarify the page. Add a clear next step, because someone who does not know you may have just landed there for the first time.

How do the dark social measurement methods compare?

Every method gains something and misses something. The table below shows what you accept when you pick one. It is a general frame, and results vary by account.

Do not settle for one method. To begin, a source question plus a tagged share button gives the most information for the least effort.

Which pages circulate most in chats?

Pages that circulate in chats usually share a certain character. Pages that explain prices or packages, short guides, contact and location pages, and campaign announcements are shared often. People want to answer a friend with a single link during a decision.

To find these pages on your site, look at the landing pages of direct traffic. The pages at the top of the list are likely at the center of sharing. Because of that, keep them simple: a clear title, fast loading on mobile, and a visible next step.

Also keep changeable details such as dates or prices current on these pages. A link that circulates with outdated information quietly erodes your credibility. A customer can pass old details to someone else without ever complaining to you.

How do you connect brand searches with dark social?

A person who hears your name in a chat may type it into a search engine instead of tapping a link. So dark social also shows up in brand-name searches. So looking at the trend in brand searches alongside direct traffic gives one more clue.

You can follow brand searches in sources such as Google Search Console at a general level. Menu names can change, so look for the relevant section in the panel. If both curves rise together, the chance of word-of-mouth interest grows.

Still, be careful. Brand searches also respond to ads, press, and seasonality. So do not treat them as proof on their own. They gain meaning when you read them together with source question answers and deep page traffic.

How do you build content that people want to share?

Encouraging shares is part of dark social work, just like measuring. People send a link when they think it helps the recipient. So keep the title clear, the page fast, and the first screen sharp.

In messages, a link shows up as a preview card. Check that the title, description, and image look right before you publish. For example, a broken image or a cut-off title lowers the chance of sharing.

Shared content should also be readable for everyone. For that, see our social media accessibility guide. Simple habits such as alt text and captions let more people use your content.

How do you protect privacy while measuring dark social?

The line here is clear: you do not try to read private message content. You only use the clicks on your own link and the information a customer gives willingly. Besides, secret observation is unnecessary, and it damages trust.

If you collect the source question as a form field, say briefly what you use it for. Rules on personal data and consent vary by country. Therefore, this article is not legal advice, so ask a qualified professional about your own situation.

Also check how your cookie and consent settings affect measurement. Visitors who decline may be missing from analytics. That adds to the size of the "direct" and "unknown" buckets.

Also, be careful with customer records. When you write down a code or a source, store only the necessary information. Avoid collecting extra personal detail, and limit access to the team. This care lowers both legal risk and loss of trust.

Which mistakes do businesses make when measuring dark social?

Most mistakes come from haste. First they hunt for a number, then they trust the number and decide. Yet this field is read from a few weak signals together, not from a single table.

• Counting all direct traffic as dark social. Real direct visits exist too.
• Writing tag names inconsistently and splitting the report.
• Adding UTM tags to links inside your own site.
• Skipping the source question and trusting analytics alone.
• Publishing short links without checking their targets.
• Setting everything up once and never looking again.

Every mistake is fixable. The costliest one, however, is to treat an unmeasured channel as worthless and pull budget from it. Make it visible first, then decide.

Finally, there is one more mistake: celebrating the number too early. Data is thin in the first weeks, and small swings look big. Do not draw conclusions before you see a trend of at least a few weeks.

What does dark social measurement look like in practice?

The three situations below are example scenarios, not real client results. Their purpose is to show how the methods fit everyday work.

• A local restaurant (example scenario): It adds a "Share on WhatsApp" button to the menu page and a booking link with a prefilled message. It sees the code in booking messages and counts which surface brings conversations.
• A service company (example scenario): A blog post circulates in group chats. Direct traffic jumps to that post, so the company adds a source question to its form and often reads "it came up in our group" in the answers.
• An online store (example scenario): It tags the share button on product pages. After an order, it tracks the recommendation rate with a "How did you hear about us?" question and sends a small thank-you to the customer who recommended.

All three scenarios stay simple on purpose. So you can test the same logic in your own business with a smaller step. For instance, add a code to one WhatsApp link and count the incoming messages for two weeks.

The common thread is the same: they trust two or three signals pointing in one direction, not a single signal.

How do you set up a 30-day dark social measurement plan?

In the first month, the goal is a working base, not a perfect setup. Start small and add one step each week. The plan should stay at a size your team can manage.

Before you begin, pick a single success metric. For example, choose "number of conversations from recommendations" or "answers saying a link arrived by message". If you try to track everything without a metric, no signal comes out clear.

1. Record the current level of direct traffic and its landing pages in week one.
2. Next, add tags to share buttons and short links in week two.
3. After that, put a source question in the form or at checkout, and add a code to the WhatsApp link in week three.
4. Finally, place all signals side by side and report in week four.

Then, next month, simplify the plan based on which signal worked. If you want to run this alongside your wider social work, our social media management service can support it.

How do you explain dark social findings to management?

Management expects exact numbers, while you instead offer a trustworthy range. Saying so honestly builds confidence. Use a sentence like "Part of direct traffic probably comes from sharing, and these three signals support it."

Put into the report the number of recommendation-based conversations, code usage, and direct traffic landing on deep pages. If you need to compare channels, you can use our attribution model comparison tool.

That way, decisions rest on several consistent signs, not on one flawed number. Also write down what you cannot measure. This openness helps you make better budget calls in later months.

In the presentation, tell a short story instead of showing one chart. First show which pages direct traffic lands on. Then show what customers answer to the source question. Finally show how many coded conversations turned into sales. If each step supports the previous one, management understands the number is not exact and still decides with confidence.

How do neighboring features relate to dark social?

This article focused on one topic, so a brief link to neighbors is enough. Business messaging channels, template messages, and newer networks all affect sharing. Each deserves its own article, so we do not go into detail here.

For one-way communication with a community, see our guide to WhatsApp Channels for business. For template messages and ads, read the WhatsApp marketing messages guide. If you are weighing a newer network, our Bluesky article covers that side.

The same principle applies to all of them: to catch sharing, you first need to know which link goes to which channel.

Also, you can start small and grow the system. Pick one channel and one page first, then expand as you see results. Finally, a regular review meeting with your team keeps the system alive.