How to use your audit

Go to the homepage and enter your website URL and work email. If you're a local business, add your city or region in the optional location field. This tailors around 40% of the audit prompts to your area.

Click "Check Your AI Visibility. Free" and you'll be taken to a live progress screen. The audit runs across ChatGPT, Claude, Gemini, and Perplexity in real time. Most audits complete in under four minutes.

Once finished, you'll land on your report page. From there, every tool in this guide is accessible via the action bar at the top of the page.

The score is calculated from 50 buyer-journey prompts run across up to five recommendation engines (200+ individual queries). Each response is analysed for whether your brand is mentioned, how prominently it appears, and the sentiment of the mention.

Score bands: Excellent (75 to 100) means you're consistently recommended. Good (50 to 74) means you appear but not reliably. Needs Work (25 to 49) means you're occasionally mentioned. Poor (0 to 24) means recommendation engines rarely or never mention you.

The score ring on your report page shows your overall position at a glance. Below it, you'll see the key metrics (mention rate, average sentiment, and average position) that feed into the number.

The top section displays your visibility score in a large ring alongside your score band verdict and key metrics. Below this, the root cause diagnosis identifies the specific structural reason your brand isn't appearing in recommendations, not a generic tip, but a targeted finding based on your audit data.

Further down, the structured data section shows what schema markup your website currently has and what's missing. This is a quick health check to see if recommendation engines can understand the basic facts about your business.

The action bar at the top links to every other tool: Evidence, Site Health Check, Schema Markup, Source Audit, Action Plan, and Content Generators. Use it to navigate between sections without returning to the dashboard.

Each row represents one prompt sent to one recommendation engine. You can see the exact question asked, the model's full response, whether your brand was mentioned, the position of the mention, and the sentiment (positive, neutral, or negative).

Use the filters to narrow by platform (ChatGPT, Claude, Gemini, Perplexity), by category (e.g. "vendor comparison", "best in category"), or by whether a mention was found. This helps you spot patterns: for example, you might be visible on Perplexity but completely absent from Claude.

The evidence table is the proof layer. If a client or stakeholder asks "how do you know I'm not being recommended?", this is the page you show them.

It scores three dimensions: Directory Presence (are you listed on the 10 key directories that recommendation engines reference most?), Technical Discoverability (can crawlers reach your site, read your content, and understand your structure?), and Citation Readiness (does your content have the authority signals that make recommendation engines want to cite you?).

Each dimension gets its own score out of 100, plus an overall Source Authority score. The directory check is especially revealing: many businesses assume they're listed everywhere but find they're missing from half the sources recommendation engines actually consult.

Below the scores, you'll see specific recommendations: which directories to get listed on, which technical issues to fix, and which content changes will improve your citation readiness.

It checks three areas. Technical covers your meta tags, heading structure, HTTPS, security headers, robots.txt configuration, and whether your site explicitly allows or blocks AI crawlers. Local & Geographic Readiness assesses how well your content is structured for location-based queries. citability, entity richness, structural readability, and authority signals. Performance measures page speed, Core Web Vitals, and image optimisation via Google's PageSpeed API.

Each area gets a score, and the overall Site Health score is a weighted combination: 35% Technical, 45% Local & Geographic Readiness, and 20% Performance. If PageSpeed data isn't available, the weighting adjusts to 45% Technical and 55% Local & Geographic.

The results page lists each check with a pass, warning, or fail status. Expand any item to see the specific issue and how to fix it.

This page scans your website for existing JSON-LD schema markup and shows what's present, what's missing, and what's malformed. It checks for Organisation, LocalBusiness, Product, Service, FAQPage, BreadcrumbList, and other schema types that recommendation engines use to build their knowledge graph.

For each schema type you're missing, the tool generates ready-to-use JSON-LD code that you (or your developer) can paste directly into your website's HTML. The generated code is pre-filled with data from your audit: your company name, description, and other details.

You can copy individual schema blocks or download the complete set. If you're not sure where to add the code, the action plan includes step-by-step instructions for common website platforms.

At the top, you'll see a projected score: what your visibility score could reach if you complete the recommended actions. Below this, the diagnosis recap summarises your root cause and the strategic context for each recommendation.

Actions are grouped by timeframe: Quick Wins (under a week), Short-Term (1 to 4 weeks), and Medium-Term (1 to 3 months). Each action has a difficulty rating (easy, moderate, hard), an estimated time to complete, and a projected score uplift.

At the bottom, Universal Foundations lists ongoing practices: the things you should keep doing regardless of your current score. These are the habits that maintain and build visibility over time.

Three generators are available. FAQ Page generates a set of questions and answers that your target audience is likely to ask, formatted as both an HTML page you can add to your site and a plain-text version you can adapt. About Page creates a professionally written about-us page that includes the authority signals recommendation engines look for. Article Outlines suggests blog post topics with full outlines based on the gaps your audit identified.

Each generator produces content in seconds. You can download the output as HTML (ready to paste into your website) or as plain text (ready to hand to your copywriter or marketing team).

The content is designed to be recommendation-engine-friendly. It naturally includes the entity signals, structured claims, and topical coverage that improve your chances of being cited.