The credibility gap: what clients check between shortlisting your agency and meeting you

There is a window in every pitch process that agencies never see and rarely plan for. It opens when a client team agrees a shortlist, and it closes when you walk into the chemistry meeting. Depending on the process, it lasts somewhere between four days and three weeks.

In that window, three or four people who have never met you go and find out who you are. They do it privately, on their own time, using whatever surfaces are available. Nothing they find gets written down. All of it shapes the room you eventually walk into.

Agencies plan for the meeting. Almost nobody plans for the window.

Who is actually looking

The pitch document goes to the marketing lead. The window belongs to a wider and more sceptical group.

The person who has to defend the choice. Usually the marketing director or head of brand who put you on the list. Their exposure is real: if the appointment goes badly in nine months, they own it. They are not looking for reasons to be impressed. They are looking for reasons to feel safe recommending you internally.

The person who was overruled. Most shortlists contain at least one agency somebody else wanted. That person will do the most thorough research of anyone in the process, and they will be looking for the thing that disqualifies you.

Procurement. Checking that you are a real business with real clients and no live reputational problem. Fast, shallow, mostly box-ticking - unless something surfaces.

The intermediary. If a consultant is running the process, they already have a view of you, formed over years, from trade press and industry gossip rather than from your website. That view was set long before this brief existed.

And, increasingly, an AI engine. Someone on the team asks ChatGPT or Perplexity who the best agencies are for the category and sees whether your name comes back. This is now routine rather than novel, and it is the only check where the absence of an answer is itself an answer.

None of these people are your buyer in the sense the pitch document assumes. They are auditors. They are trying to establish whether the story you are about to tell them in a room is corroborated anywhere outside your own control.

Everything on your website is testimony. Everything else is evidence.

The four surfaces that get checked

The specific behaviour is remarkably consistent, and it is worth understanding as four separate surfaces rather than one blur of "online presence". Each one answers a different question, and an agency can be strong on one and invisible on the rest.

1. The founder's name in a search bar. Not the agency name - the human's. The question being answered is whether there is a person with a track record behind the pitch, or a brand with a logo. What comes back on the first page decides whether the rest of the check is friendly or forensic.
2. The trade record. Campaign, PRWeek, The Drum, Creative Boom. The question is whether the industry treats you as a source or as a subject. Being quoted explaining something is worth several times being announced doing something, because one implies the industry rates your judgement and the other only proves you have a press release budget.
3. The machine answer. What a large language model says when asked to recommend agencies in your category. The question is whether you exist as an entity the model has seen described consistently across enough independent sources to repeat with confidence. We have written separately on how agencies get cited in ChatGPT; the short version is that models cite consensus, and consensus is built in other people's publications rather than your own.
4. The back-channel. A text message to someone who has worked with you. This is the only check you cannot influence in the window, and it is the one that carries the most weight. It is also the reason the previous three matter more than they look: what a peer says about you is heavily shaped by what they have recently seen you say in public.

Awards sit slightly apart from these. They function as shorthand rather than evidence - a fast way for someone to justify a choice to a boss who has not done the research. Useful, but they answer a different question, and they answer it less well than a byline the reader actually remembers.

What "nothing" reads as

The failure state is not a bad result. It is a thin one.

A founder with a LinkedIn profile last updated in 2023 and a headline that describes their job rather than their view. A news page with four posts, three of them about office moves. No trade coverage in eighteen months. An AI engine that names four competitors and not you, or worse, describes you accurately but generically, in language lifted straight off your own homepage.

None of that is disqualifying on its own. Together, it produces a specific and damaging inference: this agency is fine, but it is not distinctive, and choosing it will require me to do the work of explaining why.

That inference is expensive, because the person who forms it is the same person who has to sell you internally. You have just made their job harder before the meeting has started.

The tell-tales of a credibility gap

Agencies rarely learn they have this problem, because the feedback is structurally unavailable. Nobody tells you that you looked thin. What you get instead are symptoms that most agencies misread as bad luck:

You reach the final three often and win less than a third of them. You get told the process was "incredibly close". You are asked more foundational questions in chemistry meetings than you expect - who your other clients are, how long people have been with you, who would actually work on the account - which is what happens when the room arrives with less prior context than you assumed. Your best pitches are the ones where somebody in the room already knew you personally, and you have never examined why the win rate diverges so sharply between those and the rest.

Read together, those are not variance. They are the shape of a shortlist that arrives without a prior view of you.

Why you cannot close it inside the window

This is the part that makes it structural rather than fixable.

Every one of the four surfaces has a lead time longer than the window itself. Trade coverage takes weeks from pitch to publication in the best case, and the peg has to exist independently of your commercial need for it. Model answers update slowly and are built from sources published across months. A peer's opinion of you is the residue of years. Even a founder's own feed needs a quarter of consistent publishing before it reads as a point of view rather than a burst of activity before a pitch.

Which means the work that decides the window has to be finished before you know the brief exists. Agencies routinely discover this in the wrong order - the flurry of activity that follows a lost pitch is the most reliable sign that the loss was decided in the window rather than the room.

The unglamorous conclusion is that credibility is a calendar problem. The agencies that win the window are the ones running the work continuously at a low level for reasons unconnected to any specific brief, so that it is simply true by the time anybody checks. We set out how those surfaces compound in the agency growth stack, and the trade-press half of it in the trade-press placement playbook.

What good looks like at the moment of the check

The bar is lower than agencies fear and more specific than they expect. Someone spending twenty minutes on you should be able to finish with three things.

A sentence about what you believe that they could repeat to a colleague without checking their notes. That comes from positioning, and it is the load-bearing element - we wrote about writing the positioning statement that holds up when a client asks the hard question.

A piece of third-party evidence they did not have to take your word for. One byline in a title they recognise does more here than a case-study page, because you did not control whether it ran.

A named human they can picture in the room. Not a leadership grid. One person whose view of the category is legible within a couple of minutes.

Three things, and most agencies have none of them assembled in a form a stranger can find in the order they will look.

What this changes about the year

If the window decides more than the meeting, then the new-business budget is allocated wrongly at most agencies. Pitch production is expensive, visible, and applies to one opportunity. Visibility is cheaper, invisible, and applies to every opportunity that has not arrived yet - including the ones where you never learn you were considered and quietly dropped before the shortlist call.

That second category is larger than agencies believe and it is entirely invisible in the pipeline, which is why it never gets funded.

The reframe worth taking away is that this is not marketing your agency. It is making sure that the true version of your agency is discoverable by people who will never ask you directly, at a moment you will not be told about.

If you want to know what a client team currently finds when they check, that is a short and fairly uncomfortable exercise, and we are happy to run it with you. Drop us a line, or read how we work on the method page.