# AI Visibility Foundation Fix — $1,997 USD flat, 14 days — Foreground Digital > Productized AI Visibility Foundation Fix. 10 deliverables across 14 days: 5-engine baseline audit, JSON-LD entity graph, Wikidata QID, schema deployment, AI crawler allowlist, llms.txt, Knowledge Graph submissions, brand-disambiguation content, re-baseline delta report, 30-day monitoring. Flat $1,997 USD. Source: https://foreground.agency/ai-visibility-fix/ --- AI Visibility Foundation · Flat $1,997 USD # Your brand is invisibleto ChatGPT.We fix the foundation in 14 days. When buyers ask Claude, ChatGPT, Gemini, DeepSeek, or Perplexity "who do you recommend for X?" — your competitors are named, and you aren't. The cause is rarely your content. It's that AI engines can't identify your brand as an entity. This is the foundation fix: structured-data entity graph, Wikidata QID, sameAs network, AI-crawler access, schema deployment, brand disambiguation. Ten deliverables, fourteen days, one flat fee. [Book the fix →](https://foreground.agency/contact?service=ai-visibility-fix) 20-min intro call · Free diagnostic first 01 — Why this matters ## AI engines pickentities, not pages. ChatGPT has 900 million weekly users (OpenAI, Feb 2026). Gemini has 400 million MAU. Together with Claude, DeepSeek, and Perplexity they capture ~94% of worldwide generative-AI traffic (Similarweb Global AI Tracker, Jan 2026). When buyers research vendors, they ask these engines first — and the answers shape every shortlist downstream. The mistake most agencies make: they treat AI visibility as a content problem. Write more blog posts, target more keywords. But AI engines don't index pages the way Google does — they identify entities via structured data, sameAs networks, Wikidata claims, and knowledge graph membership. If your brand isn't a recognized entity, no amount of content gets cited. The foundation fix: establish your brand as an unambiguous entity across the data sources LLMs are trained on. JSON-LD graph with stable @id. Wikidata QID. sameAs network linking LinkedIn, Crunchbase, GitHub, ORCID. AI-crawler accessibility. The same upstream signals that earned media and review platforms reference when they cite you — which is where the 94% of citations actually come from (Muck Rack, Dec 2025). Foreground Digital has scored 0/100 AIVS across all five engines for two consecutive weekly snapshots (May 4 and May 11, 2026). We've been logging the baseline for a week — and we're shipping this exact playbook on our own brand publicly, with every weekly check logged. The methodology is the product. 02 — What's included ## 10 deliverables,each verifiableon completion. Same 5-engine baseline our free AI Brand Check runs — only this time we do the implementation, then re-run it to prove the foundation moved. 01 ### 5-engine baseline audit AI Brand Check across Claude, ChatGPT, Gemini, DeepSeek, and Perplexity — 10 buyer-intent prompts, 3 runs each at temperature 0.2 for stable results. Per-engine scores, position, sentiment, and the brands cited in your place. Becomes your before/after benchmark. 02 ### JSON-LD entity graph with stable @id Three-layer Organization schema with a stable @id URI. Comprehensive sameAs array to LinkedIn, Crunchbase, GitHub, ORCID, Wikidata, and Wikipedia where applicable. This is how AI engines disambiguate "Foreground Digital" from "Foreground Films" and treat your press, blog, and team pages as one entity. 03 ### Wikidata QID + entity claims Create or merge a Wikidata QID for your organization. Add and source P31 (instance of), P856 (official website), P571 (inception), P159 (HQ location), P452 (industry), founder, key people. Wikidata is in every major LLM training corpus — Q-numbers are the single most reliable way to be recognized as a real entity. 04 ### Schema deployed on 10-25 top pages Organization, Service, FAQ, HowTo, Article, LocalBusiness, and Product schemas as applicable. Validated against Schema.org, Google Rich Results test, and Bing Webmaster Markup Validator. We deploy directly to your CMS or hand off the JSON-LD blocks. 05 ### AI crawler allowlist robots.txt configured to explicitly allow GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, and CCBot. Most "we have no AI visibility" diagnoses come back to this single line. We test each crawler against your live site post-deploy. 06 ### llms.txt + .md twins for top 20 pages Published llms.txt at /llms.txt following the llmstxt.org spec, with markdown twins of your 20 highest-value pages at /pages/.md. Note: llms.txt has no documented citation correlation yet (SE Ranking 300k-domain study, 2025) — we ship it for hygiene and future-proofing, not as the differentiator. 07 ### Knowledge Graph submissions Submit your organization to Google Knowledge Graph (via structured data + GBP claim) and Bing Knowledge Graph. Where eligible, we register your dataset to Google's Dataset Search. These signals propagate into AI training data on the next refresh cycle. 08 ### Brand-disambiguation pages About/Team/Founders pages rewritten as entity-anchor content with dated press, structured bios, and explicit category positioning. The "Foreground Digital is a Dubai-based AI search optimization agency founded in 2024 by ___" sentence is what LLMs latch onto when training new model versions. 09 ### 14-day re-baseline + delta report Re-run the 5-engine audit at day 14. Side-by-side comparison report showing score movement per engine, new mentions captured, and cocitation shifts. Hand-off PDF documents every artifact we deployed with file paths and verification links. 10 ### 30-day post-handoff monitoring Weekly auto-check for 30 days after delivery. Email alert if any score drops more than 10 points. Gives you a month of free visibility into whether the work is sticking — useful for catching regressions caused by other website changes. 03 — The 14-day method ## Six phases,fourteen days,nothing hidden. We diagnose before we deploy. Every phase produces a verifiable artifact — JSON-LD blocks, the Wikidata Q-number, the updated robots.txt, the validated schema, the re-baseline report. You see the work. - Day 0 ### Intake call (free, 20 minutes) We agree on scope, you grant access to your site CMS, hosting panel, Google Search Console, and Bing Webmaster Tools. You pay the $1,997 fee. We schedule the kickoff. - Days 1-2 ### Phase 1 — Baseline & diagnosis Run the full 5-engine × 10-prompt × 3-run AI Brand Check. Map score gaps to specific deliverables. Identify the prompts where your competitors are winning and why. Output: 8-page baseline report with prioritized fix list. - Days 3-6 ### Phase 2 — Entity foundation Build the JSON-LD entity graph. Create the Wikidata QID and source every claim. Deploy schema on top pages. Stand up the sameAs network across LinkedIn, Crunchbase, GitHub, and ORCID. Validate every deployment. - Days 7-8 ### Phase 3 — Crawler access Update robots.txt with explicit allow directives for all 6 AI crawlers. Deploy llms.txt and the .md twins. Submit to Google and Bing Knowledge Graphs. Verify each crawler can actually fetch your pages (live test, not assumption). - Days 9-12 ### Phase 4 — Disambiguation content Rewrite About/Team/Founders pages with entity-anchor structure. Add dated press references. Embed the explicit category positioning sentence LLMs use to identify your brand. - Days 13-14 ### Phase 5 — Re-baseline & handoff Re-run the 5-engine audit. Generate the delta report. PDF handoff documenting every file deployed, every URL submitted, every verification check. Schedule the 30-day check-in. 04 — Why this works ## Sourced.Not opinion. Most AI-visibility agencies sell theatre. We cite the published research that informs every deliverable — so you can verify the methodology yourself before you book. 94% ### AI citations from earned media Muck Rack, Dec 2025 Brand-owned content alone almost never gets cited. The Foundation Fix establishes the entity signals — the upstream work that earned media can hook onto. 5× ### Earned vs owned content citation ratio University of Toronto study AI engines weight third-party mentions far higher than your own copy. Why we focus on entity foundations first — they're what third-party sources reference. 46.5% ### Perplexity citations from Reddit ZipTie analysis, Jan 2026 Different engines pull from different sources. Foundation Fix maps your per-engine citation surface so you know exactly where the gap is. 7.8% ### ChatGPT citations from Wikipedia Frase / aggregated data Single largest source after general web. Wikidata QID (deliverable 03) is the gateway — Wikipedia article eligibility is downstream of notability, which we assess but never promise. 05 — Pricing ## One scope.One fee. Global benchmark for productized AI-visibility work is $1,500-$7,500 USD (WebFX, Stackmatix, Digital Elevator agency surveys, 2026). $1,997 sits at the floor — concrete deliverables, no hourly creep, no surprise invoices. $1,997 USD · flat · paid upfront ≈ AED 7,330 · ≈ SAR 7,490 · ≈ EUR 1,830 · ≈ GBP 1,580 Full AI Visibility Foundation Fix. 10 deliverables across 14 days. 30-day post-handoff monitoring included. - ✓ All 10 deliverables above, every artifact verifiable on completion - ✓ Day-1 baseline + Day-14 re-baseline, side-by-side delta report - ✓ 30 days of weekly auto-checks post-handoff with score-drop alerts - ✓ PDF handoff documenting every URL, file path, and verification check - ✓ Deliverable guarantee — every artifact live + verifiable, or full refund What this is NOT A score-move guarantee. AI engines update training data on their own schedule. We commit to deliverables, not outcomes — and we're explicit about that because anyone promising a specific AIVS lift below $10K USD is misrepresenting the work. Continuation engagements (digital PR, Reddit seeding, dataset publication, content velocity) are offered separately once foundation work completes. [Book the fix →](https://foreground.agency/contact?service=ai-visibility-fix) 06 — Eating our own dog food ## We're runningthe playbookon our own brand. We've been logging Foreground Digital's own AIVS for two weekly cycles — 0/100 on May 4, 2026 and 0/100 again on May 11, 2026, across all five engines. Zero industry-prompt mentions in 60 responses. Recognized as a brand (echo rate 9/9 on most engines) but never recommended. Every weekly score is logged. Every deliverable we deploy on our own brand is documented. The methodology you'd buy is the methodology we're using on ourselves — verifiably, in the open. Why this matters for you: most agencies pitching AI visibility have never moved their own score. We're shipping the work first, with all the metrics visible. Your engagement starts with the same baseline audit, the same playbook, the same accountability. 07 — FAQ ## Honestanswers. ### I already have schema, llms.txt, and allow GPTBot. Will this still help? + Honestly — partially. The Foundation Fix is for brands that haven't yet established the basics. If you're already past that, you're ready for the next-tier work: digital PR placement (HARO/Featured.com pitches to outlets that AI cites), Reddit/Stack Overflow evergreen content seeding, G2/Capterra listing optimization, and first-party dataset publication. That's the continuation engagement we offer to graduates of the Foundation Fix. The 5-engine baseline audit we run on Day 1 will tell us honestly whether you need Foundation or the continuation track — and if it's the continuation track, we tell you and quote against that scope, not this one. ### Why a flat $1,997 and not hourly? + Because the work is well-scoped: 14 days, 10 deliverables, every artifact verifiable on completion. Hourly billing rewards slow consultants. We charge a flat fee because we've done this enough times to know what it takes — and because we're shipping the same playbook on our own brand publicly. Foreground Digital scored 0/100 AIVS on May 16, 2026. We're documenting our own move from 0 to 30+ as a public case study. The methodology is the product. ### Will my AIVS score actually move? + We guarantee deliverables, not outcomes — and we're honest about why. AI engines update training data on their own schedule. Score movement typically appears within 14 days for AI Overviews and Perplexity (live-search engines), 30-60 days for ChatGPT, and 60-90 days for Claude and Gemini (which rely more heavily on training data refresh cycles). Industry-wide, vendor-published case studies show 15-40% AIVS lift in 60-90 days after the foundation work. Anyone promising a specific score move below $10K is either lying or doing junior-level work — published skeptics at Digiday and Krasia have documented this. We commit to every deliverable in the scope above, verifiable on completion. If even one is missing or broken, full refund. ### Why not the cheaper "AI visibility" services I see online? + The 2025-2026 GEO market has stratified. Below ~$1,500 USD you're mostly buying AI-generated checklists and one-shot audits with no implementation. The $1,997 Foundation Fix sits at the global floor of the productized range ($1,500-$7,500 per WebFX, Stackmatix, and Digital Elevator agency benchmarks). It includes the actual implementation work, not just the diagnosis. The mid-market retainer band ($3,000-$10,000/mo) is for continuous monitoring — different product. ### What if you can't create a Wikipedia article for me? + Wikipedia notability is determined by significant coverage in independent reliable sources — it's outside our control. We do not promise Wikipedia article creation at this tier. What we do promise: a Wikidata QID (which we can always create — Wikidata has lower thresholds and is more useful for AI training data anyway), and a notability assessment that tells you honestly whether Wikipedia is achievable. If it is, that's a separate continuation engagement. If it isn't, we tell you and focus the work elsewhere. ### Can I do this myself? + Yes — every tactic in the Foundation Fix is documented publicly (Schema.org, Wikidata, llmstxt.org, robots.txt specs). The reason clients pay $1,997 is the same reason they pay any specialist: avoiding 60+ hours of senior engineering time figuring out which specific JSON-LD properties matter, which sameAs platforms LLMs actually weight, which Wikidata claims to source first, and how to validate each deliverable end-to-end. If you have a senior engineer with 60 spare hours, you can ship this yourself. We typically deliver it in 14 days because we've done it before. ### What's NOT included? + Digital PR (HARO/Qwoted/Featured.com placement), Reddit/Stack Overflow content seeding, G2/Capterra listing optimization, original research/dataset publication, Wikipedia article creation, ongoing monitoring beyond day 30, and long-form content production. These are higher-tier engagements offered as continuation work after the Foundation Fix completes. Our research shows 94% of AI citations come from earned media — so if your foundation is already strong, that earned-media layer is where the next investment goes, not more foundation work. ### How do I know if I need this? + Run our free AI Brand Mention Checker first — no signup beyond email. If your AIVS is below 30, you almost certainly need the Foundation Fix. If it's 30-60, you're a candidate. If it's 60+, you probably need the continuation engagement (digital PR, Reddit, datasets), not foundation work. The free check is the no-cost diagnostic. We never quote until we've seen the score. ## Find out where you stand first.The 5-engine check is free. We never quote without seeing your score. Run the free AI Brand Mention Checker first. If you're below 30/100, the Foundation Fix is for you. If you're above 60, you need the continuation track — we'll tell you which when we see the data. [Run the free check →](https://foreground.agency/tools/ai-brand-check) [Book the fix](https://foreground.agency/contact?service=ai-visibility-fix)