Comparison

GEO Agency vs In-House vs Done-for-You: Costs, Timelines, Control

An in-house GEO team gives the most control and the slowest start; an agency retainer gives breadth across SEO, content and PR with GEO as one workstream; a done-for-you service gives a single measured outcome (brand named in ChatGPT for a defined cluster) with the least management overhead. Choose by speed needed and who owns the result.

7 min readPublished October 8, 2026Published by Odys Global

The right choice depends on three things: how quickly you need ChatGPT to start naming you, how much control you want over the work, and whether anyone inside the company can own it. In-house maximizes control and minimizes speed. An agency trades some control for breadth. A done-for-you service trades breadth for a single measured outcome with minimal management on your side.

The direct answer: match the model to who owns the result

Ask who, twelve months from now, is accountable if a competitor is named in 70% of ChatGPT answers for your core prompts and you are named in 10%.

If the answer is “our head of marketing, and she has the time”, build in-house. If the answer is “our agency, as part of everything else they do for us”, a retainer with a GEO workstream fits. If the answer is “whoever we hire for exactly this, measured on exactly this”, a done-for-you service fits.

The reason to decide by ownership rather than price is that generative engine optimization fails most often through neglect, not through wrong tactics. A dashboard that nobody reads and a content calendar that never mentions ChatGPT are the typical outcomes when the work is nobody’s job. Pick the model where it is clearly somebody’s job.

The three models compared

Dimension In-house team GEO or SEO agency retainer Done-for-you service
What you buy Salaries, tools, training Hours and deliverables across SEO, content, PR, with GEO as a workstream A defined outcome for a defined prompt cluster and market
Cost drivers Headcount, tool subscriptions, earned-media budget Retainer size, scope of disciplines, markets, reporting depth Number of clusters, markets, starting gap to competitors
Time to start A quarter or more to hire, train and build a method Weeks to onboard Days to baseline
Time to first movement Months One to three months, category dependent Targets first movement within the first week, consistent mentions across most of the cluster within one to three months, and a stable top-three position by month three to four
Control over the work Full Shared; you approve deliverables Low; you approve the cluster and read the re-audit
Measurement Whatever you build or buy Agency reporting, often rank-tracker style About 100 answers per prompt per market, re-run identically monthly
Accountability Internal To the scope of work To the mention rate and position of the cluster
Knowledge retained All of it stays in the company Some, through reporting and handovers Little about the method; all of the measurement data
Biggest risk Side-tasking it to someone already busy GEO becoming a line item inside SEO reporting Choosing a provider who will not show the sampling method

The timeline rows carry a caveat that applies to every model: nobody controls ChatGPT. Cited sources rotate heavily, with 40 to 60% changing monthly according to the Digital Authority Partners study, so any provider promising a fixed outcome by a fixed date is overselling.

When in-house wins

In-house wins when GEO is strategic enough to deserve a permanent owner and the company already has the adjacent skills.

That usually means a company with a content team, a PR function or agency, and a marketer who understands entity consistency and prompt clusters. For them, GEO is a new objective for existing people rather than a new discipline, and the knowledge compounds inside the company.

In-house also wins on control and on sensitive categories. A law firm, a healthcare group or a financial services firm may need every public statement about the brand reviewed by compliance. An internal owner who can walk down the hall beats any external party on turnaround there.

The costs are the obvious ones (salary, tools, an earned-media budget) and the less obvious one: time to competence. Expect a quarter before the method is stable and a second quarter before you can trust the trend lines. If you go this route, start with the DIY audit method so the first baseline exists before any hire is made.

In-house wins when

You have content and PR capacity already, the category is heavily regulated, or GEO will be a multi-year priority where retained knowledge matters more than speed.

When an agency wins

An agency wins when GEO is one of several things you need and you want them coordinated.

Most companies do not need a GEO specialist; they need their SEO, content and digital PR to start producing the signals AI engines reward: consistent facts about the brand, coverage in the roundups and review sites buyers compare on, recent content, and structured facts on the site. A good agency already does most of this and can add the GEO lens (which prompts, which sources, which competitors) without a new vendor. The overlap is covered in GEO vs SEO.

The risk is dilution. GEO measured in rankings and clicks looks like it is failing even when it is working, because nearly half of consumers Google the brand after an AI names it rather than clicking through (45% in a 2026 Idea Grove survey). If the agency reports GEO inside the SEO dashboard, the work gets deprioritized in the first quarter it does not show up as traffic. Insist on separate metrics: mention rate, position, and branded search lift.

Cost depends on retainer size and scope. The GEO component rarely has its own price; it is folded into content and PR hours. That makes it hard to compare with a done-for-you quote, so ask the agency to isolate the hours they would spend on GEO specifically.

An agency wins when

You already have a retainer you trust, you want GEO coordinated with SEO and PR rather than run separately, and you can live with movement over one to three months.

When done-for-you wins

Done-for-you wins when you want the outcome, you want it measured properly, and you do not want to manage the work.

The model is narrow by design: a prompt cluster (say 30 ways buyers ask for your category), a market, a baseline, work to move the numbers, and a monthly re-audit on the identical method. Our version collects about 100 real ChatGPT answers per prompt per market from genuine mobile connections in the target country and classifies each one for brand named, position, link to your domain and competitors named. It targets first movement within the first week, consistent mentions toward up to 90% of prompts in the cluster, and a top-three position. Those are targets, never guarantees; how we do it is explained on your call, and the method’s limits are set out on the methodology page.

What you give up is breadth and visibility into the process. A done-for-you provider is not going to rewrite your blog or run your PR. It is accountable for one number on one cluster, and you judge it by whether that number moves and whether you can reproduce it yourself in ChatGPT. No lock-in matters here: the re-audit, not the contract, should be what keeps the engagement alive.

What to check before choosing any provider in this category:

  • The written sampling method, including sample size, location, device and login state.
  • Prompt-level reporting you can reproduce.
  • A re-audit on the identical method, not a reworked one.
  • A clear statement that nothing involves buying mentions, fake reviews or paid placement dressed as coverage. Buying mentions vs earning the recommendation explains why those underperform.

Done-for-you wins when

A competitor already owns your core prompts, you need movement in weeks rather than quarters, nobody in-house can own it, or you cannot run ads and need this channel to work on its own.

When to combine them

The common mature setup is two of the three.

Done-for-you plus in-house owner. A marketer owns the dashboard and the branded-search landing page while the provider runs the priority clusters. The marketer learns what moves from the re-audits and can bring more of it in-house over time.

Agency plus done-for-you. The agency keeps SEO, content and PR running and adopts the provider’s cluster as a shared target. The provider’s re-audit becomes an independent check on whether the agency’s coverage work is reaching the sources ChatGPT uses.

In-house plus a one-off audit. A company building in-house buys a single baseline audit to start from a measured position instead of a guess, then runs its own program and re-measures with its own tools.

Which to fund first: a decision rule

Ask three questions in order.

Is a competitor already winning your prompts? Run ten buyer prompts in ChatGPT, or check what ChatGPT currently says about your brand across a measured cluster. If a competitor is named in most answers and you are not, speed matters, which points away from in-house as the first move.

Who can own it weekly? If a named person has four hours a week and authority over content and PR, in-house or agency can work. If nobody does, done-for-you is the only model that does not depend on your calendar.

How is it going to be measured? If the answer is “in the SEO dashboard”, you will underfund it within two quarters. Whichever model you pick, give GEO its own metrics first; how to measure your brand’s visibility in ChatGPT explains what they should be.

The rule compressed: if speed and a clean outcome matter most, start done-for-you and build in-house knowledge from the re-audits; if coordination matters most, fold it into the agency with separate metrics; build in-house only when GEO is a permanent priority with permanent people.

What to do next

  • Write the name of the person accountable for ChatGPT visibility in twelve months. If there is no name, you have your answer.
  • Price a single cluster in a single market with each option you are considering, so the quotes are comparable.
  • Read how long it takes to get recommended by ChatGPT so your timeline expectations match the model you choose.

Frequently asked questions

What does a GEO agency actually do?

A GEO agency typically audits how AI engines describe your brand, builds prompt clusters, fixes entity consistency across your site and third-party profiles, produces content designed to be retrieved and quoted, and runs digital PR to earn coverage in the sources AI engines cite. Most offer it as a workstream inside a broader SEO or content retainer rather than as a standalone, outcome-priced service.

How much does GEO cost?

It depends on scope more than on the vendor type. The cost drivers are the number of prompt clusters and markets, how far behind competitors you start, how much earned media the category requires, the depth and frequency of measurement, and whether you pay for hours or for an outcome. Ask any provider to price a single cluster in a single market first so you can compare like with like.

Can one in-house marketer handle GEO?

Yes, for a small number of clusters, if they are given time, a measurement method and authority to change content and pitch press. The common failure is assigning GEO as a side task to someone already running SEO and paid, with no sampling method and no budget for earned media. Then the dashboard fills with numbers nobody acts on.

Is a done-for-you GEO service the same as an agency?

Not quite. An agency sells hours and deliverables across several disciplines. A done-for-you service sells a defined outcome for a defined scope: a baseline audit of a prompt cluster, work to move the mention rate and position, and a monthly re-audit on the identical method. The accountability is to the number, and the client is not expected to manage the work.

How long before I see results from each option?

In-house usually takes a quarter to hire or train, build a method and ship the first changes. Agencies typically start producing within the first month and show movement over one to three months, depending on the category. A done-for-you service built around a cluster targets first movement within the first week, consistent mentions across most of the cluster within one to three months, and a stable top-three position by month three to four, though starting position and category always matter.

Your next move

Find out what ChatGPT says about your brand before you spend anything.

Apply in about two minutes. We check your keywords, then run your real buyer prompts live on a 30-minute call. If we cannot create the result for your keywords, we tell you on the call, not after an invoice.

See if my keywords qualify →
2-minute application30-minute live audit callWe decline keywords we cannot move

Playbooks by industry

Check my keywords →