Will ads appear inside AI answers?

The platforms have said they are testing advertising in assistant surfaces, and their revenue depends on it. Plan for paid placements to exist there rather than assuming a permanently ad-free answer.

Which should a new business do first?

Usually paid, because you need to learn what converts before you spend months writing pages about it. Ads are also the fastest way to discover the questions worth answering.

Can answer work reduce paid spend?

Over time, on the queries it covers, and slowly. Treat it as reducing dependence rather than as a line-item saving you can forecast.

AEO vs paid

AEO vs paid search

One is a tap you turn on and off. The other is an asset that takes months and then keeps working. They fail in opposite ways.

The short answer

Paid search buys immediate presence at a price per click that rises with competition and stops the moment you stop paying. Answer engine work is slow, unguaranteed and cumulative. If you need pipeline this quarter, paid is the honest answer. If you are only ever buying it, you are renting a position you could partly own.

What ranked on 23 September 2026: Results for this comparison are written almost entirely by agencies selling one side. The useful framing, which is cash flow and time horizon, appears in almost none of them.

How to think about the split

The decision is a cash-flow question more than a marketing one, and it changes with how long your business can wait.

Paid gives measurable attribution today; answer work gives poor attribution and better durability
Paid costs scale linearly with volume; answer work has a fixed effort cost and non-linear returns
Assistants are beginning to carry sponsored placements, which will make this comparison direct rather than theoretical
A high-consideration purchase gives answer work more room, because buyers research before they click anything
When the ad account pauses, the traffic stops the same day. A page that gets quoted keeps being quoted until something changes

Things that sound right and are not

ClaimWe will stop ads once organic works.
Rarely happens cleanly, and the plan tends to end with both underfunded. Decide the ongoing split rather than treating one as temporary.
ClaimAI search will kill paid search.
Platforms have advertising businesses to protect and are already testing placements in assistant surfaces. Expect the inventory to change shape, not to disappear.
ClaimAnswer work has no cost because we do it in-house.
It has a labour cost, and an opportunity cost in whatever that person stops doing. It is cheaper than ads at volume, not free.

Where to start

01Work out what you are paying per acquisition through paid, and what a month of the equivalent labour costs.
02If you cannot survive two quarters without new pipeline, keep paid funded first and do the cheap answer work alongside it.
03Start answer work on the queries where you already pay the most per click, because those are the expensive positions worth owning.
04Keep the brand and bottom-funnel campaigns running regardless; they are the ones assistants cannot intercept.
05Review the split quarterly against the sampled prompt set and the ad account together, not separately.

Questions

Will ads appear inside AI answers?: The platforms have said they are testing advertising in assistant surfaces, and their revenue depends on it. Plan for paid placements to exist there rather than assuming a permanently ad-free answer.
Which should a new business do first?: Usually paid, because you need to learn what converts before you spend months writing pages about it. Ads are also the fastest way to discover the questions worth answering.
Can answer work reduce paid spend?: Over time, on the queries it covers, and slowly. Treat it as reducing dependence rather than as a line-item saving you can forecast.

Doing this yourself and want a second opinion on what the assistants say about you? Tell me what you are seeing, or read the AEO Playbook for the whole framework.