AEO for Small Business: 7 Free Tactics to Get Cited by AI

Type “answer engine optimization tool” into Google and you get ten paid products in a row. Writesonic, Profound, Surfer AEO, Meltwater, the new HubSpot AEO at $50 a month, and a handful of $200-plus enterprise options. Click any of them and the pitch is identical: “AI is the new search, you need a tool to track it.” Skip the next page of results and there’s almost nothing for the actual reader who isn’t ready to buy software — a small business owner trying to figure out what to change on their site this week.

This post is for that reader. Seven free tactics, no tools required, ranked in order of return-per-hour-invested. Each one is something you can actually ship between coffee and lunch.

The honest setup: AEO results take weeks to show up. The before/after numbers below are from real practitioners who shipped these tactics in March and April and are seeing citations now. If you read this and ship nothing, nothing happens. If you ship the first three this week, by July you’ll be in the answers your customers are already asking ChatGPT, Gemini, and Perplexity.

What “AEO” actually means (in 60 words)

Answer Engine Optimization is the practice of structuring your website so that AI assistants — ChatGPT, Google’s AI Overviews, Gemini, Perplexity, Claude — cite you when they answer questions in your space. It’s not a replacement for SEO. It’s a layer on top: same content, additional structure, different surface where you show up. AI engines crawl the same pages Google does. They just read them differently.

That difference is the whole game.

What AI engines actually weigh when picking citations

Before the tactics, a sentence of context. The engines all roughly weigh the same five things:

Answer clarity — does the page directly answer the question in the first paragraph?
Structure — schema markup, headings as questions, scannable lists
Authority — domain age, backlinks, citations from other authoritative sites
Freshness — Perplexity especially weights recently-updated pages
Topical specificity — narrow pages on tight topics beat broad pages on many

The order matters. A scrappy SMB site with a 200-word page that nails answer clarity and structure can outrank an authority site that buries the answer at the bottom of a 2,000-word post. That is the hole the tactics below fit through.

Tactic 1: Answer the question in your first 40–60 words

This is the single highest-leverage change you can make and it’s almost free of effort. Take any page on your site that targets a question (any page with “what is,” “how do I,” “best way to,” or any specific question phrase in its title or first heading). Rewrite the first paragraph so it answers the question in 40–60 words. Plain English. No setup. No “in this post we’ll cover.”

Before:

“Welcome to our guide on choosing a small business CRM. CRM software has come a long way in the last decade, and there are now many options on the market. In this article, we’ll explore the various factors you should consider when making your choice…”

After:

“The best CRM for a 5-person consultancy in 2026 is HubSpot’s free tier — it covers contact management, deal pipeline, and email tracking with no per-user fees up to 1 million contacts. For sales-led teams over 10 people, Pipedrive’s Essential plan at $14/user/month is the next upgrade. Service-led teams should look at Zoho One. Detail below.”

The “after” version gives ChatGPT, Perplexity, and Google AI Overviews exactly what they need to quote you. The “before” version gives them a paragraph to skip.

Why this works: AI engines extract the cleanest answer they can find. If your page leads with the answer, you get extracted. If it leads with throat-clearing, the engine extracts a different site that doesn’t.

Time to ship: 20 minutes per page. Start with your five highest-traffic pages.

Expected lift: real before/after data from practitioners showed pages restructured this way picking up Claude and ChatGPT citations within days — no other changes. One indie SaaS founder reported every paying customer in his first 12 days came from ChatGPT, not Google, after switching to this format.

Tactic 2: Add FAQ schema (the highest-leverage 30 minutes you’ll spend)

Schema.org’s FAQPage type — the documentation for the structured data that AI engines weight most heavily Source: Schema.org — FAQPage

FAQ schema (the FAQPage type from schema.org) is the single piece of structured data with the highest citation rate in AI-generated answers. Pages with FAQ schema appear in ChatGPT, Perplexity, and Google AI Overviews significantly more than equivalent pages without it. The format is JSON-LD inside a <script> tag in your <head>. You write it once per page and you’re done.

The minimum viable FAQ schema:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "How much does a 1-person bookkeeping firm cost per month?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A 1-person bookkeeping firm typically charges between $300 and $800 per month for small businesses with under $1M in annual revenue. Higher-volume or industry-specific work (e-commerce, construction) ranges $800 to $2,000."
      }
    },
    {
      "@type": "Question",
      "name": "When should I switch from a freelance bookkeeper to an in-house hire?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Most businesses switch when bookkeeping reaches 30+ hours per month or annual revenue clears $3M — whichever comes first."
      }
    }
  ]
}
</script>

That’s the whole pattern. Repeat per page with 3–5 questions each. The questions should be the actual questions your customers ask — not the ones you wish they asked.

How to validate it: paste the page’s URL into Google’s Rich Results Test. If the test passes, you’re done.

Google’s free Rich Results Test — paste a URL, see whether your schema validates Source: Google — Rich Results Test

Where to put the questions: answer them in your page body too. Schema without matching body content is treated as spam by Google and is mostly ignored by the AI engines. The pattern is question in an H2, answer in 40–80 words below it, same question and answer mirrored in the schema.

Time to ship: 30 minutes per page including validation.

Expected lift: practitioners report citation-rate uplifts in the multiple-fold range for pages that previously had no structured data. One agency-side case study on a local dentist’s site moved from “never cited by ChatGPT” to cited in 73% of "dentist in [city]" AI responses after adding FAQ schema and service-page structure.

Tactic 3: Rewrite your H2s as questions

Most SMB sites use H2s as topical labels: “Pricing,” “Features,” “Our Approach.” Those headings are SEO-neutral and AEO-negative. Rewrite them as the questions your prospects are typing into ChatGPT.

Before:

“Pricing”
“Features”
“Our Approach”

After:

“How much does it cost?”
“What’s included in the standard plan?”
“How do you handle migrations from QuickBooks?”

The question-shaped H2 does two jobs at once. It matches what the AI engine is looking for when it scans your page (questions are how queries arrive). And it improves readability for humans, because nobody actually wants to read a section labeled “Our Approach.”

Where this helps most: product pages, pricing pages, service pages, comparison pages, “about” pages. Blog posts already tend to use question H2s.

Time to ship: 5 minutes per page. Just rewrite the H2s.

Tactic 4: Build city-and-niche landing pages (for local services)

If you sell locally — a CPA practice, a dental office, a plumbing company, a real-estate firm, a regional law firm — the AEO move that produces the most consistent citations is city-and-niche landing pages.

The pattern: one page per city you serve, per service you offer. A CPA in Austin who serves three cities (Austin, Round Rock, Pflugerville) with three services (small-business tax, bookkeeping, financial planning) creates nine pages. Each page is 600–900 words, leads with the answer (“Tax prep for small businesses in Round Rock, TX costs $400–$1,200 depending on entity type…”), includes the city name in the title and first paragraph, lists 3–5 FAQs in FAQPage schema, and ends with one specific call-to-action.

This is the single most-mentioned pattern across local AEO case studies in early 2026. The dentist case study above used this exact pattern. The CPA equivalents are showing 3-5x citation uplift inside 90 days for “[service] in [city]"-shaped AI queries.

Time to ship: about an hour per page if you start from a template. Build the first one slowly, then duplicate and modify.

Tactic 5: Update three pages every month — freshness is a Perplexity favorite

Perplexity weighs freshness more than the other engines. A page last updated in 2024 is at a structural disadvantage versus a 2026 update of the same page, regardless of the underlying content quality. The fix is a monthly review: pick three of your highest-traffic pages, update the year stamps, refresh any data points, add one new FAQ to each, and republish with a new “Last updated” date.

This is the most underrated tactic on the list because it doesn’t feel like new work. You’re not writing new pages; you’re maintaining old ones. But the citation impact is real, especially for Perplexity and Google AI Overviews, which both prefer recently-touched pages over comparably-authoritative older ones.

The minimum viable monthly refresh:

Change the date in the page metadata
Add “(updated [Month] 2026)” to the title
Add a one-sentence “What changed in [Month]” callout near the top
Refresh any year references in the body (2024 → 2026)
Add one new FAQ to the schema and body
Republish — same URL, same internal links

Time to ship: 15 minutes per page, 45 minutes per month total.

Tactic 6: Write one piece of original data per quarter (the citation magnet)

AI engines cite pages that contain original data more than pages that summarize others’ data. “Original data” doesn’t have to mean a 50-respondent survey. It can be:

A tally of pricing across 10 competitors in your space (one afternoon of research)
A before/after of your own results on a specific tactic (anything you can measure on your own site or your own client work)
A breakdown of how your industry typically structures something (fee schedules, contract terms, project timelines)
A small-sample comparison of three tools you’ve actually used

The threshold is low and the citation payoff is high. One original-data post per quarter, deliberately written to be cited (clear data points, easy-to-pull numbers, dates and methodology stated), is the single piece of content that pays compounding dividends for AEO over the next year.

The format AI engines extract from cleanly: a clear <h2> framing the data (“Bookkeeper pricing in Austin: a 2026 survey”), a one-paragraph methodology note (“Surveyed 12 bookkeeping firms via their pricing pages, May 2026”), a table of the data, then a 100-word interpretation. That’s the whole post. Citations follow.

Time to ship: an afternoon for research, an hour for the writeup. One per quarter is the target.

Tactic 7: Manual prompt testing (the free tool nobody talks about)

This is the diagnostic you can run today with zero tooling: open ChatGPT, Perplexity, and Gemini in three browser tabs. Type the five most important questions your customers would ask. Read the answers. Note which sites get cited.

That is your AEO baseline. It costs nothing. It takes 20 minutes. It tells you exactly which competitors you’re behind, what their pages look like, and where the obvious holes are. Repeat monthly.

The reason this beats paid AEO tools for solo operators: the paid tools are valuable when you have hundreds of prompts to track and want automated alerts on volatility. If you have five high-value queries and 30 minutes a month, manual testing produces the same useful information with no subscription.

What to look for in the manual tests:

Who gets cited? That’s your real competitive set in AI search, which is often different from your competitive set in Google search.
What format do they use? Pages with bulleted answers, comparison tables, and FAQ schema show up disproportionately.
What’s missing from the AI answer? A question that’s poorly answered everywhere is a content opportunity for you specifically.

Time to ship: 20 minutes monthly, repeating.

The honest list of things to skip

The AEO space has a lot of fake-deep advice. Three things you don’t need to bother with as an SMB:

llms.txt. It’s a 30-minute setup with no confirmed lift in 2026. No major AI provider has confirmed they use it as a ranking signal. Practitioners who tested it report low or no traffic gain on its own. It’s not harmful, but if you have to choose between writing an llms.txt and any of the seven tactics above, pick a tactic.
“Chunking content for LLMs.” Google’s mid-May 2026 AI optimization guidance explicitly addressed this — there is no special “AI-friendly chunking” that produces lift over helpful, structurally clean content. The advice is a relic of a brief window in 2024 when it might have helped. It doesn’t now.
Paid AEO tools (for under $1M revenue businesses). Otterly.ai at $29 and HubSpot AEO at $50 are reasonable upgrades when you’ve maxed out the free tactics and have brand-tracking budget. They are not the starting point. If you’re under $1M in revenue, every dollar that goes to AEO software is a dollar that could have gone to one more piece of original-data content.

What this means for you

If you’re a solo founder or freelancer: ship Tactics 1, 2, and 3 this week — they take a combined 90 minutes per page and the impact compounds. Add Tactic 7 (manual testing) as a recurring monthly ritual. Skip the rest until you have evidence the basics moved the needle.

If you run a 2–10 person local services business: Tactics 1–5 are all relevant; Tactic 4 (city-and-niche pages) is the highest-leverage. Plan one new city-and-niche page per month for a year. By the same time next year, you’ll have 12 pages stacking citations.

If you’re a small e-commerce operator: Tactics 1–3 and 5–7. Skip Tactic 4 (city pages don’t fit your model). Focus Tactic 6 (original data) on category-level posts: “Pricing across 20 [your product category] brands in 2026” is exactly the post that gets cited by ChatGPT when a shopper asks “what should I expect to pay for [your category].”

If you’re a B2B SaaS founder: Tactics 1, 2, 3, 6 are non-negotiable. Tactic 7 is essential because your prospects increasingly start in ChatGPT, not Google. The garrywilliamsss case (12 days from launch, every paying customer from ChatGPT, none from Google) shows the maximum-velocity version of what this looks like.

If you’re a marketing consultant or agency: the seven tactics above are your AEO starter pack for client engagements. Walking a client through manual prompt testing in the first meeting is the most disarming opener — they see who gets cited, they see they don’t, and they’re already sold on doing the work.

What this can’t fix

Three honest limits before we close.

AI engines amplify existing authority. If your site has zero backlinks, zero domain history, and zero topical depth, FAQ schema won’t manufacture citations. AEO compounds on top of basic site health — not in place of it.
The channel is changing faster than the playbooks. Anything written today might be out of date by August. The seven tactics here are durable bets, but a Google model update or a ChatGPT routing change can move the line overnight. Stay subscribed to one good AI search newsletter and re-test your baseline monthly.
Citations don’t equal sales. Getting cited in a ChatGPT answer is necessary but not sufficient. The user still has to click through, land on a page that delivers what was promised, and convert. Treat AEO as a top-of-funnel win that needs landing-page work behind it.

The bottom line

The AEO tool-vendor SERP makes this look like a paid-only game. It isn’t. The seven tactics above are 100% free, take under five hours total to ship the first round, and produce citation lifts that compete with $50–$400-a-month subscription tools. The tools exist because tracking gets harder at scale, not because the underlying work requires software.

If you’re an SMB starting from zero, the order is: Tactic 1 (answer-first writing), then Tactic 2 (FAQ schema), then Tactic 7 (manual testing) as a monthly check. That sequence — write the answer first, mark it up with schema, watch what happens in the engines — is the entire AEO loop, free, with no scary jargon.

When you’re ready to layer in monitoring tools, HubSpot’s free AEO Sensor is the easiest first stop (we walked through it in our launch post this week). When you’re ready to invest in deeper AI search fluency, our SEO Mastery course covers the structural foundations that AEO builds on top of, our Canva AI 2 for SMB Marketers walks through the visual side of content built for AI summary surfaces, and our Claude for Small Business course shows how to use AI inside your business — including for content production at the volume AEO requires.

The first three tactics this week. Tactics 4–7 over the next 90 days. Then check the manual baseline. The number of competitor brands ahead of you in the AI answers will be smaller by then.