Frequently asked questions

Three classifications, and they are not interchangeable. Recommended means the engine names your business as the answer (or one of the top answers) to the query. This is the outcome that wins customers. Mentioned means the engine names your business somewhere in the response (a citation, a footnote, a passing reference in a longer list) but not as a recommendation: better than invisible, not enough to win the customer. Not found means your business does not appear in the response at all. The full report breaks the count out by engine and by query, so you see exactly where you are winning, where you are mentioned but losing, and where you are absent.

Most often it is one of three things. Your website does not clearly state what you do, who you serve, and where, so the engine cannot match you to a query. Your site has no FAQ pages, no schema markup, and no llms.txt, so the engine cannot extract clean answers from your content. Or you have very few external mentions (Google reviews, directory listings, press), so the engine has nothing to corroborate that you exist. The full diagnostic report tells you which of these is hurting you most.

SEO optimises for Google's search results page, where users scan blue links. AI search engines optimise for direct recommendations: the engine reads your site and your competitors' sites, decides which one fits the query best, and names a single business. The signals overlap (clear content, fast pages, structured data) but the format and the testing method are completely different. You can rank #1 on Google and still be invisible to ChatGPT.

AEO stands for Answer Engine Optimisation. It is the practice of optimising your business for direct AI recommendations rather than for Google's blue-link results. The signals overlap with SEO (clear content, fast pages, structured data) but the testing method and the desired outcome are different. SEO wins clicks. AEO wins direct mentions inside the answer itself, when Google AI Overview, ChatGPT, Perplexity, Gemini, Claude or Grok decide which single business to recommend. GetRecommended.io tests the AEO side specifically: live queries, six engines, recommended versus mentioned versus not found.

There are two tiers. The free scan runs 3 search queries (1 you enter and 2 we generate from your industry and location) across 3 major AI engines (ChatGPT, Perplexity and Gemini) in under 60 seconds. No account, no card. The full diagnostic report runs 10 queries across all 6 engines (adding Claude, Grok and Google AI Overview), and adds a website technical audit, a competitor benchmark and a remediation checklist. Both versions show you whether your business was recommended, only mentioned or not found, plus who was recommended in your place.

Six in total: Google AI Overview, ChatGPT, Perplexity, Gemini, Claude and Grok. These cover the engines Australian, New Zealand, United Kingdom, United States and Canadian small business customers are actually using day to day. The free scan tests 3 of them (ChatGPT, Perplexity and Gemini). The full diagnostic report benchmarks your visibility across all 6.

AI engines do not give the exact same answer every time. Ask the same question twice and the wording can change. So rather than reading the wording, we record the outcome: recommended, mentioned, or not found. That outcome stays the same across repeated runs in the vast majority of cases. We run live queries at the moment of your scan rather than relying on old cached data, so the result reflects what your customers would see right now. If an engine cannot answer (rate limited, not available in your region, or temporarily down), we mark it Unavailable rather than guess. And if an engine makes up a fact about your business, the report flags it: a wrong AI answer about you is itself part of the visibility problem worth fixing.

Almost immediately. The moment your payment succeeds the scan runs and the report generates, typically in a few minutes, not days. You get an email with a magic-link sign-in to your customer dashboard the moment it is ready, and the report is also delivered as a downloadable PDF. The report is AI-generated with no human review step in between, which is what makes the turnaround fast, and which we disclose at checkout.

Because AI engines are non-deterministic and personalised. The same question can return a different answer on a different run, on a different account, in a different location, or a week later: the engines weigh things slightly differently each time and tailor answers to what they know about the person asking. Your report is a point-in-time snapshot of what the engines returned for neutral, non-personalised queries at the moment of your scan. So if you search yourself and see something different, neither result is wrong; you are seeing the same non-determinism that makes any single AI ranking impossible to guarantee. That is also why we record the outcome (recommended, mentioned, not found) rather than the exact wording, and why we suggest re-scanning every few months.

Through a licensed search-data provider that returns Google's AI Overview results, not by scraping Google ourselves. That keeps data collection within terms and gives us a consistent, repeatable read of what Google's AI Overview shows for a query. The other five engines (ChatGPT, Perplexity, Gemini, Claude and Grok) are queried through their own access. Our full method, including which licensed sources we use, is set out in our Methodology Statement.

Your score is built from four pillars: AI Presence (how often you appear at all across the engines), Content Quality and Depth (how extractable and answer-shaped your content is), Authority and Trust (reviews, mentions and the signals engines use to corroborate you) and Technical Readiness (schema, llms.txt, sitemap, mobile and the mechanics engines rely on). Each pillar is scored with the working shown, and together they make the headline score. The Methodology Statement describes the pillars and their limitations in full.

No. Nobody can guarantee that and we will not pretend otherwise. AI engines control their own algorithms and change them constantly. The report tells you exactly what is hurting your visibility right now and what to fix. Implementing those fixes gives you the best possible chance of being recommended, but the engines have the final say. Results are a point-in-time snapshot and vary by engine, prompt, date, model, region and account.

The full report includes the executive summary, full visibility results across 6 AI engines and 10 queries, a website technical audit, a competitor benchmark of the 3 to 5 businesses being recommended in your space, a prioritised remediation checklist with quick wins flagged, and a library of educational videos walking you through each fix. Delivered as a downloadable PDF and accessible from your customer dashboard for the life of your account. Pricing is 297 Australian Dollars; local pricing is set per region for New Zealand, the United Kingdom, the United States and Canada, and customers outside these five regions can still purchase the report.

This is more common than you might think, especially for newer businesses, businesses with weak websites, or businesses in categories dominated by larger brands. The report is designed for exactly this case. The remediation checklist effectively becomes a launch sequence: llms.txt, schema, FAQ pages and the prioritised content gaps to fill. Mechanical fixes show up in AI responses within days to a few weeks. Trust signals (Google reviews, external mentions) take longer to build but compound. Most customers see meaningful change within the first three months of working through the report.

Yes. AI engines are increasingly used for B2B research (Claude particularly, but also Perplexity and ChatGPT). The methodology is the same: queries reflecting how your buyers actually search, scored across all six engines. B2B results sometimes show different patterns (more weight on long-form content, less on Google reviews, more on industry directories) and the report adapts to your business profile. Service businesses, software, agencies, consultancies and professional services are all in scope.

The report is honest about this. If you compete in a category where AI engines almost always recommend two or three giant brands (national banks, the largest insurers, big retail chains), the report will say so. The play in that case is usually long-tail: winning specific, narrower queries where the giants are not specialised. The report identifies those queries and tells you what to build to win them. Most small businesses that come to us are not actually trying to outrank the giants for the broadest query, they want to win the specific question their ideal customer is asking.

Your report is accurate for the moment it was run: a point-in-time snapshot, not a permanent verdict. The fixes it recommends (clear content, schema, llms.txt, reviews, the specific gaps it identifies) stay valid far longer than any single engine result, because they are about making your business legible to any answer engine. The visibility scores themselves will drift as engines update and as you and your competitors change, which is exactly why a re-scan every three to six months is the right rhythm to check progress.

It means there is limited verifiable evidence about your business in the places AI engines look, at the time of the scan. It is not a judgement on how good your business is. Plenty of excellent businesses score low simply because their website, structured data, reviews and external mentions do not yet give the engines enough to confidently recommend them. A low score is a to-do list, not a verdict: the report turns it into the specific, prioritised fixes that raise it.

The competitor data is a neutral, sourced, point-in-time observation, not a ranking, an endorsement, or a guarantee. It reflects what publicly available sources and the engines returned at the moment your report ran, from licensed data providers and public profiles only: we do not access any private or proprietary data about other businesses. Aggregates like review counts can be slightly stale by the time you read them. If a business believes something about it in a report is wrong, it can ask us to correct or remove the data by emailing hello@getrecommended.io.

No. The report is AI-generated and delivered without a human reviewing it first, which we disclose and ask you to acknowledge at checkout. The workflow is built to be reliable enough to ship without that step, and that is what makes delivery near-instant. If anything in your report looks wrong, reply to the email it arrived in and a human will look at it directly.

The full diagnostic report is 297 Australian Dollars, one-off, with no subscription and no recurring billing. Local pricing is set in your region's currency at checkout: Australian Dollars in Australia, New Zealand Dollars in New Zealand, British Pounds in the United Kingdom, United States Dollars in the United States, Canadian Dollars in Canada. Each region has its own price set per market, not a live currency conversion. Customers outside these five regions can still purchase the report, and the methodology, engines tested and report content are identical. The free 60-second scan stays free regardless of where you are.

One-off. You pay once, 297 Australian Dollars, no subscription, no recurring billing, no automatic renewal, no upsells inside the report. Local pricing is set per region in Australian Dollars (Australia), New Zealand Dollars (New Zealand), British Pounds (the United Kingdom), United States Dollars (the United States) and Canadian Dollars (Canada). Customers outside these five regions can still purchase. If you want to scan again in three or six months to check progress, that is a separate purchase at the same price.

Yes. Local pricing is set in your region's currency at checkout: Australian Dollars in Australia, New Zealand Dollars in New Zealand, British Pounds in the United Kingdom, United States Dollars in the United States, Canadian Dollars in Canada. The methodology, engines tested and report content are identical across regions. Customers outside these five regions can still purchase the report. Australia is our primary market because that is where most of our early customers are, but the report is built for any small or mid-sized business with a website and customers researching online before buying.

Your card statement will show GetRecommended.io as the merchant. Your receipt arrives in your inbox within minutes of payment, with the magic-link sign-in for your customer dashboard. If your bank shows the charge in your local currency (Australian Dollars, New Zealand Dollars, British Pounds, United States Dollars or Canadian Dollars), that is normal: each region has its own price set in the local currency at checkout.

We will refund any report that has not been generated yet, no questions asked. Once the report has been generated and delivered, refunds are case-by-case. If the report has a clear quality problem (missing sections, wrong business, broken delivery) we will fix the report and refund the difference between expectation and delivery. If you simply did not like what the report told you, we will not refund the work, but we will offer to discuss the findings.

AI engines change. Your content, competitors and review profile change. A re-scan every three to six months tends to be the right rhythm: long enough that fixes have had time to land, short enough that you catch competitor moves and engine shifts before they cost you. A re-scan is a fresh report at the same price (297 Australian Dollars in Australia, with local pricing set per region for New Zealand, the United Kingdom, the United States and Canada, and available to customers in other regions too), not a discounted re-issue, because everything is genuinely re-tested. The first signal that you need a re-scan is usually a quiet drop in customers mentioning they found you on ChatGPT.

If you have a question about something in the report, reply to the delivery email. We read replies and answer questions about your specific report, typically within one business day. We do not run a 24/7 support queue, and we do not do the implementation work for you, but we will help you understand any finding in the report. The customer dashboard at getrecommended.io/dashboard also keeps the PDF and the educational videos available for the life of your account, so you do not need to chase a download link later.

No. A low or zero visibility result is not grounds for a refund: the report is doing its job, which is to accurately show you the current state of AI search for your business. We do not control what the engines recommend, and finding out you are invisible is exactly the information you paid for. What we will refund is a report with a genuine quality or delivery problem (wrong business, missing sections, broken delivery), and your non-excludable rights under the Australian Consumer Law (and the equivalent law in your country) always apply. Full detail is in our Refund Policy.

If you buy a full report and publish a qualifying public post about it, you can claim a cashback. You use one of our post templates, tag our official account, include the required disclosure that you received a cashback offer for sharing, and submit the live public link from your dashboard within 30 days of purchase. We verify it within 10 business days, and once approved the cashback is paid back to the card you bought with. The exact amount and currency are shown in your dashboard. The full terms are in Section 14 of our Terms of Service.

No, and the distinction matters. The share cashback is a promotional reward for posting publicly about your report. For practical reasons it is processed as a partial refund back to your original card, but it is not a consumer-law refund and it is separate from our Refund Policy: claiming it does not affect your statutory rights, and not qualifying for it is not a refund dispute. Section 14 of our Terms of Service sets out the full distinction.

Yes. The report is licensed for your business, which means you can share the PDF with your team, your developer, your marketing agency or your accountant: anyone helping you implement the fixes. The only restriction is that you cannot resell it or distribute it as your own work product. The customer dashboard at getrecommended.io/dashboard hosts the same PDF and the educational videos for the life of your account.

We store the minimum needed to deliver and support the report: your business name, website, industry, location, contact email, the queries we ran, the responses each AI engine returned and the report we generated. We do not sell your data, we do not feed your business name back into AI engines as training data, we do not share customer lists and we do not run remarketing pixels that follow you around the web. Data lives in our database (Supabase, hosted in Australia). To delete your account and report, email the address on your receipt and we will remove your data within thirty days.

Only business-level public facts: whether a business appears in AI answers, its public profiles, and publicly available review aggregates, sourced through licensed providers. We do not collect personal data about other businesses' owners or staff, and we do not republish individual review text. Any business that appears in a report can ask us to correct or remove its data by emailing hello@getrecommended.io.

No tracking cookies and no cross-site advertising pixels. We use essential cookies for signing in to your dashboard (Supabase) and for processing payment (Stripe), plus Plausible for privacy-friendly analytics, which is cookieless and does not track you across other sites. We do not run Google Analytics, a Meta pixel, or remarketing trackers.

The mechanical fixes (schema, llms.txt, meta descriptions, FAQ content) start showing up within days to a few weeks as engines re-crawl your site. Perplexity and Gemini tend to reflect changes first; ChatGPT typically follows over two to six weeks. The trust signals (Google reviews, external mentions, citations from other sites) take longer because you have to earn them, and tend to compound from week six onwards. Customers who work through the report systematically usually see meaningful change within the first three months.

Yes, significantly. AI engines use reviews as a trust signal when deciding which business to recommend. A business with a strong, recent Google review profile will tend to win the recommendation over an otherwise identical business with none. Trustpilot, industry directories and review aggregators all count. The report shows you exactly where your review profile is hurting you and which platforms to prioritise.

llms.txt is a plain-text file at the root of your website that summarises what your business does in a format AI systems can read quickly. It is a proposed standard, not a guarantee. Google has confirmed that no special files or markup are required to appear in AI Overviews, so llms.txt should never be treated as a magic lever. But it is a cheap, fast win that some AI assistants do read, and adding one alongside strong on-page content and reviews removes any excuse for an engine to misread you. The full report includes a ready-to-use llms.txt for your business.

Schema markup is structured data added to your website's HTML that tells engines exactly what your pages are about. For example, schema can tell an engine: this page is about a service called X, offered by a business called Y, in location Z, costing $W. Without schema, the engine has to guess from your text. With schema, there is no ambiguity. The report tells you exactly which schemas your site needs and gives you the code to add them.

The report is written in plain English and includes the exact code snippets where code is needed. You can hand it to any web developer or virtual assistant. The quick wins (llms.txt, meta descriptions, FAQ content) are usually a few hours of work for someone non-technical. The deeper fixes are a few hours for a developer. The educational videos in your customer dashboard walk through each fix step by step, so most customers can ship the quick-win section themselves in a single afternoon.

Yes. The fixes are platform-agnostic: Wix, Shopify, Squarespace, WordPress, custom builds all support the additions we recommend. Where a platform has a quirk (for example, Shopify's restrictions on root-level files), the report includes the platform-specific workaround. The educational videos walk through the fixes step by step so you (or whoever you hand the report to) can implement without having to translate platform jargon.

Because AHPRA's advertising rules (and the National Law behind them, with equivalents overseas) prohibit using testimonials and patient reviews to advertise regulated health services. So for AHPRA-regulated businesses we deliberately do not recommend collecting or showcasing reviews and testimonials the way we would for an unregulated business. Instead the report points you to the trust signals you can use: clear credentials, registration details, qualifications, professional memberships and authoritative, accurate information about your services.