When AI started compressing our delivery times, we ran into an uncomfortable problem. A brand voice guide that used to take three days now took one. A landing page that was a week of work landed in two days. The output quality was the same — sometimes better — but the hours on the invoice had shrunk. Clients noticed. Some appreciated it. A few started asking why the price hadn’t dropped too.
How we price AI-assisted projects is something Choco Media had to work out in public, in front of real clients, with no industry consensus to lean on. This post is the honest version of that reckoning — the model we moved to, the client conversations it required, and the language that makes it land without defensiveness.
If you run an agency, freelance, or sell any kind of creative or strategic service where AI has changed your workflow, this is for you. The core argument: pricing on time was always a proxy for value. AI just made that proxy visible.
Why time-based pricing breaks with AI in the workflow
The traditional agency model prices deliverables by estimating hours and applying a rate. It is a reasonable heuristic when human effort is the main constraint. The problem is that hours and value were never the same thing — hours were just easier to count.
AI changes the hours side of the equation without changing the value side. A well-structured brand voice guide is worth the same to a client whether it took eight hours or eighteen. What the client is buying is the output, the expertise behind it, and the accountability for getting it right. None of those things cost less just because a language model drafted the first pass.
- Speed is a feature, not a discount trigger. Faster delivery has value — it means a campaign launches sooner, a product page starts converting earlier, a brand system gets adopted before the team invents its own workarounds.
- AI does not remove expertise — it amplifies it. Our strategic input, creative judgment, and quality control still account for most of the outcome. AI is a production tool, not a replacement for the thinking.
- Hours-based pricing creates perverse incentives. The more efficient you get, the more you earn per hour in real terms — but the client sees a bill that looks like you worked less. Outcome pricing removes that tension entirely.
Moving away from hourly billing was not primarily an AI decision. It was overdue. AI just forced our hand.
The model we moved to: outcome-based retainers and fixed-scope project pricing
We restructured around two formats. Retainer clients pay a monthly fee for a defined scope — a set of deliverables or a level of ongoing involvement — not a bank of hours. Project clients get a fixed price tied to a specific output: a landing page, a brand system, a campaign brief, a full content quarter.
How we set retainer fees
We start from the value the client gets, not from our time cost. A retainer that manages paid media for a client spending €15,000 per month should be priced in relation to what good management is worth to them — not what it costs us to deliver. If we can deliver it efficiently with AI in the workflow, that is margin, not an obligation to discount.
- We anchor retainers to outcomes: traffic growth, lead volume, return on ad spend improvement, content velocity.
- We scope clearly: this retainer includes X posts, Y ad creative reviews, Z strategy sessions per month.
- We review annually, not when a client notices we have gotten faster.
How we set project prices
Fixed-scope projects are priced by deliverable, with a simple internal check: what is this worth to the client if it performs well? A homepage rewrite that lifts conversion by half a percentage point on a site doing €200,000 per year in revenue is worth significantly more than the hours to write it.
- We publish price ranges for common project types so clients self-select appropriately before we get on a call.
- We define what is included and what triggers a scope extension before we start.
- We do not discount retrospectively because a draft came together quickly.
The question we ask ourselves when pricing a project: if this performs exactly as intended, what is that worth to the client over twelve months? The project price should be a fraction of that — not a function of our calendar.
The client conversation about AI efficiency
Some clients will ask directly: if AI writes a first draft, why are we paying the same? It is a fair question and it deserves a direct answer, not a deflection.
The answer we give goes roughly like this: AI writes a draft the way a capable junior team member might — quickly, competently, but without the strategic context, brand knowledge, and judgment that determines whether it actually works. What you are paying for is that judgment, applied to your specific situation, with us accountable for the result. The draft is not the work. The thinking behind it is.
What helps this conversation land
- Show the editing layer. When we deliver AI-assisted work, we annotate key decisions — why we changed the hook, why we cut a section, why the CTA was repositioned. This makes the thinking visible and distinguishes the output from a raw model response.
- Reference outcomes, not process. “This is priced at X because it is expected to do Y” is a stronger frame than “this took Z hours.” Clients who buy outcomes rarely fixate on delivery speed.
- Be honest about the tools. We do not hide that AI is in the workflow. Clients respect transparency, and it avoids the awkward position of defending a price that looks inconsistent with a process you have obscured.
Our AI content creation service page explains this framing openly — clients know what they are getting before we have the pricing conversation, which means the conversation almost never becomes adversarial.
The internal calculation: margin, not just price
Outcome-based pricing changes how you think about margin. When you price on value, faster delivery means better margin — and that is the right incentive structure. Investing in tools, systems, and processes that improve efficiency becomes a direct profit driver, not just a quality-of-life improvement for the team.
The internal calculation we run for any project or retainer:
- What is this worth to the client? Anchor to value, not cost.
- What does it take to deliver it well? Think in terms of expertise and quality gates, not hours.
- What is a fair price that we can defend honestly? Not the maximum the market will bear, but a price that reflects genuine value without being inflated.
- What is our margin at that price? If the margin is thin, the problem is either scope (too broad) or pricing (too low) — not the tools we use to deliver.
In client work we have found that this framework leads to better client relationships over time. Clients who understand what they are buying and why it costs what it costs are less likely to negotiate on price and more likely to extend and renew.
Where we still use time as a reference point
We have not abandoned time entirely — it is still useful as a sanity check and a scope boundary, just not as the primary pricing unit.
- Scope estimation: We estimate effort internally to make sure a fixed price is sustainable. If a project type routinely takes more than we estimated, we adjust the price on the next proposal — not the delivery.
- Scope creep conversations: When a client asks for significant additions mid-project, we reference effort to explain why there is a scope extension. “That is roughly the same size as the original project” is concrete and easy to understand.
- Retainer calibration: When renewing, we review what was actually delivered against what was scoped. If the scope has grown organically, the retainer should reflect it — and a light effort estimate helps that conversation feel grounded rather than arbitrary.
Time is a useful input, not the output. The distinction matters and is worth maintaining explicitly in internal conversations.
How AI changes what counts as expertise
One thing we did not anticipate: AI in the workflow raises the visible ceiling of what expertise looks like. When a strategist can produce a detailed brand document in a day, clients start expecting that depth as standard. The bar moves upward.
This is mostly a good thing. It forces everyone working with AI to invest in the parts of the work that cannot be accelerated — strategic judgment, client-specific knowledge, creative taste, quality standards. Those are the things that differentiate good agencies from mediocre ones anyway. AI just makes the differentiation more visible more quickly.
- Agencies that use AI to produce generic work faster will find margins compressing as clients notice the pattern and start shopping on price.
- Agencies that use AI to go deeper — more research, better iteration, tighter editing, faster turnaround on revisions — will find clients willing to pay a premium for outputs that actually perform.
- The pricing conversation follows naturally from which category you are in.
Our AI automation services are built around the second approach — using AI to deliver more depth, not just more volume. That choice shapes everything about how we price and position the work.
What we would tell a freelancer or small agency starting this transition
If you are still pricing on hours and AI has started compressing your delivery times, the transition to outcome-based pricing is uncomfortable at first. A few things that made it easier for us:
- Start with new clients. It is easier to introduce new pricing to someone who has not anchored to your old model. Do not lead with “we have changed how we price” — just present the new model as how you work.
- Build a reference set of outcomes. “Clients who bought this typically saw X” is a more persuasive anchor than a deliverable description. It ties price to result rather than to effort.
- Price for repeatability. If you deliver a type of project frequently, price it so you can invest in doing it better — better templates, better processes, better quality checks. Thin margins on repeatable work compound badly over a year.
- Document your thinking, not just your output. A strategy memo, an annotation layer, a decision log — these make your expertise visible and justify pricing that does not track to hours.
- Expect the conversation and prepare for it. Some clients will push back. Having a clear, honest answer ready — not defensive, just true — resolves most of these conversations without renegotiation.
The transition takes a few months to feel normal. Once it does, it is difficult to imagine going back.
The pricing conversation is a positioning conversation
How you price signals what you think your work is worth. Discounting because AI shortened delivery time sends the message that you were padding hours before — which is usually not true, but that is the story the discount tells.
Holding prices and articulating value clearly sends a different message: we use the best available tools to deliver good work efficiently, and the price reflects the outcome, not the clock. That is a positioning statement, not just a billing policy. It shapes what kind of clients you attract and what kind of work you end up doing.
- Pricing on outcomes attracts clients who think in terms of return.
- Pricing on hours attracts clients who think in terms of cost reduction.
- The two groups have very different conversations when something does not go exactly as planned.
In client work we have found that agencies who make this transition clearly and confidently attract better-fit clients over time. The pricing conversation becomes easier, not harder, because the client knows from the start what they are paying for and why.
If you are working through this transition and want to think through how it applies to your service mix, reach out — we are happy to talk through what has worked for us and what we would do differently if we were starting it today.