Skip to content
upshift
Marketing Agencies

Clients are asking to buy your AI workflow instead of hiring you. Nobody has figured out how to price that.

What’s happening

The question is starting to arrive in client conversations, and it catches owners flat-footed. An agency owner put it to peers this summer: clients are asking whether they can buy access to the agency’s AI workflows instead of retaining the agency to run them, and he had no idea how to price selling an agent. The retainer assumed the client kept paying the agency to do the work. This asks the agency to sell the machine that does the work and walk away.

The ask is disorienting because it reprices the agency’s own assets. The same owner described the shift plainly: his people are still the number-one asset, but the number-two asset is now the library of AI workflows the team built, thousands of hours of prompts and process he never expected to be worth much on their own. Agencies have always priced time and retainers. Almost none have priced the codified capability sitting on their own drives, and now a client is asking to buy exactly that.

Two instincts show up, and both leak value. One is to hand the workflow over as a one-time deliverable, priced like the small thing it looks like instead of the capability it is. The other is to refuse and insist the client is really buying judgment, not a workflow. The judgment point is correct. One owner in the thread put it well: you can buy the x-ray, but you still cannot read it. It is also not a pricing model. It gives the client no way to buy what they asked for, so they go build a worse version themselves or find someone who will sell it to them.

The competitive clock is the reason this cannot wait. If an experienced operator can distill years of expertise into an agent, so can a competitor, and so can the freelancer who used to lose the pitch on capacity. The capability is becoming a product whether or not the agency sells it. The only open question is whether the agency names and prices its own version first, or watches a cheaper, thinner copy get sold to the client who was ready to buy it from them.

Where this sits: The Three Shifts

Why the obvious responses don’t work

Hand over the workflow as a one-time deliverable

This prices a capability like a favor. The workflow took thousands of hours to build and can replace the retainer it came from, so selling it at deliverable rates means charging the least for the thing that costs the client the most to lose. Price for the impact it produces, not the file it ships as.

Refuse and insist they’re buying your judgment

Being right is not a business model. The client asked to buy something and you said no, which does not remove the demand, it routes it elsewhere. They build a rougher version in-house or hire the competitor who will sell it. Judgment is the moat, but the client still needs a way to pay you for the capability.

Add a token markup to the existing retainer

This prices the tool’s fuel instead of the outcome. A per-usage markup invites the same scrutiny every AI surcharge invites and caps the agency at cost-plus on an asset whose value is the result it drives, not the compute it consumes.

What’s working instead

The agencies getting this right are turning the workflow into a named, priced offering of its own: defined access to the capability, paired with a retainer to operate and improve it, priced on the business result it produces rather than the tokens it burns. That structure captures the value of the asset the agency already built, keeps the judgment that makes the workflow work as the ongoing relationship, and answers the client’s question with a product instead of a favor or a flat no. It is the productization move the whole market is circling: sell what the work produces, and stop pricing the work by how long it takes.

The pattern is the same across every firm that gets this right: they stop optimizing the old model and build new offerings around what AI cannot do. The Workshop is the facilitated day we do this work with you. You leave with 2–3 offering briefs, specified and priced. Your team tests them with named clients, then builds what earns it.