GEO: How to Get ChatGPT, Gemini and Perplexity to Recommend You
GEO in practice: how to get ChatGPT, Gemini and Perplexity to cite your business. AI crawlers, quotable content, brand mentions and a 90-day plan.
Read
An SEO audit is a structured review of a website that answers three questions: can Google crawl and index it, does Google understand what each page is about, and does the content answer customers’ questions better than the competition? A good audit doesn’t end with 300 warnings exported from a tool — it ends with a short list of fixes ranked by their impact on traffic and revenue. Below is the 9-step checklist we work through on business websites and online shops.
An SEO audit is a technical and content diagnosis of a website from a search engine’s point of view. It covers indexing, URL and link structure, content, performance, structured data and off-site signals. Unlike ongoing SEO work, it has a clear start and finish: it ends with a report and an implementation plan.
Run one in four situations:
| Tool | What it’s for in an audit | Cost |
|---|---|---|
| Google Search Console | Indexing, queries and pages, field Core Web Vitals, links, Generative AI performance report | Free |
| PageSpeed Insights | Per-URL performance: real-user data plus a lab test | Free |
| Screaming Frog SEO Spider | Full-site crawl: status codes, titles, headings, canonicals, redirects | Free up to 500 URLs, then an annual licence |
| Rich Results Test | Structured data validation | Free |
| Google Analytics 4 | Organic traffic, engagement, conversions | Free |
| Bing Webmaster Tools | Indexing and errors in Bing | Free |
For a small business site, the free tools are enough. For an online shop with thousands of products, an unlimited crawler is a must.
This is the foundation: if Google can’t crawl or index your pages, nothing else in the audit matters. Check:
noindex or a password.noindex made it to production.priority and changefreq, and only uses lastmod when it’s consistently accurate (Google: build a sitemap).Then work through the Page indexing report in Search Console. The most common statuses (Search Console Help) and how to read them:
| Search Console status | What it usually means | What to do |
|---|---|---|
| Crawled – currently not indexed | Google has seen the content but hasn’t indexed it yet; not always a problem | For key pages: strengthen the content and internal links |
| Discovered – currently not indexed | Google knows the URL but postponed crawling, often to avoid overloading the server | Check server performance; cut unnecessary URLs (filters, parameters) |
| Duplicate without user-selected canonical | The same content lives at several URLs | Add a canonical or a redirect |
| Duplicate, Google chose different canonical than user | Conflicting canonical signals | Align canonicals, internal links and the sitemap |
| Soft 404 | The page returns 200 but looks empty or broken | Return 404/410 or add real content |
| Server error (5xx) | The server didn’t respond properly | Hosting, caching, resource limits |
Only worry about crawl budget on large sites. Google’s rough guide is sites with more than a million unique pages changing weekly, or more than 10,000 pages changing daily (Google: managing crawl budget). A typical business site just needs a current sitemap and an eye on the indexing report.
Google indexes and ranks the mobile version of your content, crawled with its smartphone agent (Google: mobile-first indexing). In practice, that means three checks:
For JavaScript-heavy interfaces, and especially 3D and WebGL websites, this step is non-negotiable: a 3D scene tells a crawler nothing, so text, headings and links must exist in plain HTML.
Your crawler will flag missing, duplicate and “too long” titles. Read that list with care. Google sets no length limit for titles — it truncates them to fit the screen — and it may build the title link from other elements such as the H1, anchor text or WebSite structured data. Consistency matters more than character counts.
<title>;Google suggests judging content through three questions: who created it (bylines and author bios), how it was made (including whether AI was involved) and why — primarily to help people? (Google: helpful, reliable, people-first content). Of the four elements of E-E-A-T — experience, expertise, authoritativeness, trust — trust matters most, although E-E-A-T is not a single ranking factor in itself.
In a content audit we mainly look for:
| Metric | What it measures | Good score |
|---|---|---|
| LCP (Largest Contentful Paint) | When the largest content element renders | 2.5 s or less |
| INP (Interaction to Next Paint) | How quickly the page responds to interactions | 200 ms or less |
| CLS (Cumulative Layout Shift) | Layout stability while loading | 0.1 or less |
According to web.dev, thresholds are assessed at the 75th percentile of page loads, separately for mobile and desktop. Start with real-user data in Search Console’s Core Web Vitals report — it groups URLs by template, so you’ll see straight away whether, say, every product page is affected. Lab tests are for diagnosing causes. We cover fixes in detail in Core Web Vitals: how to speed up your website.
Organization or LocalBusiness, consistent with your Google Business Profile;Product with price and availability, plus BreadcrumbList;Article with author and publication date;Watch out for stale advice. According to Google’s documentation changelog, FAQ rich results stopped appearing in Search on 7 May 2026, following earlier retirements including HowTo and several niche types. If an audit tool tells you to add FAQPage markup “for a rich result”, that advice is out of date. We explain which schema types still make sense in our article on structured data, llms.txt and robots.txt for AI crawlers.
An audit in 2026 can’t ignore AI answers. Check:
Tool reports treat a critical error and a cosmetic one the same way. That’s why we sort every finding into one of four groups:
| Priority | Examples | Why |
|---|---|---|
| Critical — fix now | noindex on key pages, robots.txt blocks, 5xx errors, missing redirects after a migration |
Pages drop out of results or lose traffic |
| High | Duplicates and cannibalisation, poor Core Web Vitals on key templates, orphaned service pages | Direct impact on rankings and conversion |
| Medium | Titles and descriptions, redirect chains, structured data, content gaps | Better CTR and better understanding of the site |
| Low | Meta description length, alt text on decorative images | Housekeeping — fix when convenient |
When we started on shadowbrand.pl in July 2026, the domain had next to no search visibility. Search Console now shows 52,200 impressions and 1,680 clicks at a 3.2% CTR, with daily impressions reaching 1,200–1,600 in November. On new domains the audit is shorter but matters more: indexing and architecture mistakes caught at launch never get the chance to do damage.
Want someone to run this checklist for you and hand back a ready-to-use fix plan? We’ll carry out an SEO and GEO audit of your website or shop and can implement the fixes too — tell us about your project.
Run a full audit once a year and always before and after a migration, CMS change or redesign. In between, a monthly look at the Page indexing, Performance and Core Web Vitals reports in Search Console is usually enough.
As a starting point, yes; as a diagnosis, no. Automated tools flag symptoms, but they don't know your business goals, can't set priorities and often report things that don't matter, such as meta description length. The value of an audit lies in interpretation and an action plan.
It depends on the number of pages, the type of site (a business site or a shop with faceted navigation), the CMS, the number of language versions and whether the audit covers a migration or implementing the fixes. At OrbitBuilder we quote each audit individually after a short call and a look at the site.
No. Google states that robots.txt isn't a mechanism for keeping pages out of Search — a blocked URL can still appear in results, just without a description. To keep a page out of the index, use a noindex meta tag or password-protect it.
Check 301 redirects from old URLs to their equivalents (ideally one-to-one), make sure no staging noindex went live, submit the new sitemap in Search Console, verify canonicals and hreflang, and look for externally linked URLs that now return 404. Watch the Page indexing report and traffic to key pages for several weeks.
Not for rich results — Google stopped showing FAQ rich results in Search on 7 May 2026. A questions-and-answers section still helps readers and structures your content, so keeping the FAQ section itself on the page is worthwhile.
Insights

GEO in practice: how to get ChatGPT, Gemini and Perplexity to cite your business. AI crawlers, quotable content, brand mentions and a 90-day plan.
Read
How to get cited in Google AI Overviews and AI Mode: technical requirements, query fan-out, llms.txt myths and the new Search Console AI report, sourced.
Read
How to rank a Kraków business in Google and Maps: Business Profile setup, compliant reviews, location pages and structured data — a practical guide.
ReadWe grow rankings in Google and optimise for AI search engines. No “#1 in a week” promises — just Search Console data you can verify.
We use essential cookies and, with your consent, analytics (Google Analytics, Hotjar) and marketing cookies to understand how you use the site and improve it. Details in our privacy policy.