Agency pricing: from inputs to outputs to outcomes

There is a conversation happening quietly inside a lot of agencies.

It doesn't show up in the reports or in team meetings, but it's there: the sense that the model you've been billing with for twenty years no longer describes what the agency actually does today.

For decades, much of the professional services industry built its economics around a simple premise. Selling time. Hours, teams, projects, monthly fees.

Artificial intelligence is starting to break that link between time invested and value created.

If an agency can produce in three hours what used to take twenty, billing by the hour creates an uncomfortable paradox: the more efficient it becomes, the less it should invoice.

And that opens a much deeper question than which AI tools to adopt. How should an agency charge when its productivity is no longer tied to the number of people required to produce a result?

The answer emerging here isn't to abandon the retainer.

The monthly fee still holds real advantages for both sides. For the agency it means predictable revenue, the ability to plan resources, and a long-term relationship. For the client it means budget predictability and continuous access to specialized capabilities.

Which is why it's unlikely to disappear.

What's changing is what that fee represents.

Historically, many agency contracts were built, implicitly, around capacity. A certain number of people. A certain number of hours. A certain number of deliverables.

AI is starting to decouple those variables.

An agency can significantly expand its capacity for research, analysis, production, personalization, or testing without expanding its headcount in the same proportion.

When that happens, measuring the value of a service purely by the effort required to produce it starts to lose meaning.

The conversation moves from "how many hours did you work?" to "what value did you create?"

A useful way to organize this evolution is to think about agency pricing in three layers.

Input-based pricing. The client pays for the resources used: hours, people, seniority, technology, allocated capacity. It's the traditional model across much of professional services.

Output-based pricing. The client pays for what they receive. A campaign. A study. A set number of content pieces. A technology implementation. A production volume. Internal efficiency becomes the agency's responsibility, and that's where it starts to pay off.

Outcome-based pricing. The client pays, fully or partially, for the impact achieved. Pipeline generated. Qualified leads. CAC reduction. Higher conversion. Incremental revenue. Cost savings. Any business indicator both sides can measure with confidence.

At that third level, one of the most interesting shifts in the agency model appears. When part of the compensation depends on the result, agency and client start sharing the same economic equation.

Jumping straight from a retainer to charging purely on results sounds appealing, and it carries a fundamental problem.

An agency rarely controls the entire result.

It can generate excellent leads and run into a sales team that can't convert them. It can drive a substantial lift in traffic while working for a company whose product has conversion problems. It can improve CAC while prices, the market, or the competition shift underneath it.

Pure performance pricing hands the agency risks it doesn't control.

A more balanced alternative is to build the model in layers.

Layer 1, base fee. A recurring fee covering the strategic and operational capacity needed to run the account. It shouldn't be a low number dropped in to win the client. It has to make the relationship sustainable.

Layer 2, performance component. A variable portion tied to indicators agreed in advance: CAC reduction, qualified pipeline growth, conversion rate improvement, attributable incremental revenue, growth over an agreed baseline. When the client gains more, the agency shares in that upside.

Layer 3, specialized projects. This is the growth source most agencies leave on the table. Ongoing work generates information. The agency uncovers problems, opportunities, and bottlenecks that weren't in the original scope. Instead of absorbing them for free inside the retainer, the classic scope creep, they can become projects of their own. Research. AI implementations. Automation. CRO. Data infrastructure. New campaigns. Expansion into other markets or channels.

The client isn't buying more hours. They're buying an answer to a problem that surfaced.

There's an economic consequence of AI that gets discussed far less than productivity.

AI can improve an agency's margin, but only if its pricing model lets it keep that efficiency.

Picture two agencies delivering exactly the same result. The first needs a hundred hours. The second, thanks to automation, data, and better processes, needs forty.

If both bill by the hour, the second one invoices less precisely for being better.

That incentive is absurd.

Under a model based on outputs, value, or results, that same efficiency becomes a competitive advantage. The agency delivers faster, operates on better margins, and can reinvest part of that profitability into technology, talent, and knowledge.

AI stops being a tool for making the same work cheaper and becomes part of the agency's economic model.

For years, many agencies grew in a linear way. More clients required more people, more people generated more hours, more hours generated more revenue.

AI challenges that chain. An agency can produce considerably more output without growing its structure in the same proportion, and at that point growth stops depending on headcount.

That shift forces a review of one of the most important questions in the business.

Are we charging for the effort we put in, or for the value we're able to create?

Agencies that keep selling capacity will probably use AI to make the same work cheaper. The ones that redesign their model around expertise, technology, intellectual property, and results will be able to use it for something far more interesting: creating more value per person in the organization, and keeping a share of that value.

It may turn out to be one of the deepest changes AI brings to the agency business. Not only changing how they work. Changing what they sell and how they charge for it.

Back to Blog