Selling AEO to a client who only asks about rankings
A practical script for the conversation, the two slides that land, and the trap of promising movement you cannot deliver in a quarter.
Every agency currently has a version of the same problem. You can see that AI search is taking a growing share of your clients' discovery, you know the work required is real, and the client is on a monthly call asking why keyword fourteen dropped two places.
The instinct is to educate — to explain how retrieval works, why generative answers differ from a results page, what an entity is. That conversation almost never lands, because it asks the client to care about mechanics before they've accepted there's a problem. The sequence has to run the other way.
Open with their own data, not the category
Before the meeting, run their five most commercially important buying questions through three engines and screenshot the answers. Not brand queries — the unbranded ones a prospect would actually type. Then put the screenshots on one slide, with their competitors' names visible and theirs absent.
This is the whole pitch. It takes twenty minutes to prepare and it does something no amount of category education can: it makes the problem concrete, specific and about them. Clients who have politely ignored six months of AEO explainers tend to sit up when they see a competitor named in a recommendation they thought was theirs.
Five real buying questions. Three engines. The answers as they appear, with competitor names highlighted. No commentary on the slide — say it out loud instead.
Answer the rankings question honestly
The next question is always some version of "so does this replace what we're paying you for?" The honest answer is no, and saying so buys you more credibility than any amount of positioning.
Traditional search still drives real volume for most businesses, and rank tracking still measures something true about it. What has changed is that it no longer measures everything, and the part it misses is growing. The framing that works is additive: same discipline, second surface, and a meaningful overlap in the underlying work.
That overlap is worth being specific about, because it's your strongest commercial argument. Crawl access, clean information architecture, structured data, clear entity signals and genuinely useful content help both. You are not proposing a second programme of work. You are proposing to measure a second outcome from largely the same programme, plus a defined set of additions.
Set the timeline before they set it for you
This is where agencies get into trouble. AI visibility work is slower than a client expects and slower than a competitive pitch will imply. Engines have to re-crawl, models have to start drawing on the new material, and entity work depends on third parties you don't control.
Four to eight weeks for the first measurable movement, two quarters for a convincing trend. Say it in the first meeting, write it in the proposal, and repeat it when someone asks in week three. An agency that sets this expectation early looks careful; one that sets it in month two looks like it's making excuses.
Promising visible AI visibility gains inside a quarter. Someone in your pitch process will be tempted, because a competitor implied it. It is the fastest route to a lost account in month four.
Report on three numbers, not thirty
The reporting failure in this category is volume. A twenty-page AI visibility deck full of engine-by-engine breakdowns will be skimmed once and never opened again. Three numbers, reported the same way every month, build a habit.
Presence rate
The share of tracked buying questions where the client is named at all.
Share of voice
Their slice of the mentions against the three or four rivals that keep appearing beside them.
Readiness score
The one number that moves when you do the technical work, which is how you show progress while the visibility numbers are still lagging.
That third number matters more than it looks. It's the metric you control, it responds within weeks, and it gives the client something to see during the months when presence hasn't moved yet. Without it, the first quarter of an AEO engagement is a long silence.
On pricing it
Two approaches are working. The first is a one-off audit and fix plan — a fixed-price piece of work that produces a document, a prioritised backlog and a baseline. It's easy to sell, easy to scope, and it converts well into retained work once the client sees the size of the backlog.
The second is folding AI visibility into an existing retainer as an additional reported outcome, with a modest uplift for the monitoring and the extra implementation. This is the better long-term structure and a harder first conversation, because it looks like a price increase unless the first slide has already done its job.
What isn't working is selling AEO as a separate product line with its own brand and its own retainer. Clients experience it as a second SEO invoice, and the overlap in the underlying work makes it hard to defend when they look closely.
Whichever route you take, put white-labelled reporting in front of the client from month one. The agencies doing well here are not the ones with the best methodology. They're the ones whose clients see a consistent, legible number every month and have quietly started asking about it themselves.
- →Lead with their own answers and their competitors' names. Education comes after.
- →Say plainly that this doesn't replace SEO. The overlap in the work is your best argument.
- →Set the four-to-eight-week expectation in meeting one, not month two.
- →Report three numbers monthly. Readiness carries the story while visibility lags.