Strategy & Planning · 5 min read · September 24, 2026 · Hafsa Tariq, gaash.ai
Building a GEO strategy in 30 days: seven decisions for the first month, written on one page
A GEO strategy is a short set of decisions made before the work starts: which buyers' questions it is for, which services and places to lead with, where the business stands today, who owns each part, what it will not do, whether to do it in-house, and how it will be judged. A month is enough to make them and write them on one page. The tasks follow from the page, and the 90-day roadmap lays them out week by week.
A strategy is decisions; the tasks come after
Most GEO plans start with tasks: add markup, write an FAQ, fix the listings. Tasks without decisions behind them produce busy months and no way to tell whether any of it mattered. Make the decisions first, in the first month, and write them down. Then the 90-day roadmap turns them into work, week by week, and each task has a reason on the page above it.
Decisions one and two: the questions, and the services and places
The strategy is for a set of questions, the ones your buyers ask an assistant when they need what you sell. Write ten to twelve of them in the buyers' own words, mixing three kinds: the problem a buyer has before knowing any names, a comparison, and your own name. They become the panel you measure every month, so choose them carefully and then leave them alone. Setting up prompt monitoring covers how to write them.
Then choose what to lead with. Not every service, and not every town: the work you want more of, in the places you actually serve. A strategy that tries to be named for everything at once cannot tell which page moved which question. Two or three services and the area around one location is a fine first scope for a small business.
Decision three: where you stand today
Take a baseline before anything changes: put the panel's questions to ChatGPT, Claude, Gemini and Perplexity, three times each, and record who is named. Check what a crawler receives from your site, and compare your name, address and phone across your own pages and two listings. Then place yourself on the ladder getting onto the AI shortlist describes: never mentioned, mentioned but not recommended, recommended but seldom first.
A thin baseline is not a verdict. In the probe that accompanied our audit of 9,483 small business websites, one buyer question per business went to Claude with web retrieval on, and 14.9% of 9,240 businesses were named. Most began outside the answer.
Decision four: who owns what
| Part | The owner is responsible for | Why it needs an owner |
|---|---|---|
| The facts source | The one place the name, address, phone, hours, services and prices are written, from which everything else is copied | Facts drift when several people keep several copies |
| The site's text | Service pages, answers under question headings, dates | Most of what answers use is here |
| Access and markup | robots.txt, bot settings, and the JSON-LD matching the pages | A security update can undo it overnight |
| Listings and reviews | Map listings, directories, replies to reviews | Each goes stale on its own schedule |
| The monthly reading | Running the panel the same way and writing what moved | Without it nobody knows whether the rest worked |
In a small business one person may own all five. That is fine, as long as it is written down, because the work that nobody owns is the work that stops the month a busy season starts.
Decision five: what you will not do
A short stop list saves more time than any task. Write down what the strategy rules out, so nobody spends a month on it.
- No paying for a promised place in the answers; nobody outside the assistants' makers can deliver one.
- No pages for towns you do not serve, and no pages that differ only in a place name.
- No reviews written or rewarded by the business, and no platform ratings copied into markup.
- No special file bought as the fix; Google says none is needed for its AI features, and the pages come first.
- No rewording the panel's questions after a disappointing month.
The ten most common GEO mistakes explains the reasoning behind each.
Decisions six and seven: who does it, and how it is judged
Decide whether to do the work in-house or hire it, by hours rather than by preference. The foundation, access, identity, markup and facts in text, is a few afternoons for someone who can edit the site. Answer pages are days. The monthly reading is a morning a month, and the part most often dropped. If you hire, our prices are the $490 audit, credited in full against optimization from $1,900, and monitoring at $290 a month, which you can cancel in any month; whichever you choose, the decision belongs on the page with its cost.
Then decide how the strategy will be judged, before the first result arrives. Set a direction rather than a figure, such as being named on the problem questions for the lead service by more of the four assistants. Track the numbers the seven GEO KPIs sets out, read them monthly, and review the strategy itself each quarter.
The one-page strategy
Everything above fits on one page. Keep it where the people who do the work can see it, and change it only at the quarterly review.
GEO strategy, [business name], written [date]
Questions (fixed panel): [ten to twelve, in buyers' words]
Lead services and places: [two or three services] in [area]
Baseline, taken [date]: named on [n] of [n] questions; rung: [mentioned / recommended / first]
Owners: facts [name]; site [name]; access and markup [name];
listings and reviews [name]; monthly reading [name]
We will not: [the stop list]
Who does the work: [in-house hours a month / hired, at what cost]
Judged by: [the direction]; KPIs read monthly on [day]
Next review: [date, one quarter on]Common questions
Can a GEO strategy really be built in 30 days?
The decisions can, and they should be, because they shape everything after. The results take as long as they take: nobody outside the assistants' makers can promise when, or whether, an answer changes. A month of decisions, followed by work and a monthly reading, is how you find out.
Does a small business need a written strategy?
It needs the decisions, and writing them on one page is the cheapest way to keep them. Without it, the panel's questions drift, nobody owns the listings, and a year later there is no way to say what was tried. One page is enough.
What if the baseline shows we are never named?
That is the common starting point, not a verdict. In our audit's probe, most businesses were not named for a plain question about their own trade in their own town. Start with the rungs that come first, access and one identity, and read the panel again after a month.
How often should the strategy change?
Review it each quarter, with three monthly readings in hand. Change the lead services, the owners or the stop list when the readings or the business give a reason. Leave the panel's existing questions alone; add new ones as new lines, so the series stays comparable.
Sources
The probe figure is this site's own, from the article named below; it was one question per business on Claude. Prices are this site's own, from the audit and pricing pages.
Sources: our audit of 9,483 small-business sites, with its method · the 90-day roadmap · how the visibility score is measured
Read on