Whose face carries the channel

A go-to-market programme I work alongside spent a week making three decisions in a row, and not one of them was about tooling. Which person's profile a campaign runs from. Whether a clean second profile without a network would do. What that profile needs before it is allowed to send anything. Each was argued on credibility, not on capacity, and together they describe something most companies still book under marketing operations: the channel is a person, and people are not interchangeable.

Three decisions, none of them technical

The first: an outreach programme aimed at one decision-making group was taken off the most prominent profile in the company and moved to a colleague whose professional history sits closer to that group. The prominent profile is not weaker. It is aimed elsewhere — years of posts, connections and endorsements have taught both the platform and the readers what that person is about, and the new group is not it.

The second: somebody proposed a neutral second profile, created for the purpose, no history attached. It was rejected in the room within a minute. No followers, no posts, no shared connections — it reads as a prop, and the first thing a recipient does is look.

The third: before the new profile is allowed to send anything, it gets a proper header image and a position line that says what the person actually does. That is not cosmetics. It is the part that turns a private account into a channel, and it has to exist before the first message, not after.

Underneath all three sits the same constraint. Reach in this channel is attached to a person's accumulated record, and that record cannot be transferred, faked in a week, or spent twice.

The second channel behaves the same way

An analysis published this month looked at 89,000 profile URLs cited by AI search products and found the network showing up in roughly 11 percent of answers — around 14 percent in one product, 13 percent in another, 5 percent in a third. Almost all of it came through individual profiles rather than company pages. It is a vendor study, so treat the exact percentages as an order of magnitude rather than a measurement.

The direction is what matters, because it points the same way as the room did. Two different distribution systems — a feed algorithm and a model deciding what to cite — independently ended up weighting people over entities. If you are building visibility, you are not building the company's; you are building someone's, and then borrowing it.

What this costs the organisation

The honest version of this is uncomfortable for anyone who plans channels like media buys.

• Credibility is not fungible. The person with the largest network is not automatically the right sender; the right sender is the one whose record matches the audience, even when their network is a tenth of the size.
• It is finite per person. Every campaign that goes out under someone's name spends a little of what makes that name worth reading, and the spending is invisible until it isn't.
• It cannot be hired quickly. A new profile needs months before it carries anything, which means the decision about whose name appears is a decision about how long you are prepared to wait.
• It leaves when the person leaves. That is the part no channel plan wants written down, and it is true regardless of whether anyone writes it down.

What to do with it

Go through the outbound you currently run and write, next to each stream, whose name it goes out under and which audience that person's record actually speaks to. Where the two do not match, the fix is usually not better copy. It is a different sender, and a conversation with that sender about what they are lending you.

And if the answer for a given audience is "nobody here has that record yet", that is a real finding. It means the work is not a campaign. It is twelve months of someone publishing under their own name — or a decision to reach that audience somewhere else entirely.