Dark social: the marketing that works but does not show up in reports
25 September 2026
An article shared on WhatsApp can move from one chat to another, reach the right person and generate a new lead. The campaign has worked, but that journey disappears from the reports: this is dark social, the invisible side of digital word of mouth.
What reports do not tell us
Sometimes we open Analytics and notice an increase in direct traffic. No campaign appears to have generated it, social media cannot take the credit and neither can the newsletter.
Where did those visitors come from?
Sometimes the answer emerges when we speak to the people who completed the form: “A colleague sent me the link”, “I received it on WhatsApp” or “Someone mentioned it in a group”.
The journey did take place. The analytics tools simply could not follow it.
Imagine an article published on a company website. A reader finds it useful, copies the link and sends it in a chat. One of the recipients reads it and shares it with someone else. A few days later, that person returns to the website and requests further information.
The report may show a direct visit and a conversion. Everything that happened beforehand — the sharing, recommendation and discussion — remains hidden.
That invisible part has a name: dark social.
What dark social really is
The term sounds more mysterious than the concept actually is. Dark social is not a hidden social network, nor does it refer to anything secretive or unclear.
It is simply the name given to content shared in private spaces which, for that very reason, is difficult to track.
WhatsApp, personal emails, Messenger, corporate chats and private groups are part of our daily lives. They are also the places where we are most likely to send something directly to a specific person.
A public post may receive a distracted like. Sending a link to a colleague, on the other hand, is a more intentional action. It means believing that the content may be useful or worth discussing.
From a marketing perspective, this is a strong signal. From an analytics perspective, however, it is often almost invisible.
Where certain decisions actually happen
In B2B marketing, people rarely see an advert and decide to buy immediately.
They gather information, compare alternatives, read in-depth content and discuss their options with others. When a product or service requires a significant investment, several people may be involved in the decision: marketing, senior management, purchasing, sales or technical teams.
Many of these conversations do not take place beneath a public post. They happen via email, during meetings or in private chats.
The same is true in specialist industries. A professional discovers a new technology and recommends it to a colleague. A salesperson forwards a case study to a potential customer. A marketing manager shares a proposal with senior management and asks: “Could this be relevant to us?”
The content continues to do its job, but away from the platforms being monitored.
This is what makes dark social particularly interesting. It is not simply traffic that is difficult to attribute; it often includes the very conversations that shape a decision.
Direct traffic is not always so direct
When we see “direct” in a report, we tend to imagine someone typing the website address into their browser.
That is partly true. However, the same category may also include visits whose source Analytics has been unable to identify.
If we open a link received in a chat or personal email, for example, the source information may be lost. The visit reaches the website, but the system cannot determine with certainty where it originated and may classify it as direct traffic.
Of course, not all direct traffic is dark social. It would be a mistake to make that assumption automatically.
Some situations, however, deserve a closer look. If an internal page with a long, difficult-to-remember URL receives a large number of direct visits, it is unlikely that everyone typed the address manually. It is more realistic to think that the link was copied, saved or shared somewhere.
The report shows us the arrival, but it has lost part of the journey.
Can we measure dark social?
Not completely. And this may be the most difficult part to accept.
We are used to thinking that, with the right tools, every step can be tracked. In practice, this is not always possible. People move between devices, applications and private conversations. They copy links, forward them and return to a website days later.
We can still gather more clues.
Using UTM parameters correctly in campaigns, newsletters and company-distributed content helps us understand where a link originated. It will not track every subsequent share, but it does provide a useful starting point.
We can also configure sharing buttons more effectively, compare data from different tools, monitor increases in branded searches and investigate unusual traffic peaks on specific content.
Most importantly, we can do something very simple: ask.
A question such as “How did you hear about us?” included in a form or asked during the first sales conversation can reveal journeys that no analytics system could reconstruct.
“Someone I work with told me about you” is not as precise a source as google / cpc, but it tells us something very important: the brand has entered a conversation.
Reading data without being misled
Dark social does not make reports useless. It simply reminds us that a report is not a complete representation of reality.
Data tells us what the tools were able to record. To interpret it properly, we also need feedback from the sales team, information gathered from customers and a practical understanding of the buying journey.
Otherwise, we risk considering a piece of content ineffective simply because it did not generate many attributable clicks. Or we may give all the credit to the last channel used, ignoring the steps that brought someone there.
This is not an invitation to justify every result with something we cannot see. It is an invitation to read the figures more carefully and avoid asking them for answers they cannot provide.
A campaign must be measured, but it must also be considered within the context in which people discover, share and evaluate a message.
That context is not made up of public clicks alone. It also includes forwarded links, recommendations between colleagues and conversations in which the brand is not directly involved.
Dark social reminds us that the journey leading to a decision is often more complex than the platforms suggest. Forwarded links, recommendations and private conversations can contribute to a campaign’s results without appearing clearly in the reports.
Measurement remains essential, but data should be interpreted alongside information gathered from customers, the sales team and other touchpoints. What we cannot attribute precisely is not necessarily without value.
