Agency pricing models compared: hourly, fixed, retainer, value and productized
Agencies price in five ways: time and materials, fixed price, retainer, value-based and productized. They differ in who carries delivery risk and where the margin ceiling sits. Hourly is safest and capped. Fixed and productized pay best on repeat work. Retainers buy predictability. Value pricing needs a measurable result. Most agencies should run two or three together.
A pricing model is a rule for dividing risk. Each of the five models an agency can use gives the client something different to buy and leaves a different party holding the cost when the work goes sideways. Choosing between them is one of the few decisions that changes your margin without changing your team.
This page puts the five side by side. Each has its own detailed article, linked from the billing hub and from the sections below.
The five models in one table
| Time and materials | Fixed price | Retainer | Value-based | Productized | |
|---|---|---|---|---|---|
| Client buys | Hours | A defined outcome | Reserved monthly capacity | A business result | A standard package |
| Price built from | Rate x hours worked | Estimated hours x rate, plus contingency | Monthly hours x rate, less commitment discount | A share of the client’s gain | A published price per unit |
| Delivery risk sits with | Client | Agency | Shared | Agency | Agency |
| Margin ceiling | Capped by rate and hours | High if you finish early | Moderate, steady | Highest | High at volume |
| Revenue predictability | Low | Low to medium | High | Low | Medium to high |
| Sales effort per deal | Low | High | Medium, once | High | Low |
| Needs from you | Time tracking, trust | Accurate estimates, tight scope | Usage reporting, overage discipline | Measurable value, senior buyer | Narrow offer, steady demand |
| Breaks when | Client audits every line | Scope is vague | Usage and fee drift apart | You are easy to substitute | You accept exceptions |
Read the last row first. Every model works until a specific condition fails, and knowing that condition is more useful than knowing the model’s advantages.
The same work under each model
To make the differences concrete, take one piece of work: 100 hours of effort by your estimate, a standard rate of $150, and a loaded cost of $90 an hour.
| Model | How you would price it | Revenue if it takes 100 hours | Revenue if it takes 130 hours |
|---|---|---|---|
| Time and materials | $150 an hour | $15,000 | $19,500 |
| Fixed price, 20% contingency | 100 x $150 x 1.20 | $18,000 | $18,000 |
| Retainer, 10% discount, overage billed at $165 | 100 x $150 x 0.90 | $13,500 | $13,500 + (30 x $165) = $18,450 |
| Value-based, client gain of $120,000 | 25% of gain | $30,000 | $30,000 |
| Productized, after practice cuts effort to 70 hours | Published price | $16,000 for 70 hours | $16,000 |
And the margin each produces at a cost of $90 an hour.
| Model | Profit at 100 hours | Profit at 130 hours |
|---|---|---|
| Time and materials | $15,000 - $9,000 = $6,000 | $19,500 - $11,700 = $7,800 |
| Fixed price | $18,000 - $9,000 = $9,000 | $18,000 - $11,700 = $6,300 |
| Retainer | $13,500 - $9,000 = $4,500 | $18,450 - $11,700 = $6,750 |
| Value-based | $30,000 - $9,000 = $21,000 | $30,000 - $11,700 = $18,300 |
| Productized, 70 hours | $16,000 - $6,300 = $9,700 | Same, if the process holds |
A few things stand out. Time and materials is the only model where an overrun raises your profit, which is exactly why clients distrust it. Fixed price beats hourly when the estimate holds and loses to it when the estimate slips by 30%. The retainer looks weakest per month and earns its place over twelve months of revenue you did not have to resell. Value pricing dominates the table, in the minority of cases where it is available.
Time and materials
Suits: new agencies, unfamiliar work, clients whose requirements are still forming, and any engagement where the client wants to steer week by week.
Does not suit: clients with a hard budget, or buyers who need one number for an approval process.
The model depends on trust and on reporting. Give a budget estimate, report against it weekly, and warn early. Your revenue ceiling is headcount times billable hours times rate, so the two things to manage are your utilization rate and your realized rate.
Fixed price
Suits: work you have delivered several times, with written requirements and a client who makes decisions.
Does not suit: first-of-a-kind projects, heavy integrations with systems you have not seen, and clients who discover what they want by looking at what you built.
Fixed price pays you for accuracy. An agency that estimates well and controls scope earns more per hour on fixed price than its rate card says. One that estimates loosely subsidizes its clients. The break-even math and a scoring sheet for the decision are in fixed price vs time and materials. The scope itself belongs in a statement of work.
Retainer
Suits: ongoing, variable work after a build: support, iteration, optimization, content, advisory.
Does not suit: a single large deliverable, or a client whose needs come in two bursts a year.
A retainer trades a little rate for a lot of predictability. The trade is good when the discount is tied to a minimum term, rollover is capped and overage is billed. It is bad when any of those is missing. Sizing, discount schedule and sample clauses are in the pillar on retainer pricing.
Value-based
Suits: engagements tied to a number the buyer owns, where you have a track record that makes you hard to replace.
Does not suit: commodity work, procurement-led purchases, and outcomes nobody can measure.
It is worth being honest about how often this applies. Most quotes described as value-based are fixed prices built from hours. The test, and the conversation that makes real value pricing possible, are in value-based pricing.
Productized
Suits: agencies with one type of job they do repeatedly for one type of client, and enough inbound demand to fill a standard offer.
Does not suit: generalists, or shops whose appeal is that they will build anything.
The margin comes from repetition and from removing the sales process. The failure comes from customization creeping back in. The operational requirements are in productized services.
Hybrids worth using
Most engagements use more than one model. These combinations hold up.
| Combination | How it works |
|---|---|
| Fixed discovery, then fixed build | Discovery removes the unknowns, so the build can be priced with normal contingency |
| Fixed build, then retainer | The project is priced once, the ongoing work monthly |
| Retainer plus project carve-outs | Anything above a size threshold, such as 20 hours, is quoted separately |
| Productized entry offer, then custom work | A fixed-price audit or workshop leads into a scoped project |
| Fixed price plus hourly change requests | The scope is fixed, changes are priced as they arise |
One hybrid to avoid: time and materials with a not-to-exceed cap. You are paid by the hour if you come in under and you absorb the loss if you go over. It combines the ceiling of hourly billing with the risk of fixed price.
A second arrangement gets called a pricing model and is really a staffing model: placing your people inside the client’s team at a monthly or daily rate. The pricing is time and materials. The commercial and legal questions are different, and they are covered in staff augmentation.
Which mix suits which agency
| Agency | A sensible mix |
|---|---|
| Solo consultant | Day rate or fixed-fee engagements, plus one or two advisory retainers |
| New studio, under five people | Time and materials while you build estimating data, first retainers from project clients |
| Established dev shop, 10 to 30 people | Fixed price by phase for builds, retainers for support, hourly for changes |
| Marketing or content agency | Deliverables retainers as the base, fixed-price campaigns on top |
| Specialist consultancy | Fixed and value-based engagements, a productized diagnostic as the entry point |
The pattern across all of them: use time and materials to learn, fixed price to profit from what you have learned, retainers to make the revenue steady, and the two premium models where your position supports them.
Choosing for a specific deal
Four questions settle most cases.
- Can the scope be written down and approved? If no, time and materials or paid discovery.
- Have we done this before? If no, time and materials, or fixed price with heavy contingency after discovery.
- Is the work ongoing? If yes, a retainer.
- Is there a number the buyer owns, and are we hard to replace? If yes to both, have the value conversation.
Whatever the answer, the price has to clear your cost floor. That calculation, for a solo consultant and for a small firm, is in how to price consulting services.
Common questions
- What are the main agency pricing models?
- There are five: time and materials, where the client pays for hours worked; fixed price, a set fee for a set scope; retainer, a recurring monthly fee for reserved capacity; value-based, a fee tied to the worth of the outcome; and productized, a standard package at a published price.
- Which pricing model is best for a new agency?
- Start with time and materials while you learn how long your work takes, and keep careful time records. Move repeat work to fixed price once you have delivered it three times. Add retainers as soon as a project client needs ongoing help.
- Which pricing model has the highest margins?
- Value-based and productized work have the highest ceilings, because the price is detached from hours. They also need the most to go right. Fixed price on familiar work is the most dependable way for an ordinary agency to earn more than its hourly rate.
- How do I move a client from hourly to a retainer or fixed price?
- Use the data you already have. Six months of hourly invoices show average monthly usage, which sizes a retainer. Three similar hourly projects show typical hours, which supports a fixed price for the fourth. Present the change as predictability for the client's budget.