Guides
The first 30 days of an AEO engagement
An audit-led first-month AEO plan for agencies that need to establish a baseline, prioritise work, and set an honest client reporting rhythm.
A first month of AEO work should end with a client-approved baseline and a prioritised work plan, not a promise that an AI answer will change on a particular date. For an agency, that makes the first 30 days an onboarding sequence: agree what to measure, find the conditions the client can change, assign the work, and decide how the next review will judge it. This guide covers that engagement sequence. For the evidence and prioritisation method inside the readiness-audit deliverable, see AI readiness audit methodology for agency remediation.
Consider a fictional agency, Northline, starting work with a B2B software client. The client has seen a competitor named in several buyer conversations and wants to know why, but nobody has yet agreed which questions matter, what the current answer set looks like, or which site conditions are worth fixing first.
Start with the client decision, not a list of AI platforms
The first useful AEO question is the decision the client's buyer is trying to make. Northline begins by turning that decision into a small, documented prompt set, such as the comparisons, problem questions, and category queries the client's sales team hears most often.
The point is not to create a large prompt inventory in week one. It is to establish a scope the client can recognise and an agency can repeat at the next review.
Record the scope in the kickoff note:
- the client domain and the products or services in scope
- the buyer questions to measure
- the competitor set, if the client has named one
- the AI platforms included in the baseline
- the date, geography, and language assumptions that can affect a result
A baseline is a starting record, not a verdict on the client's brand. Generated answers vary, and a single answer does not explain why an engine chose a source or whether it will make the same choice again. The agency's job is to make that uncertainty visible before it becomes a reporting dispute.
Build a baseline that someone can review later
A credible baseline combines the observed answers with the conditions on the site that may affect whether it can be read and understood. The two are related, but they are not interchangeable.
For Northline, the week-one deliverable is a short baseline memo with three parts:
- Measurement scope. The agreed prompts, platforms, competitor set, and run date.
- Observed pattern. A plain description of what appeared in the sampled answers, including where the pattern was mixed or unclear.
- Readiness findings. The site conditions that merit technical or content investigation, ordered so the client can see what should happen first.
This keeps the agency from presenting an AI visibility score as a diagnosis. A score can indicate where to investigate, but the client still needs the evidence, the owner, and the next verification point.
Maverank's AI readiness model is useful here because it separates recurring answer tracking from a readiness audit that produces a prioritised fix list. The product tracks prompts across ChatGPT, Claude, Gemini, and Google AI Overviews, while the audit evaluates site conditions that can be reviewed and changed.
Audit the conditions that can block useful work
An AEO audit should give the agency a practical sequence, not a long checklist for its own sake. Northline starts with conditions that can make later content work hard to evaluate, then moves to structure, authority, and answerability.
For example, the first review can ask:
- Can relevant crawlers fetch the pages the client wants cited?
- Does the page deliver its primary content without relying on a browser to assemble it?
- Can a reader and a machine identify what the page answers, who stands behind it, and which claims are supported?
- Are there obvious gaps between the buyer questions in the prompt set and the client's public content?
These questions do not guarantee a future citation. They help the agency distinguish a concrete remediation project from a vague request to make a brand appear in AI answers.
The AEO glossary explains the category term, but the client conversation should stay grounded in the specific page, prompt, and owner in front of the team. If crawler access is the problem, it belongs with the technical owner. If a comparison page does not answer the buyer's question, it belongs with the content owner.
Turn findings into a 30-day work plan
The last half of the month is where a baseline becomes a service the client can understand. Northline turns each selected finding into a small work item with an owner, an expected effort, and a verification method.
A practical first-month plan can look like this:
| Week | Agency deliverable | Client decision |
|---|---|---|
| 1 | Kickoff scope and first prompt baseline | Confirm the questions, platforms, and owners that define the engagement. |
| 2 | Readiness audit and evidence review | Choose the few findings worth addressing before broader content work. |
| 3 | Prioritised remediation brief | Approve the work items, technical owners, and content owners. |
| 4 | First client review and next measurement date | Agree what will be rechecked and what the result can, and cannot, show. |
This is deliberately narrower than an ongoing retainer. The first month establishes how the agency will measure and deliver the work, while a retainer is where the team repeats the measurement, verifies completed fixes, and adapts the plan as evidence changes.
For Northline, a sensible week-three brief might include one crawlability fix, one page-level content improvement, and one question that needs more evidence before anyone changes the site. That last item matters. An agency earns trust when it can say that a finding needs investigation rather than turning every observation into billable implementation.
Report the first month as a method and a decision
The first client review should make the work legible without manufacturing certainty. Show the prompt scope, the observed baseline, the selected findings, the work owners, and the date of the next measurement.
Do not frame the review as proof that a technical change caused a future answer to move. AI platforms change, prompts vary, and external sources can change independently of the client's work. The defensible statement is smaller: the agency measured a defined set of questions, completed or scheduled specific remediation, and will recheck the same scope on an agreed date.
Agencies that need a client-facing rhythm can use Maverank for agency delivery to keep prompt tracking, readiness findings, and reporting in one workflow. The product's white-label PDF reports and scheduled delivery are available from the Studio plan, while the client conversation still needs the agency's own judgement about scope and limits.
Frequently asked questions
What should an agency deliver in the first month of AEO work?
An agency should deliver an agreed measurement scope, a dated baseline, a prioritised set of readiness findings, assigned remediation work, and a date for the next review. The deliverable should explain what was observed and what will be verified without promising a specific AI-answer outcome.
How many prompts should an AEO baseline include?
Start with a small set of buyer questions the client and agency can both defend as relevant. The right number depends on the service scope, but a repeatable and documented set is more useful than a large unreviewed list.
Does fixing a readiness finding guarantee more AI citations?
No. A readiness finding identifies a site condition the team can investigate and improve, but AI platforms do not publish stable ranking factors or guarantee future citations.
When should the agency remeasure AI visibility?
Set the next measurement date during the first-month review and compare the same documented prompt scope before changing the conclusion. A repeatable measurement rhythm is more useful to a client than frequent snapshots with changing assumptions.
Start with one scoped client baseline
A first-month AEO engagement works when the client can see what is being measured, what the agency recommends changing, and how the next review will test the work. Start a Maverank trial when you need to turn that baseline into an ongoing agency workflow with tracked prompts, prioritised readiness findings, and client-ready reports.