In this guide
- Where should your questions come from?
- Which questions deserve an answer on your site?
- How do you write an answer an AI assistant will quote?
- Should FAQs live on one page or across the site?
- Does FAQ schema still matter?
- How does an answer turn into an enquiry?
- How do you measure whether FAQs produce leads?
- What are the common mistakes?
- Frequently asked questions
- Put your FAQs to work
AEO for service businesses turns FAQs into leads when each answer does two jobs: it resolves the question clearly enough that an AI assistant or searcher can quote it, and it hands the reader an obvious next step. Source questions from real sales conversations, answer them in the first sentence, place them on the pages where buyers decide, and track enquiries.
Most service businesses treat FAQs as a dumping ground at the bottom of the site. That wastes your best sales material. The questions prospects ask before they buy are exactly the questions people now type into ChatGPT, Perplexity and Google's AI Mode. Answer engine optimisation (AEO), explained in full in our AEO and GEO pillar guide, simply means structuring those answers so machines can find, trust and repeat them.
Where should your questions come from?
From the conversations you already have with buyers, not from a keyword tool. Keyword tools show what people search; your inbox shows what stops them from booking.
Spend an hour collecting questions from:
- WhatsApp and DM threads. Scroll back three months and copy every question a prospect asked before paying.
- Sales or discovery calls. Ask whoever takes calls to list the five questions they answer every week.
- Lost deals. The reason someone didn't proceed ("too slow", "unclear what's included") is a question you didn't answer early enough.
- Review text. Customers often mention what reassured them.
- Lead-form custom questions. If you run Meta lead ads, the answers people give to your qualifying questions reveal their worries.
We run our own enquiries through a lead centre in the GRWW Ops Console that pulls website form submissions and Meta lead-ad enquiries into one place. Reading them side by side makes the repeated questions obvious, and that same habit works with a spreadsheet.
Which questions deserve an answer on your site?
Prioritise questions that sit close to a buying decision, then the ones that filter out poor-fit leads. Group what you collected into four buckets:
| Question type | Example (home renovation firm) | Where it belongs | Next step to offer |
|---|---|---|---|
| Price and scope | "How much does a kitchen renovation cost?" | Service page, near the top | Price-range guide or quote request |
| Process and timing | "How long does a bathroom refit take?" | Service page, below the process section | Book a site visit |
| Risk and trust | "Are you licensed and insured?" | About page and service pages | Link to licences, reviews, past work |
| Fit and eligibility | "Do you take on small jobs?" | Service page or dedicated FAQ | Short form or WhatsApp chat |
Questions about general topics ("what is a load-bearing wall?") can earn traffic, but they rarely produce enquiries directly. Write those as blog posts later; keep your service pages focused on buying questions.
How do you write an answer an AI assistant will quote?
Lead with a direct, self-contained answer in one or two sentences, then add the detail that makes it trustworthy. A model lifting a passage needs it to make sense out of context.
A simple formula:
- Answer first. "Most kitchen renovations we quote fall within a typical range of X to Y, depending on cabinetry, appliances and layout changes."
- Name what changes the answer. Two or three factors, in plain words.
- Give a concrete example or rule. "Moving plumbing usually adds a week."
- Add the next step. "Send us your floor plan for an itemised estimate."
Avoid answers that start with "It depends" and stop there. "It depends" is true, but it's not quotable and it's not helpful. Give a range, label it as an estimate, and explain what moves it.
Our guide to writing content AI assistants quote covers sentence-level techniques like defining terms and putting numbers next to their context.
Should FAQs live on one page or across the site?
Both, with service pages doing most of the selling. A buyer reading your "Commercial cleaning" page should find the five questions about commercial cleaning right there, not hunt for a separate FAQ page.
Use this rule of thumb:
- Service pages: 4–8 questions specific to that service, placed after the main description and before the final call to action.
- Central FAQ page: company-wide questions (payment terms, service areas, guarantees, how to get started) plus links into each service page.
- Blog posts: deeper explanations for questions that need 800+ words.
Avoid copying the same answer across many pages. If two services share a question, write it once in full and summarise with a link elsewhere. Our post on service pages that rank and convert shows how FAQs fit inside a full service-page layout.
Does FAQ schema still matter?
Yes for machine readability, but not for the old rich-result boost. In 2023 Google limited FAQ rich results to a small set of well-known, authoritative sites, mostly government and health. Most businesses no longer get the expandable FAQ listing in search results.
FAQPage structured data still labels your questions and answers clearly for any crawler that reads it, and it costs little to add when your FAQ content is already on the page. Just make sure the markup matches the visible text word for word. Check it with the Schema Markup Validator after every change.
How does an answer turn into an enquiry?
By ending at the point of highest intent with the lowest-effort way to continue. Someone who just read your pricing answer is ready for a quote, not your newsletter.
Match the next step to the question:
- Price questions: link to a quote form that asks only for what you need to price the job.
- Timing questions: offer a calendar link or "check availability" button.
- Trust questions: show two relevant reviews and a link to past projects.
- Quick questions: a WhatsApp or chat button.
Our own Meta ads data supports offering a messaging option. Across our client accounts, click-to-message campaigns usually delivered the cheapest cost per enquiry, while lead forms delivered more leads to qualify. People are comfortable opening a chat with a question; your FAQ can do the same on your site.
Keep the call to action specific. "Get a fixed price for your bathroom" beats "Contact us".
How do you measure whether FAQs produce leads?
Track enquiries that start from pages with FAQs, and ask new leads which question brought them in. You won't get perfect attribution, especially when an AI assistant answered part of the question without a click, but you can see direction.
| What to track | How | What it tells you |
|---|---|---|
| Enquiries per service page | Form tracking and UTM-tagged chat links | Which pages convert readers |
| Clicks on FAQ calls to action | Event tracking in your analytics tool | Which answers prompt action |
| "How did you find us?" | One optional field on your form | AI assistant mentions you can't see in analytics |
| Repeated questions in sales calls | Monthly review of call notes | Gaps your FAQs still don't cover |
| AI answer mentions | Monthly prompt checks | Whether assistants repeat your answers |
Review the list each quarter. Retire questions nobody asks, rewrite answers that generate confused follow-ups, and add the new ones your team keeps hearing. Even a simple spreadsheet of enquiries by source, reviewed monthly, beats guessing.
What are the common mistakes?
Most FAQ sections fail because they're written to sound good rather than to answer.
- Questions nobody asks ("Why are you the best?").
- Answers that dodge pricing entirely.
- Accordions that hide answers from crawlers by loading them only after a click via JavaScript.
- Dozens of questions on one page with no grouping.
- Outdated answers (old prices, discontinued services) left live.
If you serve a local area, pair this playbook with our guide to getting recommended for "best near me" searches.
Frequently asked questions
How many FAQs should a service page have?
Four to eight is usually enough. Pick the questions that most often come up before people book, and move broader topics to blog posts. More questions isn't better if they bury the important ones.
Should I publish prices in my FAQs?
Publishing typical ranges usually helps: it pre-qualifies buyers and gives AI assistants something concrete to quote. Label ranges as estimates and explain what moves the price. If you truly can't share numbers, explain how you price and how fast you can send a quote.
Can AI tools write my FAQ answers?
They can help draft, but the substance must come from your experience: your process, your timelines, your exceptions. Generic AI answers sound like every competitor and give assistants no reason to cite you. Edit every draft against what you actually do.
Do accordions hurt SEO?
Not if the answer text is in the page's HTML when it loads. Problems arise when answers are fetched only after a click, which some crawlers never trigger. Check by viewing the page source and searching for an answer's wording.
How often should I update FAQs?
Review them quarterly and whenever prices, services or policies change. Add a question each time your team hears the same one three times in a month.
Put your FAQs to work
Your best leads already ask the same questions; answering them well, in the right place, with the right next step is one of the cheapest growth moves a service business can make. If you'd like help restructuring service pages and FAQs, talk to us about a fixed-price project.