Skip to content

Deliverable-based pricing: how agencies move off the billable hour

Most agencies already suspect the hour is the wrong thing to sell. What stops them is not conviction but mechanics: what exactly replaces it, how you put a number on it, and what you say to the client who has been reading your timesheets for three years.

The Tave teamPublished September 28, 20269 min readAgency profitability

What deliverable-based pricing actually is

Deliverable-based pricing means the client pays a fixed price for a defined piece of work — a landing page, a brand identity, twelve social posts a month — regardless of how many hours it takes you to produce it. The price is attached to the output, not to the time.

The Association of National Advertisers uses almost exactly that definition in its Trends in Agency Compensation research: an output-based fee is negotiated for a specific project or set of deliverables, without regard to the agency labor time involved. It is not a fringe idea invented by software vendors. It is how a growing share of the biggest advertisers in the world now pay their agencies.

Two things it is not. It is not value-based pricing, where the price is set by the client's expected return — that is a further step, and a harder one. And it is not a fixed quote on a vague brief. A deliverable price only works when the deliverable is defined tightly enough that both sides can tell when it is done. Most of this post is about that definition.

Where agencies actually are today

Among agencies, the hour is still the single most common model, but it no longer has the majority. Ignition asked 273 agency managers and executives how they price for The 2025 Agency Pricing & Cash Flow Report. Hourly billing and productized packages came out level.

28%
of agencies bill hourly — and another 28% sell productized or subscription packages or tiered bundles; 25% price by project and 10% by retainer (Ignition, The 2025 Agency Pricing & Cash Flow Report, n = 273).

The buyer side has moved further and faster. In the ANA's 2022 survey of 101 client-side marketers, 82% used fee-based compensation, up from 68% in 2016 — and among the very largest advertisers, the kind of fee changed completely.

53%
of marketers spending $500 million or more a year now pay agencies fixed or output-based fees, up from 5% in 2016 (ANA, Trends in Agency Compensation, 18th edition, n = 101).

The same research found that smaller advertisers are much more likely to stay on labor-based fees: 76% of them still do. That matters for a small agency, and it cuts both ways. Your clients are mostly not the $500-million kind, so some of them will ask for hours out of habit. But the direction of travel is clear, and the ANA gives two reasons for it that apply at any size: output fees remove the review and haggling over agency labor time, and some marketers simply want to pay for what an agency produces rather than how long it takes.

Why the billable hour breaks for a small agency

The case against the hour is usually made in philosophical terms. For a five-to-twenty-person agency it is more practical than that. Three things go wrong, and all three get worse as you get better at your job.

It punishes you for getting faster

The fifth landing page you build takes less time than the first. Templates, component libraries, a better brief, and now AI tools that draft copy and code in minutes. Under hourly billing every one of those improvements is a price cut you hand to the client. The more efficient the agency, the less it earns for the same work.

It turns every invoice into an audit

A client cannot judge whether a logo should have taken fourteen hours or twenty. They can only see the number, and the number invites the question. That is the administrative cost the ANA names: time spent reviewing and defending timesheets instead of doing work. For a small agency the person defending the timesheet is usually the owner.

It hides where the money goes

When everything is billed by the hour, nothing has a margin of its own. You know the agency made or lost money last month, but not which kind of work made it. Deliverable pricing forces you to know the cost of each thing you sell — which is uncomfortable for a quarter and then the most useful number you have.

Step one: define the deliverable before you price it

A deliverable is priceable when a stranger could read its description and agree on whether it has been delivered. Write each one down with the same five parts:

  • A name the client would use. "Campaign landing page", not "Frontend sprint".
  • What is in it. One page, up to 1,200 words of copy supplied by the client, one form, desktop and mobile layouts.
  • How many revision rounds are included. A number, written down. "Two rounds of consolidated feedback" is a term; "revisions as needed" is an open-ended promise.
  • What is not in it. Copywriting, stock photography, integrations beyond the form. The out-list is the part that prevents arguments later.
  • What done means. Approved by the client in writing, or live on their domain — whichever you pick, pick it in advance.

Start with the three things you sell most often and that vary least from client to client. For most agencies that is something like a landing page, a social content package and a brand identity. Bespoke work — a complex platform build, a research project — can stay on project quotes for now. You do not need to convert everything at once, and you should not try.

Step two: price it from your own hours

Here is the part that surprises people: moving off the hour does not mean you stop tracking time. It means time stops being the thing you sell and becomes the thing you measure cost with. Your timesheet history is the best pricing data you own.

  1. Pull the hours logged on the last five to ten instances of the deliverable. Note the typical number and the worst one.
  2. Work out your blended cost per productive hour: salaries, contractor fees and overheads for a month, divided by the hours your team actually spends on client work in that month — not the hours they are paid for.
  3. Multiply both the typical and the worst-case hours by that cost. That is the range of what the deliverable costs you to make.
  4. Set a price that still leaves you a margin you are happy with on the worst case — not the typical one. You are now carrying the overrun risk the client used to carry, and the price has to pay for that.
  5. Sanity-check the price against what the client would have paid you hourly for a bad run of the same work. If they are close, you are offering the client certainty at roughly their old worst case.

A worked example

An illustration, with invented inputs and real arithmetic. A small studio has built six campaign landing pages. The hours logged were 24, 26, 30, 31, 34 and 41 — typically about 31, worst case 41. The studio's blended cost is €48 per productive hour, and it used to bill clients €85 an hour.

  • What it costs. 31 hours × €48 = €1,488 for a typical page; 41 × €48 = €1,968 for the worst run.
  • What hourly billing charged. 31 × €85 = €2,635 for a typical page; 41 × €85 = €3,485 for the worst run.
  • The deliverable price. €3,400, two revision rounds included. Margin on a typical page: 56%. On the worst run: 42%.
  • When the team gets faster. At 24 hours the cost falls to €1,152 and the margin rises to 66%. Billed hourly, the same improvement would have shrunk the invoice to €2,040.

That last line is the whole argument in arithmetic. Under the hour, the studio's efficiency gains belong to the client. Under a deliverable price, the client gets a fixed number at roughly their old worst case, and the studio keeps what it learns.

Two cautions. The margin figures are only as good as the cost rate, so be honest about productive hours — a team that bills 60% of its paid time has a much higher real cost per hour than salary divided by 160. And six data points is a start, not a law: revisit the price after every five or so deliveries.

Step three: make the switch without a difficult conversation

Most agencies imagine the switch as one uncomfortable meeting with every client. It does not have to be. The path that tends to work is gradual and starts where there is no history to unlearn.

  1. Quote every new client on deliverables from a fixed date. They have no hourly habit to break.
  2. Keep logging time internally on everything, including the new deliverables. That log is what tells you whether your prices are right.
  3. For existing clients, offer the deliverable price at the next natural break — a new project, a renewal, a new year — framed as a fixed number instead of a variable one. "You will know the cost before we start" is an easy sentence to say yes to.
  4. Turn retainers into packages: a monthly price for a named set of deliverables, with what happens to unused ones written down (they lapse, or they roll into the next month only).
  5. After a quarter, compare each deliverable's real cost against its price. Raise the prices that are losing, and leave the rest alone.

Notice what is missing: a rate card on the invoice. If a client insists on seeing an hourly rate, you can tell them the number your prices are built from — but the invoice names the deliverable, not the hours. The moment hours reappear on an invoice, the client is back to auditing them.

What can go wrong, and how to handle it

Deliverable pricing moves risk from the client to you. That is the trade, and it is a good one if you manage four specific failure modes.

You underestimate

It will happen, especially on the first few. Pricing against the worst case rather than the typical case absorbs most of it. When a deliverable loses money twice in a row, the price is wrong — change it for the next client rather than hoping.

The deliverable grows

A fixed price makes scope drift more expensive for you, not less, because there is no longer an hourly meter quietly catching it. This is why the out-list and the revision count exist. Anything outside them is a change to the agreed work, and it needs its own price and the client's written approval before you start it.

The client wants hours anyway

Some will, particularly larger organisations whose procurement teams are built around rate cards — the ANA's 76% figure for smaller advertisers is a reminder that labor-based fees are still the norm on the buyer side outside the very largest. Offer the deliverable price alongside a clear statement of what is included. Most people who ask for hours are really asking for predictability, which a fixed price gives them better.

Every deliverable becomes a custom one

If each quote needs a new definition, you have renamed project pricing, not replaced the hour. Keep the menu short, keep the definitions stable, and push genuinely unusual work to a separate project quote.

The short version

Deliverable-based pricing is not a leap of faith. It is a short list of named things you sell, each with an in-list, an out-list and a revision count, priced from your own timesheet history against the worst case. Hours stay — as an internal instrument for knowing what each deliverable costs you. They just stop appearing on invoices. This is also the model Tave is built around: each deliverable carries its price and its included revisions, and logged time feeds your margin, never the client's bill.

Sources

Tave turns scope drift into a change order your client can approve in one click — so the work you actually did becomes work you actually billed.