Skip to content

The case against

When you do not need AEO yet

Plenty of businesses are being sold AI visibility while a more ordinary problem eats the same budget. Here are the cases where the right move is to spend the money elsewhere and revisit in six months.

The short answer

Skip it for now if your buyers do not research online, if your site cannot be crawled, if nobody can publish changes quickly, if you have no way to measure a baseline, or if your revenue depends on a channel you have not finished building. In each case the constraint upstream will absorb whatever an answer engine sends you.

What ranked on 18 September 2026: Across sixteen commercial searches in this category, not one first-page result argued that a reader might not need the service. That absence is the reason this page exists, and it is worth remembering when reading any of them.

Your buyers do not start with a search

Some demand simply does not route through research. If your pipeline comes from tenders, referrals, field sales or a marketplace where the platform owns discovery, AI visibility changes a number nobody in your business is paid to move.

  • Check where the last twenty deals actually came from before buying visibility anywhere
  • If the honest answer is introductions, the budget belongs in whatever produces introductions

The site cannot be read

An answer engine cannot quote a page it cannot fetch or render. This is ordinary engineering, it is unglamorous, and it is the most common blocker on sites that arrive asking about AI.

  • Content that only exists after a script runs, on a crawler that may not run it
  • Key pages behind a login, an interstitial or a region gate
  • A site where every page returns a success status, including the missing ones, so nothing can tell content from absence

Nothing gets published

If a copy change takes six weeks to clear legal, brand and a committee, no amount of external advice will produce a different outcome than last year's SEO retainer did. Fix the path to publish first. It is cheaper and it improves everything else you buy.

You cannot measure a baseline

Without a fixed question set sampled before the work starts, you will be unable to tell a model update from your own effort, and both you and your vendor will end up arguing from anecdotes. Two weeks of sampling costs almost nothing and makes the rest of the argument possible.

The channel underneath is unfinished

If your organic presence is thin, being cited is not the fastest available win. Answer engines lean on sources they already trust, which usually means you have to be findable and credible in ordinary search first.

What to do with this

  1. 01Spend the first month on retrievability and the publishing path. Those two fix the ceiling on everything else.
  2. 02Start sampling now even if you buy nothing, so that whenever you do start you can prove what changed.
  3. 03Revisit the decision when you can answer yes to three things: buyers research online, pages can be fetched and quoted, and someone can publish a change this week.

Questions

Is it too late to start AEO?
No. The surface is a few years old, the engines change constantly, and the underlying inputs are the same durable ones. Late is not the risk here. Spending on visibility while the site is unreadable is the risk.
We are a small local business. Does this apply to us?
Partly. For local demand, accurate business listings, reviews and a clear, crawlable site do most of the work, and those also feed the answer engines. A dedicated AEO programme rarely makes sense before that foundation is solid.
Our traffic is falling. Should we buy AEO to fix it?
Diagnose first. Separate informational from commercial queries and compare click-through at stable positions before concluding that AI answers caused it. Plenty of declines blamed on AI turn out to be seasonality, a migration, or a page that lost its ranking for ordinary reasons.

Related

Working on this yourself and want a second opinion? Tell me what you are seeing, or read the AEO Playbook for the full framework.