What should you ask before signing a year-long GEO or AEO retainer?

The baseline question

A twelve-month retainer signed without a baseline has no way of demonstrating it worked. Before signing, establish what is being measured, how it is captured, and who captures it. If the measurement lives inside the vendor's own dashboard and cannot be reproduced from your Search Console, your logs and your analytics, you are buying their marking of their own homework.

The month-two question

This is the one that matters. Ask directly: if the first month of diagnosis shows the real problem is a rendering or entity issue rather than an AEO issue, what happens to the retainer?

A practice with a method will say the scope changes. A retainer built on filling hours will explain why the AEO work is still worth doing. You learn more from this answer than from the entire pitch deck.

Scope and ownership

• What specifically ships each month, and what happens when it does not
• Who owns the content, the code, the schema and the monitoring configuration at the end
• Whether any deliverable depends on continued access to the vendor's tooling
• Notice period, and whether there is a break clause at three or six months
• Who does the work. Named people, not a company name

Why retainers fail in this category specifically

AEO work is front-loaded. The diagnosis, the structural fixes and the entity work are heavy early and light later. A flat twelve-month fee spreads even payment across uneven work, which suits the vendor's cash flow and rarely suits yours.

Engagements here are priced per piece of work rather than per month, and no retainer is created by default. If ongoing advisory genuinely helps after the structural work is done, it gets set up then, at the size it actually needs.