A retainer is right when the value of the work compounds, and wrong when the work completes. This post gives the test we apply to our own proposals, shows how our published rate card splits the same service into a project and a retainer, and sets out what a fixed scope must say in writing to stay fixed.
Compounds or completes: the question that decides the shape
Work compounds when doing more of it keeps producing a return: content that keeps ranking, campaigns that keep learning, a site that keeps changing and needs looking after. Stopping it is not finishing the job; it is abandoning an ongoing one.
Work completes when there is an end state, after which more of the same activity returns little. A migration is finished or it is not. A redesign ships or it does not. An audit produces a fix list and stops.
A retainer wrapped around completing work goes wrong in one of two ways. It drags on past the point of diminishing returns, or it narrows quietly into activity that exists to justify the invoice. A fixed scope wrapped around compounding work goes wrong the other way: the results fade once the work stops.
- Fits
- An end state: a migration, a redesign, an audit
- Goes wrong when
- Wrapped around compounding work: results fade once it stops
- Fits
- An audit whose fix list implies ongoing work
- Goes wrong when
- The retainer is presold before the diagnosis exists
- Fits
- Content that keeps ranking, campaigns that keep learning
- Goes wrong when
- Wrapped around completing work: it drags on or pads activity
The proposal test: is it build-shaped or retainer-shaped?
The label on a proposal is the least reliable signal, because “package” and “programme” are used for anything. We read three things instead, and you can read them in any proposal you receive, including ours.
| Read this | Build-shaped (a project) | Retainer-shaped |
|---|---|---|
| What success is measured against | A delivered thing: pages shipped, a migration live, a fix list closed | A moving number: organic sessions, cost per conversion, pipeline |
| Can the criteria read “done”? | Yes, once, and they stay done | No; next month brings the same standard to clear again |
| The team’s time in month six against month one | Front-loaded, tapering towards handover | Roughly level for as long as the engagement runs |
If a proposal cannot say what the team’s time looks like in month six, it has not decided which shape it is. That is worth asking before you sign, because the answer tells you whether you are buying a result or a standing commitment.
How our rate card splits the same work two ways
We publish most services in both shapes on our pricing page, and the line between them is drawn by what the work is, not by what we would rather sell. Three examples from that rate card:
| Service | Project shape | Retainer shape | What decides between them |
|---|---|---|---|
| SEO audit fixes | Fix sprint: a fixed quote, four to eight weeks, the review re-run at the end | Monthly fix programme, three-month minimum, for lists too long for one sprint | How many findings came back red: six fixes and forty are different jobs |
| Technical SEO | Fix programme: a fixed quote, four to eight weeks | Monthly maintenance, three-month minimum | Release cadence: a site that deploys daily breaks more things than one that deploys quarterly |
| Migration SEO | Fixed quote, run alongside the build | None; it ends when the move ends | URL count and how much of the structure changes |
The SEO audit row is the clearest case: one fix list, shipped as a sprint or worked through monthly. In our rate card the sprint runs 4 to 8 weeks, and the programme has a 3-month minimum.
We recommend the sprint for a short list, and the programme for a long one or a site that keeps generating new faults. The choice follows the list, not our preference.
What a fixed scope must say in writing
A fixed-scope proposal names a deliverable, a price and a definition of done: all three, not just the first two. “Audit and fix list” at a fixed price is a project. “Audit and fix list, then ongoing implementation at the same rate” is not, because the second half has no end.
Our sample fixed project scope asks for four more things: dependencies, acceptance evidence, change handling and payment milestones. Acceptance evidence is the useful test. Could a third party check the delivered work against the scope and agree it is complete? If yes, the scope was real.
Change handling is what keeps a fixed fee honest. Without a written route for work that turns out larger than expected, a project either loses money quietly or gets rushed to protect the fee. Neither serves the client.
When the work starts as a project and becomes ongoing
This is common, and it is the right order. An audit produces a fix list. Some fixes are one-off, such as a broken redirect chain; some imply ongoing work, such as a content programme. The honest move is to scope and deliver the project, then let the client decide separately whether the follow-on justifies a retainer.
What we avoid is preselling the retainer before the diagnosis exists to justify it. A client sold a retainer for work that completes notices, a few months in, that the reports are padding activity. That costs more trust than the smaller project would have cost revenue.
When a retainer is the right answer after all
This post argues against the wrong retainer, not against retainers. Choose the retainer when:
- The measure is a moving number that falls back when the work stops, such as rankings in a contested category or a paid account’s cost per conversion.
- The site changes often enough to keep breaking things, so maintenance is continuous by nature.
- The fix list is long enough that a sprint would either rush it or run past its end date.
A retainer that is too small for the work is its own problem; why we turn down small SEO retainers covers that case.
Written by
Parvessh Agarwal
Founder and Chief Executive, Vinove
Parvessh founded Vinove in 2004 with one idea: build focused technology companies that earn their place in real markets. His focus is on the long game, on companies that survive and serve real customers long after launch.
LinkedIn →
