A manufacturer selling through an OEM channel decided to open a direct channel from scratch, with no storefront and no organic history to inherit. The catalog ran to 22k SKUs. Pricing was multi-currency across several tax regions. And the product data lived where it belonged, in the Acumatica PIM, which wasn’t going anywhere.
Anonymized · NDA
Cited in AI Overviews 12 weeks after launch.On a domain with zero history.
A fluid power manufacturer opening a direct commerce channel had no storefront, no organic history, and one hard rule: the Acumatica PIM stays the source of truth. Five months later a 22k-SKU Next.js + Saleor build was live. It shipped AIO-ready, and earned its first AI Overview citations within 12 weeks of launch.
- Client
- Fluid power manufacturer · OEM channel
- Scale
- 22k SKUs
- Engagement
- 2025 · 5 months
22k
SKUs on a greenfield Next.js + Saleor build
5 mo
Kickoff to launch
12 wk
To first AI Overview citation after launch
1
Source of truth — the Acumatica PIM, wired straight through
The situation
Where they started.
The constraint
What made it hard.
A greenfield domain starts with nothing engines trust. Every citation heuristic that rewards incumbents — history, links, behavioral signals — worked against this launch. The PIM-as-source-of-truth rule ruled out a copy-paste catalog, too: if the product data ever forked from Acumatica, the project failed on its own terms no matter what traffic did.
What we ran
The work, phase by phase. What shipped, and when.
Saleor backend wired to Acumatica
The PIM stays canonical. Product data flows from Acumatica into Saleor, so the catalog never forks — 22k SKUs with one source of truth.
Next.js storefront on Vercel Edge
Server-rendered storefront with multi-currency and multi-tax-region handling from day one.
AIO-ready at launch, not retrofitted
Schema graph, direct-answer category structure, and machine-parseable spec tables shipped with the first deploy.
This is what a Website Development engagement at this scale looks like in practice.
Why it worked
The mechanism. What actually moved the number.
A greenfield domain can’t win citations on authority, so it has to win on legibility. The engines’ first crawl saw the catalog exactly the way they like it: structured specs, answer-shaped category pages, a complete schema graph. That’s why the first citations came in weeks instead of quarters. There was no remediation phase, because nothing shipped broken.
What happened
The results. As measured, dips included.
The build went live at month five. 22k SKUs, PIM wired straight through.
The first AI Overview citations on tracked category queries landed inside 12 weeks of launch, on a domain less than half a year old with no link history to lean on.
The number that mattered
12 wk
Time to first AI Overview citation, post-launch
Greenfield build · zero pre-launch organic history
Measurement notes
How these numbers were measured.
A metric without a source is an assertion. Every number on this page is listed below with how it was counted.
- Time to first AIO citation
- First observed AI Overview citation of the new domain on tracked category queries, measured from launch day via the Salesolution citation tracker.
- Scope & stack
- Verbatim from the launch retro: greenfield build, 22k SKUs, 5 months, Next.js + Saleor on Vercel Edge, Acumatica PIM integration, multi-currency / multi-tax-region.
Anonymity. Client name withheld under NDA. Most industrial distributors prefer the case study without the logo. The vertical, scale, timeline, and every number are exactly as measured. Only the identifying details are removed. Reference calls available on request.
More case studies
Other distributors. Same standard of proof.
Industrial hydraulics distributor~8,500 SKUs
Anonymized · NDA+43.5%
Qualified leads per month, on the August baseline
Aug 2024 – Jan 2025 · client CRM
1,840 to 2,640 qualified leads a month. No new ad spend.
An ~8,500-SKU hydraulics distributor was stuck at 1,840 qualified inbounds a month. AI Overviews were citing its manufacturers, not its own category pages. Six months of catalog and AI-search work later, January closed at 2,640. That’s +43.5% over the August baseline, and we didn’t add a dollar of ad spend.
Catalog AIAI Search & GEOAug 2024 – Jan 2025Full engagement
Industrial hydraulics distributor~8,500 SKUs
Named · with consent+43.5%
Qualified leads a month, across the full engagement
2022 – 2025 · client CRM
1,840 to 2,640 qualified leads a month. The full-stack rebuild behind it.
Northern Hydraulics is an industrial hydraulics distributor — ~8,500 SKUs of JIC/NPT fittings, adapters, and assemblies sold to engineers and MRO buyers. Over a three-year relationship (2022–2025) we ran the whole growth function: brand and design, a Magento 1 to headless replatform, AI-search and technical SEO, an AI-assisted catalog rewrite, pillar content, paid ads, and cold outbound. Qualified leads went from 1,840 to 2,640 a month, up 43.5%.
Full Growth OwnershipWebsite DevelopmentAI Search & GEOEditorial AuthorityOutbound Email2022 – 2025Industrial hydraulics distributor~8,500 SKUs
Named · with consent8,500
SKUs replatformed off Magento 1
6 months · schema-complete from day one
8,500 SKUs off Magento 1, onto a storefront AI can read.
Northern Hydraulics ran an 8,500-SKU catalog on Magento 1, which hit end-of-life with no security patches and product markup AI search engines couldn't parse. Over 6 months we replatformed the whole catalog to a headless Next.js storefront on Shopify Hydrogen, schema-complete from day one, and carried the JIC/NPT spec quote flow across intact. Mobile category-browse INP dropped from 600ms+ to under 200ms. This was the web-dev cut of the full engagement; the qualified-lead growth is reported on the anchor study.
Website DevelopmentAI Search & GEO2022
Your catalog, same standard
Want a number like these for your catalog?
Held to the same standard as the studies above: we measure where you stand today — organic and AI-search visibility across your top categories — and show you the baseline before we propose a thing.