The curious economics of the agency retainer

Imagine two marketing agencies presented with the same problem.

A client wants to identify potential customers for a specialist service, research their requirements, develop relevant communications and generate qualified sales opportunities.

The first agency proposes a team of strategists, researchers, copywriters and campaign managers. It estimates the work will require a substantial number of hours each month.

The second agency uses experienced professionals supported by proprietary AI intelligence. Research that previously required days can be completed far more quickly. Campaigns can be prepared, monitored and adjusted with considerably less manual intervention.

Which agency should charge more?

The conventional answer is that the first agency should. It has more people doing more work.

But this produces a peculiar conclusion: the less efficiently an agency operates, the greater the apparent justification for its fee.

Something is wrong with the economics.

How we came to sell time

The traditional agency model developed in an era when human labour was the principal constraint on producing marketing.

More campaigns meant more planning, writing, design and administration. More clients required more people.

Charging for time, whether explicitly through hourly rates or implicitly through retainers, was a practical way to price scarce resources.

The difficulty is that the customer was never particularly interested in purchasing those resources.

The customer wanted something else entirely.

More enquiries. Stronger market positioning. More sales conversations. More customers.

Hours were simply the mechanism by which the agency delivered the work.

Over time, the mechanism became the product.

Agencies began selling teams, allocated resource, monthly deliverables and contracted levels of activity. The commercial conversation shifted from what the client needed to achieve towards how much work the agency would provide.

This was understandable when production capacity and commercial value were closely linked.

AI is weakening that link.

The wrong incentive

Consider an agency whose contract specifies a certain quantity of content, campaigns and reporting every month.

Its financial obligation is to produce those outputs.

But what happens if the evidence suggests that the client doesn't need another campaign?

What if the most valuable action is to improve an existing offer, concentrate on a small group of highly relevant buyers or stop communicating with prospects who have shown no interest?

A sensible agency might recommend exactly that.

Yet the contractual model is still built around producing activity.

The result is a subtle tension between satisfying the agreement and pursuing the customer's commercial interests.

This doesn't mean agencies are deliberately doing unnecessary work. Most are trying to do the best possible job within the model they have inherited.

The question is whether the model remains appropriate.

The temptation of payment by results

One obvious alternative is to pay agencies only when they produce a commercial outcome.

It sounds attractive. The client pays for what matters, and the agency is rewarded for success.

But this creates a different problem.

Marketing agencies rarely control the entire customer relationship.

They may generate an excellent enquiry that receives no sales follow-up. They may create a promising opportunity for a product whose price is uncompetitive. They may identify buyers whose decisions depend on procurement cycles, budgets or circumstances beyond the agency's influence.

Making the agency financially responsible for every outcome it cannot control is not alignment. It is risk transfer.

There is also the danger of rewarding easily counted outcomes over valuable ones.

If an agency is paid for meetings, it has an incentive to generate meetings. Whether those meetings are commercially useful becomes a secondary question.

A flawed measure, once attached to money, can become an excellent way of producing the wrong behaviour.

A more sensible bargain

There is a middle ground.

An agency can agree to take responsibility for a defined commercial programme, rather than a quantity of labour.

The client and agency establish what the business is trying to achieve, which markets and propositions are in scope, what resources and systems are involved and how progress will be measured.

The agency receives a predictable fee for managing that programme.

It is free to choose and improve the activities required, rather than being rewarded for producing an arbitrary amount of work.

The client isn't buying hours. Nor is the agency underwriting the client's revenue.

Both are committing to a commercially focused programme with transparent accountability.

This approach requires discipline. Scope must be clear. Expectations must be realistic. Measurement must distinguish genuine progress from superficial activity.

But it is better suited to a world in which technology can dramatically change the cost of execution.

What should efficiency be worth?

A good agency should be rewarded for the value it creates, not penalised for becoming more efficient.

Equally, clients should benefit when technology makes professional services less cumbersome and more responsive.

The opportunity is not simply to automate the old retainer.

It is to reconsider what an agency relationship is for.

The most important question in a monthly review should no longer be whether the agreed hours were used or the planned outputs delivered.

It should be whether the programme is making meaningful commercial progress and what should happen next.

That seems a much healthier basis for a partnership.