The traditional agency retainer is a strange thing. You pay a fixed monthly fee, the agency assigns hours to your account, and at the end of the month everyone counts up the time spent. If the hours were used, both sides feel satisfied. If they weren’t, someone feels guilty. Neither outcome has much to do with whether your marketing actually moved the needle.
At Choco Media, we’ve structured our engagements differently from the start — and the longer we work this way, the more convinced we are that the standard agency retainer pricing model creates the wrong incentives on both sides of the table. This post explains our thinking, what we do instead, and how we decide if a new model is right for a given client.
Why the hours-based retainer fails everyone
The hours model sounds fair on the surface. You buy a block of time, we fill it with work. In practice, it generates a slow, persistent misalignment between what the client needs and what the agency delivers.
- Agencies are incentivised to be slow. Efficiency means unsold hours. There is no structural reward for finishing faster.
- Clients are incentivised to over-brief. If you’ve paid for 40 hours and only used 30, you look for work to fill the gap. The work expands to justify the cost.
- Reporting becomes a justification exercise. End-of-month reports often answer “did we use the time?” rather than “did this work?”
- Speed improvements are penalised. When we automate a task with AI that used to take four hours, we either hide it from the client or explain why the value didn’t shrink when the time did.
We’ve been on both sides of this. We know what it’s like to watch a Monday-morning Slack message calculate how many retainer hours remain. It’s not a great place to do your best work from.
What we moved toward: output-based pricing
The model we use most often now is built around defined deliverables rather than time blocks. A client knows exactly what they’re getting each month — a specific number of ad creatives, a content package of a fixed scope, a defined set of SEO actions — and we price for the output, not the hours behind it.
How we define a deliverable
A deliverable, in our context, has three properties: it is specific (both sides can point to it and agree it exists), it has a quality standard (we can both agree when it’s done properly), and it has a frequency (weekly, monthly, per sprint).
Examples from current client work:
- Eight short-form video scripts per month, each with a hook variant and a caption
- One long-form SEO article per fortnight, to an agreed brief template
- Full monthly paid media account management: weekly optimisations, a creative test per cycle, and a written performance summary
- A brand review pass on all client-produced copy before it goes live
When the deliverable is clearly defined, billing conversations become very short. Either we delivered it or we didn’t.
The incentive structure this creates
Output pricing changes how both sides behave.
For us, efficiency becomes an asset rather than a threat. If we find a better process — AI-assisted drafting, a tighter brief template, a smarter review workflow — the benefit compounds. We get back time we can reinvest in quality or in additional output, rather than watching it disappear into an hour-count.
For the client, there is no perverse incentive to over-brief. You’re not paying for unused hours. If you need something different this month, we talk about scope, not time.
The question shifts from “did we use the budget?” to “did we hit the outputs?” — and that turns out to be a much better question to build a working relationship around.
The model does require more precision upfront. We spend more time in scoping conversations than most agencies. We define exactly what a deliverable looks like before we agree to produce it. That investment in clarity pays back quickly — typically within the first month.
Where the output model still has edges
We’re not arguing this is perfect. There are genuine limitations.
Strategic work is harder to scope
“Develop our Q4 messaging” is not a deliverable in the output sense. Strategic advisory, positioning work, and research-heavy projects don’t fit cleanly into this structure. For those engagements, we tend to use a fixed-project fee rather than a monthly retainer, with a clear scope document and defined review milestones.
Iteration cycles need clear rules
If a client requests three rounds of revisions on a script and we’ve budgeted one, we need a policy. We include revision terms in every engagement — how many rounds are included, what counts as a revision versus a new brief, and what happens outside that scope.
Scope creep still happens
Defining deliverables reduces scope creep significantly. It doesn’t eliminate it. The moment a client asks “could you also just quickly…” in a Slack message, the risk is back. We handle this with monthly scope reviews — a short call to confirm the deliverable set is still the right one, or to adjust it formally. Our client onboarding process includes a scope document that becomes the reference point for these reviews.
How we price the outputs
We’re transparent about this: we price on value delivered, not on our internal cost. The internal cost matters to us — it sets the floor — but the price we quote is based on what the output is worth to the client.
A short-form video script that runs as a paid ad has a different value profile than a blog post sitting in a CMS waiting for organic traffic. We try to price accordingly, and we explain the logic when we quote.
For clients on our bespoke retainer, pricing is agreed quarterly. Rates are fixed for the quarter — the client knows what they’re paying, we know what we’re delivering, and neither side has a reason to game the other. If the scope needs to change materially, we adjust in the next quarterly review rather than renegotiating every month.
What this model selects for in clients
Output-based pricing is not for everyone. It works best with clients who:
- Know what outcomes they’re working toward (not just “more marketing”)
- Can brief clearly, or are willing to go through a structured briefing process with us
- Are comfortable making decisions in defined review rounds rather than asking for open-ended iteration
- Value speed and efficiency as well as quality
In client work we’ve found that clients who struggle with hours-based retainers — who feel they’re not getting enough value, or who feel the agency is slow — often do much better with an output model. The clarity is part of the value.
We’re selective at the intake stage. The questions we ask in early conversations are partly about budget, but mostly about whether a client can work in the way this model requires. If the answer is no, it’s better to say so before the contract is signed than six weeks in.
The question about AI
The AI context matters here and we’d be ignoring the obvious if we didn’t name it.
AI tooling has made parts of our production significantly faster. A first-draft article that once took three hours now takes forty minutes. An ad variant set that required half a day can be produced in an hour. Under an hours model, this efficiency improvement would put direct pressure on our revenue.
Under an output model, it doesn’t. We can improve our internal tooling aggressively, pass some of the benefit to clients in the form of more output for the same price, and keep some as margin. This is what any well-run production operation does when it improves its process — it’s just not what most agencies have been structured to allow.
Our AI automation work with clients follows the same logic: we help them price for value delivered rather than hours spent, because the AI efficiency is otherwise a cost rather than an advantage.
What we tell clients who ask “but what about accountability?”
This is the most common objection we hear when moving a client off an hours model. If we’re not counting hours, how does the client know we’re putting in the work?
The answer is: accountability shifts from time to output, and output is more legible than time. A client can look at a month’s deliverables and judge quality directly. They can compare this month’s results to last month’s. They can see what was produced, test it, and build on it.
What they can’t easily evaluate is whether 40 hours was genuinely 40 hours, or whether any given hour was used well. Time is a proxy for effort, and a rough one. Output is a much cleaner signal.
We also share process notes transparently — what we tested, what we dropped, what we’re planning next cycle. Clients who want visibility into the work get it; they just don’t get a timesheet. In our experience, most clients who ask for timesheets are actually asking for confidence that work is happening at all. That confidence is better addressed with clear output and regular communication than with hour logs.
Practical notes if you’re switching models
If you’re running an agency on an hours model and considering this shift, a few things we’ve learned:
- Don’t flip existing clients mid-contract. Present the new model at renewal time, with a clear explanation of why it works better for them.
- Start with your clearest deliverables. Content retainers and paid media management translate well. Broad “strategic advisory” retainers are harder to start with.
- Write down what “done” looks like. A simple quality standard document — shared with the client — removes most of the ambiguity that makes hourly billing feel safe.
- Price slightly higher than you think the hours justify. You’re selling certainty and accountability, not just time. That has its own value.
- Expect a longer scoping conversation. Budget time for it. It saves multiples later.
Our current retainer structure in brief
For transparency: our standard engagements run on monthly output-based pricing with a quarterly review. Scope is defined in a shared document before the engagement starts. Revision rounds are explicit and included. AI tooling is used heavily in production; the efficiency goes into quality and output volume, not margin-padding. We don’t charge by the hour, and we haven’t for some time.
If you’re evaluating whether a different engagement model might suit your business better, the contact page is the quickest route to a straight conversation about it.