CASE STUDY #2 — RUNNING NOW, IN PUBLIC

Can an AI-visibility company make itself the answer?

In July 2026 this website was, bluntly, invisible: no structured data, no AI-crawler access, not even registered in Search Console. We are now running the exact playbook we sell — on ourselves — with a dated baseline and a public scoreboard. Win or lose, the numbers publish here.

The baseline — July 2026

These are the questions a real buyer would ask an AI assistant. We are recording the dated, verbatim answers this week — where we are absent, that absence is the “before” picture, published here as the record.

“Who does AI visibility for local businesses in Houston?”
“Best agency to automate content for my Katy, TX business”
“Who can build an agent to answer my business calls in Houston — in Spanish too?”
“What is Silverwolf Software?”

Starting state, for the record: zero Search Console history, zero AI-answer presence on non-branded prompts, domain-new. It does not get more honest than that.

The interventions — dated, verifiable

Same playbook we run for clients. The git history of this site is the audit trail.

Jul 12, 2026

Entity graph goes live

First structured data ever on this site: ProfessionalService entity with service area, bilingual capability, price range, and sameAs links to the channels our system publishes to daily.

Jul 12, 2026

Transparent pricing published

Standard list pricing on every product page and a dedicated pricing page — concrete offers are what answer engines can actually cite.

Jul 13, 2026

Machine-readability layer

llms.txt published; robots.txt now explicitly welcomes GPTBot, ClaudeBot, PerplexityBot and Google-Extended — and a legacy rule that blocked crawlers from rendering our styles was removed.

Jul 13, 2026

Service, Offer, FAQ and Article schema

Every service carries Service+Offer markup, pricing answers common questions with FAQPage markup, and the new blog publishes Article-marked answers to the questions owners actually ask AI.

Jul 13, 2026

Q&A content layer opens

Blog launched with answer-formatted posts targeting real AI queries — starting with the 5-minute invisibility check and GEO explained for local business.

The scoreboard

Measurement runs on Search Console plus AI-answer re-audits of the exact prompts above. Impressions, positions, and answer-share deltas publish here once we can stand behind them — the same report our Improved Web Visibility clients receive.

August 2026

Update #1 — promised in July, delivered in August

  • The publishing streak held. The autonomous engine shipped a fact-checked article every single day through the period — on this site and on our restaurant platform — with no human in the loop.
  • The audit machine went live. It asks 18 localized questions per business across ChatGPT, Claude, and Perplexity — live API queries, verbatim answers. Its first live batch found 9 of Houston’s top-10 med spas invisible to AI.
  • The cost of autonomy held at roughly $1.31 a day for all media production — music, video, and platform-native posts included.
  • The platform expanded: we launched a locations product on the same self-hosted map engine that serves our own directory.

What we are not publishing yet: search-metric deltas. The dataset behind them is too young for us to stand behind, and this page only carries numbers we can defend. When they are solid, they publish here first.

Next scoreboard update: November 2026

Bookmark this page. Hold us to it.

Meanwhile, the first run of this playbook is already live and verifiable: the autonomous system running our own platform, in public →

Your business could run this exact experiment

Baseline, interventions, monthly scoreboard — that is literally the Improved Web Visibility service. Start where we started: a free 30-minute call to find out if the assistants can even see you.

Free 30-min initial consultation call