Access Is Not an Outcome: What Your Agency Retainer Should Actually Buy

There's a particular kind of agency retainer that has stopped being a commercial engine and become a subscription. You pay a fixed monthly fee. In return you get a named account manager, a monthly call, a slide deck with a lot of activity in it, and the warm feeling of knowing somebody is there if you need them.

Nothing is broken, exactly. That's what makes it so hard to challenge. But ask yourself a plain question: if you cancelled tomorrow, what would visibly stop happening?

If the honest answer is 'some meetings', you're not buying outcomes. You're paying for proximity.

How retainers drift

Retainers rarely start badly. They start with a defined problem — a site that needs work, a launch to support, a martech stack to untangle — and a sensible block of days. Then the problem gets solved, and instead of re-scoping, everyone quietly carries on. The agency keeps the revenue. You keep the relationship. The scope becomes whatever lands in the inbox that month.

This is nobody's fault in particular, which is precisely why it persists. Changing it requires a conversation that feels rude, so it doesn't happen, and another year goes by.

The cost isn't just the fee. It's the opportunity cost of work you didn't commission because the budget was already committed, and the slow erosion of internal capability while you outsource the thinking as well as the doing.

'Access' is a symptom, not a service

When a retainer is described mainly in terms of availability — dedicated account management, priority support, a Slack channel, monthly strategy sessions — you're being sold insurance rather than output. Insurance has its place. But insurance is priced on risk, and most retainers are priced on days, which means the two things drift apart the moment nothing urgent happens.

There's a version of this that's genuinely valuable: a retainer that buys you capacity you can't hire, at a seniority you can't afford full-time, with continuity that saves onboarding time. That's real. But it should still be pointed at something.

Scope in outcomes, not activities

A useful retainer scope answers three questions. What measurable thing should be different in ninety days? Who is accountable for it? What does the agency own end to end?

Compare these two descriptions of the same twelve days a quarter.

Version one: 'Monthly performance review, quarterly strategy workshop, ongoing design support, ad hoc development requests, monthly reporting.'

Version two: 'Reduce checkout abandonment on mobile from 71% to under 66% by 31 March. Deliver: two rounds of usability testing on the mobile basket, a rebuilt checkout flow, and a measurement plan signed off before build. Reported monthly against the abandonment figure.'

The first is a menu. The second is a job. Only one of them can fail, and that's the point — a scope that can't fail is a scope that can't deliver either.

You don't need every retainer to be a hard commercial number. Some work is genuinely foundational: technical debt, accessibility fixes, a design system. Fine. Then define the outcome in those terms and put a date on it. 'Core Web Vitals passing on all templates by end of Q1' is an outcome. 'Ongoing development support' is a line item.

The quarterly value test

Every ninety days, ask three questions and write the answers down.

What changed in pipeline? Not traffic, not impressions — pipeline. Which opportunities touched work the agency did, and did any of them move stage?

What changed in conversion? Pick one number that matters to the business and track it across the quarter. Form completion rate. Checkout completion. Demo requests per thousand sessions. If nothing moved, why not, and was that expected?

What changed in search visibility? Not rank for one vanity keyword — visibility across the commercial terms you actually care about, plus whether new pages are earning links and impressions.

If two of the three are flat for two consecutive quarters, you don't have a performance problem you can fix with a better meeting. You have a scope problem.

Add a fourth question, quietly, for your own benefit: how many hours of our team's time did this consume? A retainer with decent outputs that eats a day a week of internal time is more expensive than the invoice suggests.

Read the exit terms before you need them

Most retainer pain lives in the small print nobody reads at signature. Three things are worth checking on every agreement.

Notice period. Ninety days is common and occasionally defensible. Six months, rolling, is a trap. Negotiate down to thirty or sixty at renewal — most agencies will move if you ask at the right moment.

Handover. You want an explicit clause covering asset transfer, documentation, source files, analytics and ad account access, and a defined handover period at no extra charge. 'Reasonable cooperation' is not a clause, it's a hope.

Ownership. Confirm in writing that you own the work product, the design files, the code you've paid for, and the domain and hosting accounts. Any agency that pushes back on this is telling you something useful.

The conversation, without the drama

You don't need to fire anyone to fix this. Open with the shared interest: 'We want to make sure the next twelve months are pointed at things that move numbers. Can we rebuild the scope around three outcomes and review quarterly?'

Most good agencies will welcome it. They're as tired of vague retainers as you are, because vague work is impossible to be proud of. The ones who resist — who want to talk about days, access and 'ways of working' rather than results — have answered your question for you.

A retainer should be a bet you'd happily place again every quarter. If you wouldn't, it's not a partnership. It's a direct debit.