Your portfolio books Premier League, Coca-Cola, Bentley, NatWest, Innocent and Wolt. The engines that send the next brief in 2026 cannot see you.
We crawled all 33 URLs of sebantoniou.com on 27 April 2026. Two of 33 pages carry a meta description. Zero of four service pages declare an H1. Two of 16 portfolio thumbnails have alt text. The /storyboards-for-music-videos page weighs 2.93 MB. Cache-Control is set to expire on 1 January 1970. Eight schema types AI engines look for to cite a storyboard artist are absent. The portfolio is excellent. Word-of-mouth is excellent. Both are finite. The brands you want next ask AI engines for a recommendation, and right now those engines do not have your name. Ninety days fixes that.
What a 33-URL crawl found on 27 April 2026
A storyboard artist whose own portfolio thumbnails carry no alt text is the single most damaging finding. The product is invisible to the engines that decide who gets quoted when a director asks ChatGPT for a storyboard artist with named-brand commercial work. Every row below is verbatim from a live HTTP fetch of every URL in your sitemap, parsed for HTML, schema, headers, and payload size. Each row maps to a fixed milestone in the 90-day plan that follows.
| Metric | Current | Target | Severity |
|---|---|---|---|
| Pages with a meta description | 2 of 33 | 33 of 33 | Critical |
| Service pages with a H1 | 0 of 4 | 4 of 4 | Critical |
| Pages with duplicate H1 tags | / and /storyboards | 0 duplicates site-wide | Critical |
| Homepage portfolio image alt text | 2 of 16 | 16 of 16 | Critical |
| Heaviest page payload | 2.93 MB on /storyboards-for-music-videos | Under 500 KB | Critical |
| Cache-Control header | Forced to Expires: Thu, 01 Jan 1970 | Sensible max-age tuned per asset class | Critical |
| LocalBusiness schema opening hours | Malformed, fails the Rich Results test | Valid openingHoursSpecification array | High |
| Article schema publisher logo | publisher.logo.url missing | Absolute URL to a 600 × 60 brand PNG | High |
| NAP consistency | Cambridge address vs /storyboards page targeting London | One canonical NAP, geo-targeted pages declare areaServed | High |
| Schema types deployed | 0 of the 8 AI engines look for to cite a storyboard artist | All 8 deployed and validating | Critical |
| Transport security and compression | HSTS on, gzip on (the only two server-level wins) | Keep, document, add Brotli where supported | Low |
Methodology — live HTTP fetch of every URL in /sitemap.xml on 27 April 2026, full DOM parse, schema and structured-data inspection, robots and sitemap review, response-time and payload sizing across all 33 URLs. Numbers are verbatim from that crawl. The full audit log and per-URL findings are filed in the internal record.
Three waves. Ninety days. From invisible to cited.
Three parallel tracks, each with a deadline and a measurable acceptance criterion. Week one closes every audit row above — the foundation. Weeks one to three rebuild the on-page and schema layer that AI engines extract. Weeks three to thirteen turn the rebuilt site into a citation engine on the same brand-led queries directors and commissioners actually ask. The audit drives the urgency. The plan is the answer. Day 90 is the contract.
Wave 1 · Foundation
Week 1 to Week 2- Re-baseline every URL in PageSpeed Insights, verify Google Search Console and Bing Webmaster Tools, install heatmaps, snapshot a 28-day analytics baseline.
- Replace the broken Cache-Control header so the site stops telling browsers it expired in 1970. Done either through the host or fronted with a thin caching layer.
- Cut the /storyboards-for-music-videos page from 2.93 MB to under 500 KB. Same image quality, smaller bytes.
- Resolve the Cambridge address vs London targeting conflict. One canonical NAP, declared cleanly on every geo-targeted page.
- By close of week 2 — every server-level audit row green, the heaviest page light, the cache header sane.
Wave 2 · The cited layer
Week 1 to Week 3- Thirty-three unique meta descriptions, one per indexed URL. Each one written answer-shaped so it doubles as featured-snippet bait. Closes the 2-of-33 row.
- One keyword-led H1 on each of the four service pages. Removes the duplicate H1s on / and /storyboards. Closes the 0-of-4 and the duplicate-H1 rows.
- Descriptive alt text on every portfolio thumbnail. Each alt names the brand and the medium — Premier League title-sequence storyboard, Coca-Cola pitch frames, Bentley commercial animatic, NatWest broadcast, Innocent print, Wolt brand spot. Closes the 2-of-16 row.
- Eight new schema types deployed across the site so the engines can read who you are, what you make, and which brand each frame was made for. Closes the schema row.
- A new /faq page with twelve question-and-answer pairs that builds your first AI Overview footprint inside Google.
- By close of week 3 — every URL has a unique title, a meta description, exactly one H1, and validating schema. The site is legible to ChatGPT, Perplexity, Copilot and Gemini for the first time.
Wave 3 · The citation engine
Week 3 to Week 13- A pillar page — How storyboarding works: a director's guide. Two and a half thousand words. One named example per medium. Internal links to all four service pages.
- Four named-client case studies — Bentley, Premier League, NatWest, Innocent. Each runs eight hundred to twelve hundred words and embeds three frames from the actual job. The brands are already on your site as logos. We turn each one into a page directors can land on directly.
- A storyboard-vs-animatic-vs-concept-art comparison page. The question every junior commissioner asks. Your site does not answer it yet.
- The /storyboard-rates page rebuilt as a citation-ready data table. Brands brief who answers the price question first; right now your rates page is prose.
- An IMDb claim filed against the credit, then linked from /about so the engines anchor your name to the work.
- Tier-1 industry placement queued — Cartoon Brew, Skwigly, then Variety, Campaign and Little Black Book once the IMDb credit is live.
- A baseline AI-citation sweep across twenty-five target queries on ChatGPT, Perplexity, Copilot and Gemini in week four, re-run monthly.
- By close of day 90 — Seb Antoniou is the cited source in roughly one in three of the target queries, eight new long-form pages are live, and the citation rate compounds because the methodology is repeatable, not because we got lucky.
- Every URL has a unique title, meta description, exactly one H1, and validating schema.
- Eight new long-form pages live, including the four named-client case studies.
- Mobile speed score 90 or above site-wide; 95 or above on the four service pages.
- AI citation rate at or above thirty percent across twenty-five target queries on ChatGPT, Perplexity, Copilot and Gemini.
- Search-console impressions up at least eighty percent versus the day-zero baseline.
- At least one earned media link from a Tier-1 industry publication.
Most agencies still don't take generative-engine optimisation seriously in 2026. We do. A producer at home in the evening doesn't only google any more. They ask ChatGPT who is the best storyboard artist for a Premier-League-style commercial. When Seb Antoniou is the answer, you have already won the brief — before they ever reached Google.
Structured answers on every page
AI engines only cite what they can read cleanly. Clear definitions, question-and-answer blocks, and a crawl-friendly information architecture. Your existing page on rates, your service pages, and the new pillar piece all become extractable.
Author identity for Seb
The engines do not cite anonymous brands. They cite named people. Your face, your bio, your IMDb claim, and your portfolio of named-client commercial work get anchored into the page's data graph.
Provable claims, not promises
Storyboards for Premier League, Coca-Cola, Bentley, NatWest, Innocent and Wolt is a citable claim. Talented storyboard artist available for hire is not. Every claim on the rebuilt site is anchored to the brand it came from.
Clean web standards for AI crawlers
OpenAI, Anthropic, Microsoft and Perplexity all read the site from their first visit. The cache header, the schema graph, the sitemap, and the robots policy are tuned to let them through, then to let them quote you accurately.
Monthly AI-visibility check
We test a fixed list of twenty-five queries every month across ChatGPT, Perplexity, Copilot and Gemini and log when Seb Antoniou appears as a source. The number moves because the methodology is repeatable.
By day 90, the cited source in 1 in 3 of 25 target queries on ChatGPT, Perplexity, Copilot and Gemini. First hits realistically by week six.
What you own when this is over
Every line below is in writing on billyreiner.com and operationally enforced by the systems that run this company. The proposal closes on outcome and ownership. Pricing is its own conversation, and we will have it on the call — not on this page.
Any change, any page, in 24 hours, forever
Write, edit, replace, restructure, restyle. Submit a request, the change ships within 24 hours. Miss the window and the next billing period is on us. The guarantee is in writing on billyreiner.com and applies to every page in this proposal.
Continuous content production
One named-client case study, one educational pillar piece, two short blog posts every month. Researched against verified sources, edited through a seven-sweep framework, fact-checked, schema-marked, and shipped. Forty-eight new long-form assets in year one.
AI citations that compound across four engines
Every page rewritten answer-shaped. Citation rate tracked monthly across ChatGPT, Perplexity, Copilot and Gemini against the same twenty-five target queries from week four. The number moves because the methodology is repeatable.
Lead capture wired in
Booking slots embedded on the four service pages, conversion events on every CTA, heatmaps so you can see which paragraph closes the brief, and a lightweight CRM tag so the next Bentley-tier inbound is attributable to the page that earned it.
Monthly reporting in five minutes
One PDF and one short walkthrough video, the same day every month. Six numbers, three charts, and the next month's plan. No black box, no jargon.
You own the source code
If you ever leave us you take everything — source, schema, content, case-study assets. Hosted elsewhere for around five dollars a month if you choose. No retainer lock-in, no rebuild fee, no exit cost.
- 100/100 mobile PageSpeed at launch, or the launch balance is not owed.
- Live on day 7, or every day after is on us.
- Prototype before the full build. If the direction is wrong, full refund.
- Any change, any page, in 24 hours, forever. Miss the window and the next billing period is on us.
- Your published rates page lists projects in the £1,500 to £10,000 band. Mid-point sits around £4,000.
- One additional inbound per quarter that closes at £4,000 is four extra projects in year one — roughly £16,000 of incremental revenue, before we count the relationships that follow.
- The named-brand work that lives on your site already — Premier League, Coca-Cola, Bentley, NatWest, Innocent, Wolt — is the moat. Brands of that calibre brief by recommendation. Each new one is a multi-year relationship, not a one-off invoice.
- If the citation engine lands one Tier-1 brand inbound that you would not have had otherwise, the lifetime value of that one relationship outweighs the entire engagement by an order of magnitude.
The downside case is one extra project a quarter from a clean site. The upside case is the brand that asks because Perplexity, ChatGPT, Copilot or Gemini told them to. The math we will do together on the call. The page is what the math is built on.
Two ways this ends
Seb — 90 days. Three waves. From invisible to cited.
Pick a 30-minute slot below. We walk through the audit on screen, we pressure-test the plan against the 33 URLs the crawl found, and we agree what week one ships. The audit drives the urgency. The plan is the answer. The cost of the engagement is the conversation we have on the call — not the headline of this page.