Read time: 2 min

GEO agency or in-house: an honest decision guide

Daniel RothLast updated on Aug 14, 2026Talk to our team →

Key highlights

  • The real cost is not the software published GEO tools run $99 to $780 a month, and the binding constraint is the in-house hours that act on what they report.
  • Published agency link and retainer rates run from $999 to $15,000 a month, and the wide range mostly reflects what is bundled and what is disclosed.
  • In-house wins when someone senior owns it the work is cross-functional, touching engineering, content and PR, and orphaned dashboards achieve nothing.
  • The decision is reversible a sensible path is measuring in-house first, then deciding with a baseline in hand rather than a pitch deck.

In this article

Summarise this article with AIOpen ChatGPT →

Disclosure first: we sell the agency side of this choice, so read this the way you would read a barber's opinion on haircuts. We have tried to make it the framework we would want as a buyer, with published numbers where they exist.

The question is rarely whether AI visibility matters. It is whose hours will do the work, and what those hours cost in practice.

The actual cost stack

Tools: published pricing for AI visibility trackers runs from around $20 to $99 at entry, through $245 to $495 mid-tier, to $780 and up for larger plans [1]. Tools measure; none of them do the work.

Agencies: our programme starts at $999 a month, published. Elsewhere, published retainers in adjacent link building run $2,999 to $15,000 a month [2], and many GEO agencies publish no pricing at all, which is itself information.

In-house: the invisible line item. A competent owner spending even a quarter of their time is often the largest cost in the stack, and the one nobody budgets.

When in-house is genuinely the right call

You have a senior person who wants to own it, with authority across engineering, content and outreach, because the work crosses all three. You already publish well, so the gap is measurement and the third-party layer rather than writing capacity. And your category is technical enough that domain knowledge beats production speed.

In that setup, buy a tracker, fix legibility, ship question-shaped pages, and work the cited sources. Everything in the public playbooks, including ours, is executable internally by a team that will actually execute it.

When an agency earns its fee

Nobody in-house has weekly hours for it, which is the most common reality: the dashboard gets opened in month one and orphaned by month three, and a tracker nobody acts on is a subscription, not a channel. Or the work needed is mostly the grind, hundreds of outreach touches to find the few dozen worth having, where an agency's existing pipeline is the product. Or you need the measurement itself to be credible to a board, which is where verification by live API call, with published methods, matters.

The test to run in any agency pitch: ask what arrives each month, ask to see verbatim answers for a real client, and ask how their numbers were measured. The answers separate operators from invoice generators quickly.

The reversible path

This is not a marriage. The cheapest sensible sequence: measure your baseline first, in-house, with a fixed question set and a free or entry-tier tool. If the gap is small and specific, close it internally. If it is structural, you now have the exact evidence to brief an agency against, and to hold it accountable with afterwards.

Either way, insist that whoever does the work reports against the same fixed questions every month. The owner can change. The denominator should not.

Be the answer AI recommends.

We find the buyer questions where an engine names someone else, then publish and earn the citations that change the answer.

Book a call →