Why we don't sell retainers

Retainers solve an agency problem, not an operator problem. The agency wants forecastable revenue. The operator wants a metric moved. The two goals are usually in tension — & the tension shows up in scope changes, padding, & the slow erosion of accountability over the course of a multi-month project.

The mechanics of the misalignment

A retainer is “X hours per month for Y dollars.” The agency books the revenue regardless of what gets delivered. The operator gets capacity, not outcome. Both sides have to actively manage to keep the work directed at the metric, & most of the time they don’t.

The drift looks like this. Month one, the brief is fresh & the retainer is full of intent. Month two, the team is on it but a competing priority emerges & a few hours go to “let’s also look at this.” Month three, the original brief is half-done & the retainer hours are 80% spent. By month six, the operator has spent six retainer-fees & got about three months of focused work.

We’ve all worked with clients who got there. The problem isn’t that anyone was acting in bad faith — the structure rewards it. Hours billed is the unit; when hours are the unit, work expands to fill them.

What we sell instead

We sell projects. One job, one metric, one go-live, one fixed fee. You pay for the move, not the calendar. If we move the metric in week six instead of week ten, we don’t keep billing — we close, hand off, & either start the next project or step out.

The unit is the go-live

Each project names the go-live up front. “Lift cart-to-purchase from 32% to 38%, sustained across three weekly cohorts, by Tuesday May 27th.” That’s the unit. Everything in the project is in service of that go-live. Anything that isn’t gets scoped out.

Scope changes get their own project

Without hours to fill, there’s no “we have hours left this month, want us to look at X?” pressure. If X is a real job, it’s a new project. If it isn’t, it doesn’t get done. Both sides know where they stand. The operator gets a cleaner read on which work is actually moving the business; we get a cleaner pipeline of focused projects instead of slow-creeping retainers.

Honest scoping up front

We can’t pad a project with 30% buffer hours because there are no hours, just a deliverable. If the scope is wrong, we lose. So we write the scope carefully & we set the starting numbers before we start. This is the discipline behind the Map — it’s the only way fixed-fee projects work.

The exception: Scale

We have one plan that looks retainer-shaped. The Scale plan is a continuous embed for operators running multi-stack environments where work needs to be held between go-lives. Build, reliability & security work in particular benefits from continuity.

Even there, the unit isn’t the hour. The Scale plan is priced per quarter against a defined set of work held, with response SLAs & go-live capacity included. The operator isn’t buying our calendar — they’re buying continuous coverage of named work.

The Scale plan is a small fraction of our book. Most operators want Starter or Growth, both of which are project-priced & have a clear close.

What operators sometimes ask for

A few common asks & how we respond.

”Can we do a 3-month retainer to start?”

We don’t. Instead, we offer a Map (fixed fee, one-page document) followed by a first project scoped against the part the Map identified. This gives the operator a sense of how we work, without committing to time we can’t hold accountable.

”We have ongoing work, can we just put you on retainer?”

The “ongoing work” is usually two or three jobs. We propose a Growth plan (multi-job quarterly) with the jobs named. Same total spend over the same period, but each job has a metric & a go-live date. The operator gets continuity; we stay accountable.

”What if we run out of work mid-project?”

This is rare with proper Map work, but it happens. The project closes early & we don’t keep billing. The operator gets the closeout document & either re-engages on a new job or steps out for a quarter. Both are fine.

The trade-off

Project pricing means our revenue is more bumpy than retainers would produce. We accept that. The trade is a clean alignment: every dollar we earn is tied to a metric we moved.

For operators it means harder upfront scoping (the Map) but cleaner accountability (the go-live). It also means they’re not paying for our slow months. We don’t get to bill through a quiet quarter; we close projects, take the gap, reopen capacity when the next ones close.

The takeaway

Retainers feel safer because they smooth revenue. Projects feel sharper because they require honest scoping. We chose the second because the first is mostly a pricing fiction — the actual work is always project-shaped, even when it’s billed as a retainer. We just stopped pretending otherwise.