Choose a project when the SEO problem, output and finish line are stable. Choose a monthly retainer when the next priority will change as work ships and somebody must keep diagnosing, implementing, measuring and reprioritising.
Start with this:
| Situation | Better starting model |
|---|---|
| One known technical defect | Project |
| Migration or rebuild | Project with pre- and post-launch support |
| Scope or opportunity is unclear | Paid discovery |
| Internal team can execute but needs judgement | Advisory |
| Known backlog needs production capacity | Implementation sprint |
| Search spans changing technical, page, authority and conversion priorities | Managed retainer |
| Foundation needs a bounded rebuild, then ongoing growth | Project-to-retainer hybrid |
| Nobody can implement | Fix ownership before buying more SEO |
The model is not the strategy. It is the commercial container around the work.
The decisive question: is the problem stable?
A project works when you can reasonably define:
- the affected site, pages or systems;
- the desired output;
- the dependencies;
- the acceptance test;
- the decision owner;
- the end state; and
- what happens after handover.
A retainer works when evidence will change the order of work. A technical release may reveal a page problem. A stronger page may expose an authority gap. Better visibility may expose a conversion failure. The provider needs authority to move the backlog as those signals arrive.
Monthly SEO retainer vs project-based SEO
| Factor | Monthly retainer | Project-based SEO |
|---|---|---|
| Problem shape | Moving or multi-workstream | Stable and bounded |
| Priority changes | Expected through a defined process | Controlled through change request |
| Commercial structure | Recurring fee and review cycle | Fixed, milestone or phase price |
| Finish line | Continue while the economics and work justify it | Accepted output or end state |
| Implementation | Often recurring and shared | Must be explicit in the project |
| Measurement | Ongoing evidence and reprioritisation | Completion proof plus observation plan |
| Main strength | Adapts to what the market gives back | Cost and outcome boundary |
| Main failure | Repeating activity without decisions | Recommendations nobody implements |
Neither model guarantees rankings or revenue.
When a project is the right answer
Project-based SEO is strong for:
- technical audits with decision-ready tickets;
- migrations and redesigns;
- crawl, rendering or indexation repair;
- information architecture;
- service, category or location-page systems;
- analytics and conversion remediation;
- content consolidation;
- international or platform setup;
- recovery diagnosis;
- expert training; and
- a defined commercial-page sprint.
The project scope should include implementation or a credible handover path. “Complete audit delivered” is not a business result when the client has no developer, editor or release capacity.
A project must define aftercare
Specify:
- post-release verification;
- observation period;
- defect correction;
- rollback or incident support;
- training and documentation;
- open risks;
- maintenance owner; and
- whether follow-on work is optional.
A project can finish cleanly without pretending the search market stops changing.
When a monthly retainer is the right answer
A retainer is strong when:
- the website or market changes regularly;
- several workstreams interact;
- the highest-value priority is not static;
- recurring source data changes decisions;
- content and commercial pages need ongoing ownership;
- authority work requires sustained relationships;
- local or multi-location facts require governance;
- internal releases need senior review;
- incidents or migrations require continuity; or
- search is important enough to operate as a channel.
The retainer should define the operating system:
- commercial outcomes and markets;
- search surfaces;
- provider and client authority;
- adaptive backlog;
- proof states;
- implementation;
- data access;
- decision cadence;
- change control;
- ownership; and
- exit.
A fixed number of blogs and links every month is a subscription, not necessarily a managed search system.
When advisory is enough
Use advisory when an internal team already has:
- capable developers and editors;
- production access;
- reliable measurement;
- subject-matter input;
- authority and PR capacity where needed; and
- leadership able to make decisions.
The provider can own:
- diagnosis;
- prioritization;
- briefs and tickets;
- technical or editorial quality assurance;
- release QA;
- incident advice;
- training; and
- performance interpretation.
Advisory fails when the internal team is theoretically capable but has no actual capacity.
The hybrid model
Many businesses benefit from establishing the foundation before starting a retainer.
A healthier sequence may be:
- Paid discovery: verify the market, baseline, dependencies and economics.
- Foundation project: repair the platform, architecture, money pages or measurement.
- Observation: confirm implementation and collect useful evidence.
- Managed retainer: address the next priority when recurring work is justified.
This prevents an expensive diagnostic phase from being disguised as months of “ongoing optimization”.
The reverse can also work: an ongoing program may break out a migration or platform rebuild as a separate project with its own risk, budget and acceptance controls.
Compare the complete cost
Compare the complete cost of each model rather than setting a project fee beside one month of a retainer.
Include:
| Cost | Project question | Retainer question |
|---|---|---|
| Diagnosis | Is discovery included? | Is diagnosis repeated as evidence changes? |
| Implementation | Who ships the output? | How much production is owned each cycle? |
| Internal team | What handover capacity is required? | What recurring client capacity is required? |
| Tools and data | Included or separate? | Included, capped or passed through? |
| Authority | Is it in scope? | How are relevance and placement proven? |
| Rework | What correction is included? | How are defects and misses handled? |
| Observation | Who measures after delivery? | Which source-backed layers are reviewed? |
| Exit | What assets and access remain? | What does the final handover include? |
A lower fee with a large hidden implementation burden may be the more expensive model.
Use this six-question decision
1. Can we name the finish line?
If yes, a project may fit. If the answer is “grow search”, keep going.
2. Will evidence change the next priority?
If yes, you need an adaptive backlog, through a retainer, advisory relationship or staged projects.
3. Who implements?
If nobody can ship, neither model works. Add implementation or fix internal capacity.
4. Does the work need continuity?
Migrations, authority relationships, governance and recurring releases may require the same senior context over time.
5. Can the result be observed after the project ends?
Define the owner, source, observation window and attribution boundary.
6. What is the clean exit?
Projects need accepted handover. Retainers need account, asset, evidence, backlog and risk transfer.
What good execution looks like in either model
A strong retainer
- changes its priorities when the evidence changes;
- names the implementation owner;
- uses capacity around the most valuable ready work;
- turns meetings into decisions;
- proves what reached production;
- keeps KPI definitions stable; and
- uses review points supported by visible delivery.
A strong project
- ends in an accepted output or business-ready end state;
- controls scope and changes explicitly;
- identifies dependencies before work begins;
- confirms the internal capacity required;
- hands over source files, documentation and tests;
- assigns post-project observation; and
- keeps moving work out of a falsely static scope.
What should the agreement say?
For either model, define:
- commercial job;
- baseline;
- exact scope and exclusions;
- deliverables;
- responsibilities;
- implementation;
- acceptance;
- measurement;
- dependencies;
- change control;
- pricing;
- ownership;
- claims boundaries; and
- exit.
Have qualified counsel review the agreement. This page is commercial and operating guidance, not legal advice.
FAQ
Is a monthly SEO retainer better than a project?
Only when the problem and priority will keep changing and recurring diagnosis, implementation and measurement are justified. A bounded problem is often better handled as a project.
Is project-based SEO cheaper?
It can be, but compare the total cost of diagnosis, implementation, internal capacity, rework, observation and maintenance. A cheap audit nobody ships is poor value.
Can we start with a project and move to a retainer?
Yes. Paid discovery or a foundation project can prove the opportunity, working relationship and implementation path before a recurring commitment.
Which SEO work fits a project?
Audits, migrations, technical repairs, architecture, measurement remediation, defined page systems, consolidation and training can all have clear end states.
Which SEO work fits a retainer?
Changing multi-workstream growth, recurring technical and page governance, authority development, local operations, ongoing release QA and evidence-led reprioritisation.
What if our internal team implements everything?
Consider advisory or a narrower retainer for diagnosis, briefs, decisions and QA. Confirm that the team has real capacity, not just theoretical responsibility.
How long should an SEO retainer run?
While the commercial opportunity, implementation path and evidence justify recurring work, with review and exit points defined. Contract length should not substitute for visible delivery.
Buy the model the work deserves
The right commercial model makes the work easier to define, ship and judge. Start with the SEO system, then use the detailed guide to what an SEO retainer should include or establish the first 30 days of SEO onboarding.