Answers
Should I do AEO in-house, or hire an agency?
Short answer
Should I do AEO in-house, or hire an agency?
Whether AEO belongs in-house or with an agency depends on whether you can staff and sustain all three parts of the work, not just the visible one. AEO splits into on-site pages your own team can write, off-site placement on the third-party pages that carry 91.5 percent of citations by our own measurement, and a held-out control measurement that proves any of it caused a change. Most in-house attempts staff the first, attempt the second inconsistently, and skip the third entirely, which is why they produce activity without a provable result. An agency is worth it specifically for the second and third parts: relationships and process for off-site placement, and a pre-registered measurement design most in-house teams have no reason to have built before.
The short version
The part that fails first is proving cause, not producing pages.
01
The work has three parts, and only one is easy to staff
On-site work, writing answer capsules and structuring pages for extraction, is the part an in-house content or SEO hire can do well, because it is close to work they already do. Off-site placement, getting named accurately on the roundups and community threads that our own measurement puts at 91.5 percent of citations, is a different skill: it is closer to PR and community relationship-building than to publishing, and it does not compound the way owned content does.
The third part, a held-out control measurement, is the one almost no in-house team has built, because nothing else in a marketing function requires deliberately withholding work from part of a query set to prove causation on the rest. It is a statistics exercise wearing a marketing coat, and skipping it is the default, not the exception.
Three parts, one usually skipped
Most in-house attempts staff the first part, attempt the second inconsistently, and skip the third entirely.
02
What breaks first when this stays in-house
The most common in-house failure is not effort, it is measurement. A team publishes pages, checks a visibility tool weekly, sees the number move, and reports the move as the result, without a control arm to say whether the model moved on its own. Our own step-zero run found 7 of 12 queries changed outcome across five identical repeats with no work applied at all, which is the exact noise an unmeasured in-house effort has no way to subtract out.
The second most common failure is off-site placement stalling. Getting cited on a third-party comparison post or community thread usually requires an existing relationship or a track record the site does not have on day one, and a single in-house hire rarely owns both the content skill and the outreach network simultaneously.
Step zero, 12 queries x 5 repeats
A weekly check with no control arm has no way to tell this noise apart from a real result.
03
The five questions worth asking before hiring anyone
If you decide to hire rather than build the function in-house, the checklist that separates a real vendor from a rebrand is short: what specific AI-citation deliverable do they produce beyond generic backlinks, how do they measure whether an answer engine actually started recommending you, do they understand that a developer buyer researches inside documentation and package registries rather than review sites, will they name the queries they are targeting before signing, and what would they say if the number did not move.
A vendor with a real answer to all five has a legitimate claim on the category. A vendor who answers the first four and goes quiet on the fifth is telling you something about what happens when their own measurement comes back flat.
Ask before signing
Asked next
The questions that follow this one
Can one in-house hire cover all three parts of the work?
Rarely well. On-site writing, off-site relationship-building and control-group measurement are three different skill sets, and a single hire competent in all three is uncommon enough that most teams end up strong on one and thin on the other two.
What is the cheapest way to test whether in-house is working?
Run a held-out control on your own effort before paying anyone. Pick a set of money queries, work half of them and leave the other half untouched for the same window, then compare. If you cannot build that split, that gap is itself the answer to whether the function is properly resourced.
Does hiring an agency mean giving up on-site work?
No. The on-site half is usually the part an in-house team keeps even after hiring an agency for off-site placement and measurement, since it is the skill closest to work most content or SEO teams already own.
What is the single biggest tell that an in-house attempt has stalled?
A visibility score reported without a comparison arm. That is a symptom of measuring activity rather than proving cause, and it is the same failure mode a five-question vendor checklist is built to screen out on the agency side.
Before deciding either way
Free gap report
When the model answers,
be the one it names
Tell us your category and the questions your buyers ask. We run them against live AI answers and walk you through what came back. If you are already winning, we will tell you that too.
12 queries · 5 repeats each · 5 working days
Free · Walked through live · Five working days
Queries we run
Cited instead of you
reddit.com38%
g2.com26%
news.ycombinator.com21%91.5% of citations point off-site