In-house GEO vs. hiring an agency
What it actually takes to run generative engine optimization yourself, what an agency adds, and how to decide which one fits your team right now.
In-house GEO works when you already have a content team, engineering support to fix crawler and schema issues, and someone with real time each week to track citations across ChatGPT, Claude, Gemini and Perplexity. Hiring an agency works when you need all four of those skills now, on one invoice, without building a new department.
This is one of four questions how to choose a GEO agency walks through for anyone comparing vendors; here we compare the vendor option against not hiring one at all.
On this page
What does "doing GEO in-house" actually involve week to week?
Four separate jobs, done by whoever is available. Someone checks that GPTBot, ClaudeBot and PerplexityBot can still reach your pages, because a firewall update or a new WAF rule can block them silently [1][4]. Someone rewrites pages so a model can lift a clean, complete answer from them. Someone reaches out to the review sites, directories and comparison posts that answer engines already trust. And someone runs the same set of buyer questions against each engine every week and logs what came back.
None of those four is hard on its own. The difficulty is that they rarely sit with one job title. The first needs an engineer. The second needs a writer who understands the format. The third is closer to partnerships or PR. The fourth needs someone disciplined enough to run the same prompts weekly instead of checking in sporadically and drawing conclusions from a single lucky answer.
Most companies already have people who can do pieces of this. The gap is rarely skill; it is time. A content writer who already owns three other projects will not become the person who quietly pitches twenty review sites a month, and an engineer with a backlog will not treat a robots.txt check as urgent unless someone keeps asking. In-house GEO fails less often because nobody knew how, and more often because nobody owned the whole list.
What does the first 90 days actually look like if you build this yourself?
Slower than most teams expect, mostly because the first month is setup rather than output. Weeks one and two are spent confirming crawler access, picking a fixed set of twenty to forty buyer questions, and deciding where those questions will be run and logged every week. Weeks three and four are the first real rewrite pass on a handful of pages. Only in month two does outreach to third-party sites usually start, because it takes that long to identify which pages already get cited in your category and worth pitching.
By day 90 a well-run in-house build normally has a working weekly tracking habit, a handful of rewritten pages, and two or three early conversations with third-party sites — not yet placements, just conversations. That pace is not a failure. It is what building a new function from nothing actually looks like, and it is worth knowing in advance so month two does not feel like the project stalled. An agency starting the same day usually has the crawler check and the prompt set running in week one, simply because both are already built processes rather than new ones.
What does an agency actually add that a good hire wouldn't?
Two things: existing relationships, and a system already built. A GEO agency working several accounts has already pitched inclusion in comparison posts and review platforms in adjacent categories, so it starts partnerships work with warm contacts instead of a cold list. It also arrives with the tracking spreadsheet or software, the prompt-set method, and the reporting cadence already tested on other clients, instead of building that process from nothing in month one.
What it does not have is a magic lever an employee could never pull. The mechanics — crawlability, page structure, third-party outreach, weekly measurement — are the same four jobs either way. You are paying for the relationships and the running start, not for a different kind of work.
How do the costs actually compare?
Rarely the way a single line item suggests. A retainer looks like the bigger number until you price out what an in-house build actually needs.
| In-house | Agency |
|---|---|
| One salaried hire, plus benefits | One monthly retainer, fixed scope |
| Borrowed engineering time for crawler fixes | Engineering fixes scoped into the retainer |
| Borrowed writer time for page rewrites | Writing done by the agency's own team |
| No existing third-party relationships | Relationships built across other clients |
| Tooling and process built from scratch | Tooling and process already running |
The honest comparison is not salary against invoice. It is one full-time cost plus the hours other teams spend supporting that hire, against one fixed monthly fee that already includes the support functions.
There is one more cost that rarely makes it into either column: the time it takes to notice the work has stalled. An underperforming in-house hire can go unnoticed for months inside a busy team, because nobody else is checking the weekly tracking sheet either. An underperforming agency shows up in a monthly report you actually read, and a contract you can end. Neither structure removes the risk of the work quietly not happening; a retainer just makes it visible faster.
How fast can each option actually start working?
On-page fixes move at the same speed either way, because the work is not different — a page that already ranks can usually be rewritten and reindexed within weeks, in-house or through a vendor [1]. Third-party placement is the part that changes with experience. A team pitching its first comparison post is starting a relationship from zero; an agency with existing contacts across a category can sometimes get a response in days rather than weeks.
Tracking speed depends entirely on whether the habit sticks. A study that tracked 13,184 citations across the four major assistants found week-to-week churn strong enough that a single run tells you almost nothing — Claude kept only 30% of the URLs it had cited two days earlier in one test run [3]. Whoever owns this needs the discipline to keep running the same questions for a month before drawing a conclusion, which is exactly the kind of unglamorous, recurring task that in-house teams tend to let slip once a busier priority shows up.
What skills does an in-house GEO hire actually need?
Technical enough to read a robots.txt and a firewall rule
They do not need to write the fix, but they need to know what to ask an engineer for and how to verify it worked.
A writer who understands extractable prose
Short, complete answers under a clear heading, not long pages that circle a topic — this is a specific skill, not general content writing.
Comfortable asking strangers for a mention
Getting onto review sites and comparison posts is outreach, closer to PR than to SEO, and most content hires have never done it.
Disciplined enough to track weekly without being told
The measurement only means anything if someone runs the same prompt set every week for a month, on a fixed schedule, whether or not anything interesting happened that week.
When does hiring in-house clearly make sense?
When you already have three of the four skills above and are only missing one. A content team that already writes well and has an engineer on call mostly needs someone to run the tracking discipline and start the outreach — a narrower hire than building the whole function from nothing. It also makes sense when GEO is core enough to your business that you want the institutional knowledge to stay inside the company rather than live in a vendor's head.
It also tends to work better for companies with a genuinely unusual product, where an outside writer would spend a month just learning the domain well enough to write a clean, extractable answer. In that case, the ramp-up cost of an agency learning your business can exceed the ramp-up cost of a hire who already knows it and only needs to learn the format.
When does an agency clearly make more sense?
When you are missing more than one of the four skills, or when the team you have is already stretched on other work. It also makes sense simply to get started faster: a vendor with a working process can begin week one, where an in-house build usually spends its first month hiring, before a single citation is tracked. If your current SEO agency vs. in-house vs. freelancer decision already leans toward outside help for search, the same reasoning usually applies here.
There is also a scale argument. An agency running several accounts sees far more of the pattern than any single in-house hire will see in a year — which third-party pages a category's engines actually trust, which pitches get a reply, which page formats keep getting cited. A new in-house hire has to learn all of that from scratch, one account at a time, which is slower work than it looks from the outside.
Is there a middle path?
Yes, and it is common. Some teams keep the writing and the crawler checks in-house, since both are close to work they already do, and bring in an agency only for third-party outreach and the weekly cross-engine tracking — the two jobs furthest from a typical content team's normal work. If you would rather compare software against a full-service vendor instead of a hybrid team, GEO tools vs. GEO agency is the next comparison to read, and our red flags guide is worth a look before you sign anything either way.
The question I ask founders isn't in-house or agency. It's which of the four jobs you'd actually keep doing in month four, after the novelty wears off. Whichever ones you wouldn't, pay someone else to own.
Either path is a real option, not a consolation prize for the one you did not pick. What decides it is capacity, not ambition: how many of the four jobs above your current team can actually keep doing every single week, past the first month when it stops being new.
Questions people also ask
Can one in-house person run GEO alone?
One person can run the tracking and the on-page work. They usually cannot also fix crawler and firewall issues, which needs engineering time, or land third-party placements at volume, which is closer to a PR or partnerships job. Most one-person setups quietly become a part-time job for two other people too.
Is an in-house hire cheaper than an agency?
Only if you already employ the adjacent skills. A single full-time hire plus the engineering and content time they need to borrow usually costs more than a retainer once you count salary, benefits and the hours other teams spend supporting them.
What's the fastest way to start without hiring anyone?
Fix crawler access this week — it costs nothing and needs no new headcount — then decide whether the ongoing work belongs in-house or with a vendor. Our guide to what a GEO agency does lays out the four jobs so you can see which ones you already have covered.
Do agencies use the same tools an in-house team would buy?
Mostly yes. The tools are commodity software; what an agency adds on top is the people running them every week and the relationships that get you onto third-party pages. Our GEO tools vs. agency comparison goes through that split in detail.
Can you start in-house and switch to an agency later?
Yes, and it is a common path. Build the tracking habit in-house first, see how many hours it actually eats, then decide whether to keep it or hand it off. Nothing about switching later locks you into anything you built.
Sources
- AI features and your website — Google Search Central (checked 22 September 2026)
- Overview of OpenAI crawlers — OpenAI (checked 22 September 2026)
- I tracked 13,184 AI citations across 4 LLMs: here's what I learned — Orbit Media Studios (checked 22 September 2026)
- Does Anthropic crawl data from the web, and how can site owners block the crawler? — Anthropic (checked 22 September 2026)
Not sure which mix of in-house and agency fits your team?
Book a free 30-minute consultation with Ibrahem and Jeremy. We look at your site while we talk, and you leave with a prioritised plan.
Who wrote this
Ibrahem HelalCo-founder
Leads strategy and delivery.
Works on: SEO, GEO
Reviewed before publishing by Jeremy Zava, Co-founder.
- GEO Agency vs SEO Agency: What's the Difference?How a GEO agency's work differs from an SEO agency's day to day, where the two overlap completely, and how to decide which one you actually need.
- GEO tools vs. a GEO agencyAI visibility software tracks your citations across ChatGPT, Claude, Gemini and Perplexity. It won't rewrite a page or land a placement. Here's the real split.
- What Does a GEO Agency Actually Do?What a GEO agency delivers each month: crawler checks, page rewrites for AI citation, tracking against real questions, and third-party outreach.
- GEO Agency Red Flags to Watch ForEight warning signs in a GEO proposal or sales call, from guaranteed ChatGPT rankings to reports with no URLs, and what to ask instead.