A US industrial paint and safety supply store with tens of thousands of SKUs was invisible where it mattered: average Google position 22.0, ranking below distributors reselling the same catalogue. Work began 1 September 2025. Twelve months later the site sits at an average position of 9.8, holds #1 in the United States on a long list of product and brand terms, and is being surfaced by AI assistants that sent it no traffic at all a year ago.
| Query (United States) | Average position | Type |
|---|---|---|
| strobel supply | 1.0 | Brand |
| strobels supply inc | 1.0 | Brand |
| buy glasurit paint online | 1.0 | Commercial intent |
| rational r95-clx | 1.2 | Product / part number |
| 3m 9699 | 1.3 | Product / part number |
| 3m 33493 | 1.4 | Product / part number |
| rational r95 clx filter | 1.5 | Product / part number |
| 3m 33478 | 1.6 | Product / part number |
| r95-clx | 1.8 | Product / part number |
| axalta flex additive | 1.8 | Brand product |
| strobels supply | 1.9 | Brand |
| futuro for her wrist brace | 1.9 | Brand product |
| 3m 8949 | 2.1 | Product / part number |
| rt09c16g20nn | 2.1 | Product / part number |
| 3m 2123 | 2.5 | Product / part number |
| imron industrial paint | 3.7 | Category head term |
| best lead test kit for dishes | 4.0 | Informational head term |
| nason paint | 5.6 | Brand head term |
| imron paint | 6.1 | Brand head term |
| glasurit paint | 6.5 | Brand head term |
Sample of queries from Google Search Console, United States, 28-day window ending 3 September 2026. Positions are Google's own averages over the period, not a single-day check.
Product and category pages were rebuilt so a model can actually parse them: complete Product, Offer and Organization schema, real specification tables instead of manufacturer PDFs, explicit part numbers and cross-reference numbers in the copy, and answer-shaped content for the questions buyers actually type.
On top of that: clean crawl paths for AI crawlers, canonical entity pages for each brand the store carries, and internal linking that makes the relationship between a brand, a product line and a part number unambiguous.
The effect is that when someone asks ChatGPT, Gemini or another assistant where to buy a specific industrial part, the store is one of the sources that comes back, and the visit lands as a buyer rather than a browser.
Source: Google Analytics 4, session primary channel group, 8 Aug – 4 Sept 2026 vs the same period in 2025.
Faceted-navigation traps, parameter duplicates and orphan pages were cleaned out so Google spends its crawl on pages that can rank. Core Web Vitals, canonical logic and sitemap hygiene were rebuilt at the same time.
Thousands of SKUs carried the manufacturer's boilerplate, identical to every competitor selling the same part. Priority pages were rewritten with real specifications, cross-reference numbers, use cases and complete structured data.
Buying guides and comparison pages built around what customers actually search, from lead test kits to automotive paint systems, feeding internal link equity into the commercial pages behind them.
Each brand the store carries was given a proper hub page, linked consistently across the catalogue, so Google and AI models both understand what the site is an authority on.
Bot and scraper traffic blocked, GA4 re-tagged, channel grouping repaired and Search Console reconciled with analytics so every decision after that was made on real numbers.
Schema, entity pages, answer-shaped copy and clean access for AI crawlers, which turned into a measurable assistant-referred traffic channel that did not exist twelve months ago.
I take on a small number of SEO programmes at a time, usually for e-commerce catalogues and B2B sites where the products are real and the competition is selling the same thing. You get the technical work, the content, the AI-search layer and a monthly report with the actual Search Console numbers in it, not a dashboard of invented scores. Reply within 12 hours, usually two.