What's inside every deck

Pillar 1: the differentiator

AI visibility & answer readiness

The question underneath everything else: when a buyer asks an assistant, do you come up? We run a 50-prompt set of the questions your buyers actually ask across ChatGPT, Gemini, Claude, Perplexity, Google AI Overviews and AI Mode, record whether you are mentioned, cited or recommended, and measure share of voice against named competitors. Then we audit the plumbing that decides those answers, because an assistant cannot cite a page it is blocked from or a page that only exists once a browser runs it.

• Mention, citation and recommendation rates across six answer surfaces, with the verbatim prompts and responses in the deck
• AI share of voice against named competitors, prompt by prompt
• Per-agent crawler access for 17 named agents: GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Google-Extended, Googlebot, Applebot, Applebot-Extended, Meta-ExternalAgent, Amazonbot, Bytespider, CCBot and bingbot
• Search bots vs training bots separated: blocking the wrong one costs you citations
• JavaScript-dependency verdict: how much of the page exists before any script runs, and most AI crawlers never run one
• llms.txt, llms-full.txt, CDN-level AI blocking and declared sitemap
• Answer-readiness scoring: answer-first openings, question-shaped headings, FAQ blocks, comparison tables, freshness signals
• Bing indexation, the index ChatGPT search leans on

Pillar 2: the foundation

Search findability

Classic search still feeds the assistants, and it still feeds your pipeline. We report where you rank and how that has moved over twelve months, the authority pointing at you, the keywords your competitors hold and you do not, and every technical fault that keeps a page out of an index or wastes the crawl budget that reaches it.

• Keyword rankings with 12-month trends, position buckets (1-3, 4-10, 11-20, 21+) and estimated traffic value
• Backlinks and referring domains with authority scoring, plus new and lost link movement
• Keyword gap against competitors: the terms they rank for and you do not, sized by opportunity
• Full technical crawl per page: status, redirect chains, titles and lengths, descriptions, canonicals, word count, internal and external links
• Missing and duplicate titles, missing descriptions, canonical gaps, thin content, broken links, noindex detection
• HTTPS and www redirect handling, sitemap URL count and lastmod range, robots.txt coverage
• Structured data by rich-result type: what you publish, which pages carry it, and the types you are missing

Pillar 3: what visitors feel

Experience & performance

Core Web Vitals from real-user field data where Google has enough traffic to report it, and lab measurement where it does not, always labelled so you know which you are reading. Alongside it, the accessibility failures that are verifiable from the page itself, and whether the site is built for the phone most of your traffic arrives on.

• LCP, INP and CLS with the source, field or lab, stated on the slide
• Mobile readiness: viewport configuration, tap targets and mobile rendering
• WCAG checks: images missing alt text, unlabelled form inputs, heading-order breaks, links with no text, missing lang attribute, missing landmarks
• Issues ranked critical, serious or moderate with counts, not a single opaque score
• Stated plainly in the deck: automated testing catches roughly a third of real accessibility issues, so this is a floor, not a clearance

Pillar 4: can you be believed

Trust, security & brand authority

Two questions in one section. First, is the site itself safe to send traffic to: HTTPS enforcement, TLS certificate and expiry, security headers, cookie flags, mixed content. Second, does the world treat your brand as a real entity, which is what lets a knowledge panel or an AI answer attach facts to your name with confidence, and what a buyer sees when they search your name to check you out.

• HTTPS enforcement, HTTP to HTTPS redirect, TLS protocol, issuer and days to certificate expiry
• Security headers scored 0-100 and graded, with the missing ones named
• Cookies missing Secure or HttpOnly flags, and suspected mixed content
• Entity records: Wikidata including official-website match, Wikipedia, Google Knowledge Graph
• Review-platform presence and ratings on Trustpilot, G2, Capterra and Clutch where relevant, with score and volume
• Brand SERP control: what occupies page one for your own name, and what you do not own
• Reputation and news footprint, plus the social profiles your site actually links to

Pillar 5: does it sell

Conversion & messaging

Traffic that arrives and bounces is a marketing cost, not a marketing result. We read your pages the way a first-time visitor does and check whether they can tell what you do, believe you, and find the next step, then do the same teardown on the pages your competitors send their traffic to.

• Value-proposition clarity graded clear, vague or missing, from what the page actually says above the fold
• Calls to action counted per page, and the pages that carry none
• Forms found, live chat detected, phone and email visible, contact page reachable
• Pricing transparency: whether a pricing page exists and is findable
• Social proof and trust badges: testimonials, review counts, client logos, certifications and payment marks
• Landing-page teardowns for your key pages and for competitor pages, side by side

Pillar 6: can you prove it

Measurement & compliance

If nothing is measured, no fix in this deck can be proven to have worked, and if your tags run before the banner is answered, the measurement you do have is a liability. We fingerprint what is installed on the site and report the gaps between the channels you are running, the channels you are measuring, and the permission you are asking for.

• Analytics platforms detected, plus the CMS underneath
• Google Tag Manager and other tag-management containers
• Advertising pixels by platform, and CRM or marketing-automation tracking
• Consent management platform detected, and whether Google Consent Mode is wired to it
• Tracking tags that carry no consent gate, named individually, with your own markup as the evidence
• Cookies your server sets before a visitor clicks anything, named with their lifetimes and checked against what your cookie policy declares
• Stated plainly in the deck: we scan as a first-time visitor with no cookies and no clicks, so everything we name is real and a quiet result is a floor, not a clearance
• The channel gap: what you are advertising on versus what you can actually attribute

The market view: Professional and Enterprise

Competitors & paid channels

The six pillars explain your own site. This section explains the market you are being judged against. Competitors are auto-discovered from your own market rather than guessed, then benchmarked beside you, and we report exactly what everyone is running in paid without touching a single ad account. Everything here comes from public ad libraries and public data.

• Competitors auto-discovered from your market, benchmarked on authority, keywords, backlinks, AI share of voice, ad activity and content depth in one positioning matrix
• Whether you and each competitor run Google, Meta and LinkedIn ads, with the live creatives and the messaging they lead with
• How long each campaign has been running, a strong signal of what is working
• Google Ads paid keywords and estimated monthly spend, labelled as an estimate
• The landing pages behind the ads, torn down the same way as your own
• Social channels reported as investment tiers and creative velocity, never invented dollar figures, because credible social spend data does not exist at any sane price
• Content scoring and topical maps against the live top 20 results, with the topics nobody has claimed
• Industry modules auto-detected: local, e-commerce, SaaS, publisher or app
• Opportunity value: what closing each gap is worth per month