The case for embedding instead of pitching

The agency pitch is a strange ritual. A team that does not know the business is asked to make a presentation about how it would run the business, against a brief written by a procurement team that does not run the business either. The decision is made by a committee that is rarely the team the agency will actually work with. The winning deck is rarely the one that gets executed. The losing decks contain ideas that the winning agency quietly absorbs over the next eighteen months.

The ritual persists because it solves a procurement problem. It does not, in our experience, solve the actual problem, which is finding an agency partner that can do the work and stay accountable to it. The agencies that win the long-term work in our category are not the ones who perform best in pitches. They are the ones who have stopped pitching and started embedding.

What embedding actually means

Embedding is not a synonym for being present in the room. It is a specific operating posture, and it has three characteristics.

The first is that the agency is on the same operating cadence as the in-house team. The weekly review, the monthly business review, the quarterly planning session. The agency does not show up to present at these meetings. The agency participates in them, with the same access to dashboards, the same context on the open questions, and the same accountability for the answers.

The second is that the agency has named individuals doing named work, with continuity across quarters. The pitch model rewards the agency that puts the best generalists in the room for the kickoff and then quietly rotates them out to other accounts within a quarter. The embedded model requires that the people doing the work in month three are the same people doing it in month eighteen, and that they have built up the institutional knowledge that makes their work compound.

The third is that the agency is willing to be evaluated on the same metrics as the in-house team, in the same review, in the same room, with the same standard of evidence. Not on agency-side metrics (deck quality, response time, hours billed). On the metrics that matter to the business.

When those three characteristics are present, the agency is embedded. When any one of them is missing, the agency is, at best, a high-touch vendor.

Why embedding wins the long-term work

The case for embedding is straightforward and almost entirely about compounding. The work an embedded agency does in year two is materially better than the work it does in year one, because the agency has internalized the buyer cohorts, the measurement framework, the creative templates that work, the offers that do not, the calendar quirks of the category, and the political topology of the marketing organization. None of that knowledge is in the pitch deck. All of it is required to do the work well.

A pitch-model agency restarts from zero every time it wins a new account. The first ninety days are spent learning the things an embedded agency would have known on day one. The next ninety days are spent earning the trust required to act on that knowledge. By the time the agency is producing work that reflects real understanding, the contract is up for renewal and the procurement team is wondering whether to put the account back out to pitch.

The math on this is harsh. A pitch-model agency that gets one year of real productivity out of an eighteen-month engagement is operating at maybe sixty percent of its peak. An embedded agency that is in year three of an engagement is operating at full capacity, with two years of compounded context behind every decision. The output gap is not subtle, and the buyer eventually notices.

Why agencies pitch anyway

The honest reason most agencies still pitch is that pitching is the only way the procurement process selects them. The procurement process exists for legitimate reasons. It has to compare vendors against a common standard. It has to provide a paper trail. It has to defend the selection to the audit team. The pitch is the artifact that makes the selection defensible.

But the pitch is selecting for pitch craft, not for the work. The agencies that are best at the work are not always the agencies that are best at performing competence in a ninety-minute slot. The agencies that are best at the pitch are sometimes good at the work and sometimes not. The procurement process is unable to tell which is which.

The agencies that have figured out how to grow on the embedded model have, in one form or another, exited the pitch process. They get hired by reference. They get hired by trial engagement. They get hired by walking into a problem the client is currently failing to solve and solving a small piece of it as a proof point. They do not win logos by pitching. They win logos by working.

The practical move from both sides

For agencies that want to operate this way, the practical move is to stop trying to win the pitches that are run as pitches and to start engineering the entry points that come before the pitch. A small paid trial engagement with a clear scope and a clear evaluation criterion is worth ten pitches. An audit engagement that produces a documented opinion the client can act on is worth twenty. The work is harder to manufacture and longer to convert. It is also the work that builds the relationships that compound.

For clients who want to hire this way, the practical move is to stop running pitches as the default and to start running paid trials as the default. The paid trial is faster, cheaper, and produces evidence that is dramatically more useful than a deck. The agencies that show up well in trials are not always the same ones who show up well in pitches, and that difference is the entire point.

What this means for AYMI

We do not pitch in the traditional sense. We have run a total of three formal pitches in five years, all three at the explicit request of the client, all three with the same caveat: we will participate in the process, we will not pretend the process is the right way to evaluate the work, and we would rather start with a paid two-week audit than a sixty-day pitch cycle.

The clients we have stayed with the longest, and produced the best work for, are the ones who hired us off a paid audit or a trial sprint. None of them came from a competitive pitch. That is not an accident. It is the structural consequence of how the two models select for different kinds of agencies.

The work compounds where the relationship compounds. The relationship compounds where embedding replaces pitching. Everything downstream of that follows.