A geo checklist for small businesses works best as a 30-day starter sequence, not a theory lesson. Week one makes your business legible: claim your Google Business Profile, fix your name, address, and phone number everywhere, and write plain service summaries. Week two publishes the answers and prices customers ask about. Week three earns corroboration \u2014 reviews, directories, and mentions that confirm you are real. Week four tests what AI tools say about you and fixes the gaps. Do those four weeks in order, from zero, and you have the foundations answer engines cite.<\/p>\n\n\n\n
Two earlier guides cover the neighbouring ground, and this plan does not repeat them. For the concept itself, start with generative engine optimization<\/a>. If your foundations are already in place, read how to get ChatGPT to cite your business<\/a>. This piece assumes neither: you are starting from zero, and it answers a narrower question \u2014 what do you do on day one, day eight, day fifteen, and day twenty-two? The output is a calendar, not a definition. One honest warning: thirty days builds the foundations; it does not guarantee citations by day thirty-one, because engines re-read their sources on their own schedule.<\/p>\n\n\n\n Each week has one job, a short task list, and a definition of done. Resist reordering: answer pages in week two work better once your facts agree in week one, and review asks in week three land better once your pages say the right things. Budget two to four focused hours a week. If a task is already done, verify it and move on.<\/p>\n\n\n\n Day 1\u20132: Claim and complete your Google Business Profile.<\/strong> Correct category, real service list, accurate hours, current photos, and a service area that matches where you actually work. Google’s own documentation on AI features and your website<\/a> tells local businesses to keep that profile information up to date for exactly this reason.<\/p>\n\n\n\n Day 3\u20134: Fix your facts in one sitting.<\/strong> Write down one canonical version of your business name, phone number, service list, service area, and hours. Check it against your website, profile, and main directories, and correct every variation \u2014 the old phone number, the former address, the dropped service. Contradictions are the most common starter problem, and the cheapest to fix.<\/p>\n\n\n\n Day 5\u20137: Rewrite your service summaries in plain sentences.<\/strong> Open each core service page with a paragraph an AI could quote word for word: what the service is, who it is for, where you offer it. “We renovate kitchens for homeowners in Toronto” is citable; “Solutions for your dream space” contains no fact at all. End the week by checking your robots.txt does not block AI crawlers \u2014 a blocked crawler makes everything after this pointless.<\/p>\n\n\n\n Done when:<\/strong> your profile is complete, your facts match in the places you checked, and every core service page opens with a plain, factual summary.<\/p>\n\n\n\n Week two gives the machines something to quote: the answers your customers already ask you by phone and email.<\/p>\n\n\n\n Publish real prices, or honest ranges.<\/strong> The cost question is usually the most-asked and least-answered in any trade. You do not need to price every job to the dollar; a starting price or clear range, with what moves it, is quotable \u2014 “contact us for pricing” is not. Our SEO cost guide<\/a> is built to this pattern: short answer first, range, then what changes the number.<\/p>\n\n\n\n Answer the next five questions.<\/strong> After price, customers ask how long it takes, what the process looks like, what could go wrong, how to choose between two options, and what you need from them. Answer each on the relevant service page, or as a short post of its own \u2014 direct answer first, specifics after. Add a short FAQ section to each main service page while you are there.<\/p>\n\n\n\n Connect the pages.<\/strong> Link each new answer to the service and pricing pages it supports. A page nothing links to is harder for crawlers to find and weigh.<\/p>\n\n\n\n Done when:<\/strong> your prices are public, your five most-asked questions have written answers on your site, and those answers link to your core pages.<\/p>\n\n\n\n Your own site is a claim. Week three collects the evidence that lets an AI repeat it with confidence.<\/p>\n\n\n\n Start the review habit.<\/strong> Ask every happy customer, at the moment the job finishes, and reply to every review you receive. Specific, recent reviews \u2014 naming the service, the neighbourhood, the outcome \u2014 are some of the most quotable material that exists about your business. Never buy or filter reviews; the profile they put at risk is the foundation weeks one and two stand on.<\/p>\n\n\n\n List where your trade is listed.<\/strong> Chamber of commerce, trade association, supplier and brand locators, the directories customers in your field actually use. A handful of accurate listings beats a hundred scraped ones. Use the canonical facts from week one, word for word.<\/p>\n\n\n\n Add the machine-readable extras.<\/strong> Structured data \u2014 Organization and FAQ schema \u2014 helps machines confirm your details instead of guessing. An llms.txt file, a short plain-text summary of your business for AI systems, costs an hour at most; our guide to what is llms txt<\/a> explains what goes in it. OpenAI’s publisher and developer FAQ<\/a> is the straight source on how its crawlers read sites and how owners control that access.<\/p>\n\n\n\n Done when:<\/strong> the review ask is part of your job routine, your main listings carry identical facts, and the schema and llms.txt basics are in place.<\/p>\n\n\n\n Run the citation test.<\/strong> In fresh sessions, ask the AI tools your customers use: “Who should I hire for [your service] in [your city]?” and “How much does [your service] cost?” Note whether you are named, who is named instead, and which sources the answers draw on. Absence is diagnostic \u2014 the sources named instead show you what to build next.<\/p>\n\n\n\n Check indexing.<\/strong> In Google Search Console, confirm your new and edited pages are indexed. A page Google has not indexed cannot appear in its AI answers at all.<\/p>\n\n\n\n Fix the top three gaps only.<\/strong> Pick the three highest-value fixes the tests exposed and schedule them. Then set the standing rhythm: a monthly citation test, a quarterly facts audit, and new answer pages as customers ask new questions. If you would rather hand the sequence over, this four-week build is essentially how our GEO work is scoped \u2014 a one-time GEO Setup at $349 CAD plus HST, and an optional GEO Visibility Plan at $149 CAD per month for the ongoing loop \u2014 with the full breakdown on our services<\/a> and pricing<\/a> pages.<\/p>\n\n\n\n Done when:<\/strong> you have a written baseline, your key pages are confirmed indexed, and your next three fixes are scheduled.<\/p>\n\n\n\n A starter month is as much about refusal as action. Skip AI-visibility dashboard subscriptions \u2014 you have no baseline yet, so their charts measure nothing. Skip volume blogging: twenty thin posts will not outrun the five real answers from week two. Skip doorway location pages \u2014 one template per suburb \u2014 which engines now read as thin content. And skip guaranteed citations: there is no submission form and no paid placement inside an AI answer that a small business can buy. The verification these systems do is exactly what the four weeks above build.<\/p>\n\n\n\n Partly. Week one’s profile work and week three’s listings do not require a site \u2014 do those now. But weeks two and four depend on pages you control, and a profile alone gets passed over for detailed questions. If you have no website, that build comes first.<\/p>\n\n<\/div>\n<\/div>\n Roughly two to four focused hours a week \u2014 less where a task turns out to be verification rather than new work. The heaviest sitting is week two, because real prices and real answers cannot be delegated to a template. Nothing requires daily attention.<\/p>\n\n<\/div>\n<\/div>\n Foundations, not applause. Your facts agree, your prices and top answers are published, your review habit is running, and you hold a written baseline of what AI tools say about your market. Citations follow as the systems re-read their sources \u2014 over the following months, on their schedule. Judge the plan at day thirty by whether the foundations are verifiably in place; judge results quarterly after that.<\/p>\n\n<\/div>\n<\/div>\n No. The 30 days are a build, not a subscription. What continues is week four’s light rhythm: a monthly citation test, a quarterly consistency check, and new answer pages when customers ask new questions.<\/p>\n\n<\/div>\n<\/div>\n Then you are both legible, and the contest moves to depth: whose answers are more complete, whose reviews are more specific and recent, whose facts stay consistent longest. Starting now still matters, because corroboration compounds and a later start has more to catch up.<\/p>\n\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\nHow to use this geo checklist for small businesses, week by week<\/h2>\n\n\n\n
Week 1 (days 1\u20137): Make your business legible<\/h2>\n\n\n\n
Week 2 (days 8\u201314): Publish the answers and the prices<\/h2>\n\n\n\n
Week 3 (days 15\u201321): Earn corroboration<\/h2>\n\n\n\n
Week 4 (days 22\u201330): Test what the machines say, then fix<\/h2>\n\n\n\n
What to skip in the first 30 days<\/h2>\n\n\n\n
Frequently asked questions<\/h2>\n\n\n
I have no website yet. Can I still start this plan?<\/h3>\n
How much time does the 30-day plan actually take?<\/h3>\n
What should I expect to see by day 30?<\/h3>\n
Do I need to redo this every month?<\/h3>\n
What if my competitor finishes the same 30 days?<\/h3>\n
Start week one this week<\/h2>\n\n\n\n