A 500,000-part online store for a 70-year-old equipment dealer.
We rebuilt Scholten Equipment's parts store from the ground up: instant search across 500K+ parts and exploded-view diagrams, click-to-buy from the drawing, live two-store inventory, and real carrier shipping.
This video explains the Scholten Equipment website and gives a walkthrough of its parts-store experience, from equipment and model search to browsing the catalogue. It also demonstrates interactive exploded-view parts diagrams and the shopping cart, showing how customers move from finding a part to preparing an order.
Transcript
Visual walkthrough transcript — descriptions of the on-screen content, not a verbatim spoken narration. 00:00–00:05 — The Scholten’s Equipment, Inc. logo introduces the website showcase, with Built by DoodleWeb branding. 00:05–00:15 — “Find parts & equipment from the brands you trust” introduces the store. WordPress, WooCommerce, WP Engine and Typesense are shown as the technologies behind the website. 00:15–00:25 — The walkthrough demonstrates equipment and model search. Kubota tractors and model options appear, showing how customers narrow the catalogue to their equipment. 00:25–00:35 — Equipment categories include utility vehicles, mowers, tractors and construction equipment. 00:35–00:45 — “500,000+ parts, organized by equipment and category” introduces the parts catalogue, including belts, blades, filters, and tires and wheels. 00:45–01:00 — An interactive exploded-view diagram appears beside a parts list with inventory information. On-screen prompts explain “Scroll to zoom” and “Click a part # to view details,” demonstrating how the drawing helps identify the correct part. 01:00–01:05 — The shopping cart shows selected parts, including a washer and spring washer, with prices. 01:05–01:13 — The walkthrough closes with “Need a parts store like this?” and the booking address book.doodleweb.io.
- 500K+ parts & diagrams
- 546K diagram images
- 500+ CLAAS machines added
- 2 stores, 1 live inventory
Scholten Equipment is a family-owned farm equipment dealer in Lynden and Burlington, Washington. DoodleWeb built scholtenparts.com, a WooCommerce store on WP Engine with instant search, interactive exploded-view diagrams that let customers buy a part by clicking it, live two-store inventory and real carrier shipping. The store carries more than 500,000 parts and diagrams, serves 546,000 diagram images from a CDN, and added 500+ CLAAS machines automatically after launch.
What did DoodleWeb deliver for Scholten Equipment?
Scholten Equipment is a family-owned dealer in Lynden and Burlington, Washington, selling and servicing farm equipment (Kubota, CLAAS and more). Their parts business ran on phone calls and paper catalogues. We delivered scholtenparts.com: a WooCommerce store on WP Engine with over 500,000 parts and exploded-view diagrams, a search engine that answers as you type, interactive diagrams where customers click a numbered pin to add the exact part, inventory synced from the dealer's HBS business system, live UPS and USPS rates, Stripe payments, local pickup at either store, and SEO so the catalogue is discoverable. We also built the data pipeline that keeps the catalogue growing, including an automated import of the entire CLAAS parts catalogue.
Key facts
- Client: Scholten Equipment
- Location: Lynden and Burlington, Washington, USA
- Platform: WordPress / WooCommerce on WP Engine
- Catalogue size: 500,000+ parts and diagrams
- Launch year: 2026
- Services: Data audit, catalogue engineering, WooCommerce build, search, diagram viewer, ERP sync, shipping and payments, SEO
Last updated:
- Products
- 500K+
- Diagram images
- 546K
- CLAAS machines
- 500+
- Locations
- 2
Where they started.
Scholten had decades of parts knowledge locked in printed and PDF catalogues, so customers had to call just to identify a part. The catalogue data they did have was huge, duplicated and inconsistent: diagrams were copied eleven times per model, inflating a 20 GB database that crashed search and threw 502 errors. Product photos and diagrams lived on different systems. Two stores shared one inventory, but customers couldn't see it, and every shipping quote was worked out by hand.
What we built.
We rebuilt the catalogue on a lean model, one diagram shared by many models, and put Typesense instant search over roughly 700K documents. Interactive exploded-view diagrams use numbered pins that open the part and add it to the cart, and a part-to-diagram relation shows every drawing a part appears on, with images served from S3 and CloudFront. HBS inventory sync shows combined stock for both stores, alongside live UPS and USPS Ground Advantage rates, Stripe checkout and local pickup with clear pickup-confirmation wording. Location pages, SendGrid transactional email and a launch SEO layer completed the build.
Where it landed.
The store launched with the full catalogue searchable. Customers find a part by number, model or keyword in milliseconds and buy it straight from the drawing, while the team receives orders with the right pickup store or carrier label. The catalogue keeps growing automatically: the CLAAS line was added after launch without any manual data entry.
The work behind the launch.
Each engagement covers strategy, design, build, and post-launch care — orchestrated by a single senior team.
Barriers we had to overcome.
A concise look at the constraints that shaped every decision in the build.
A database eleven times bigger than it needed to be
Duplicated diagrams per model inflated the site to 20 GB and search ran out of memory. We de-duplicated to one diagram shared across models and cut the index to roughly 700K lean documents.
Hosting limits on big queries
The host kills any query over 16 KB, so bulk operations had to be chunked and term counting deferred. We built our own keyed maintenance endpoints for audits, cache repair and category fixes.
Stale caches hiding good data
The persistent object cache showed intact diagrams as empty products ("Price on request"). We built an audit tool that compares the database with what the site serves and repairs the difference.
A catalogue that only existed inside a manufacturer portal
CLAAS parts live in a session-based web app with no export. We built a Playwright pipeline that walks the catalogue, captures every drawing with its part positions, and publishes it through WooCommerce with prices from the dealer's price file.
What we built.
The pieces that shaped the final build — each tied back to a real business outcome.
Instant search
Results appear as customers type a part number, model or keyword, so they find the right part without calling the counter.
Interactive exploded-view diagrams
Numbered pins on every drawing open the exact part and add it to the cart, turning the catalogue itself into the checkout path.
Part-to-diagram relations
Every part page shows the drawings it appears on, so buyers can confirm fit before they order.
Live two-store inventory
Stock synced from HBS shows availability across Lynden and Burlington, so customers know what they can get today.
Real carrier rates and local pickup
Live UPS and USPS rates replace hand-written quotes, and local pickup routes each order to the right store.
Automated catalogue pipeline (CLAAS)
A repeatable import adds whole manufacturer catalogues without data entry, so the store grows as the dealership does.
What changed.
What changed for the team and their audience after launch — measured in real behavior, not vanity metrics.
- Customers self-serve part identification
- Orders arrive with the correct pickup store or carrier
- Both stores' stock is visible online
- Catalogue grew by 500+ machines after launch with no data entry
- Site stays healthy at 500K+ products
- Emails reach the inbox from the store's own domain
Tech stack & delivery.
Inside the build.
A few selected screens from the live product.
I honestly couldn't believe all of this was possible. Our entire parts catalogue, every diagram, every part number, is now online and searchable, and customers can buy straight from the drawing. How professionally and how fast DoodleWeb launched this store is mind-blowing.
Numbers we are proud of.
Verified outcomes and delivery details from the engagement.
| Label | Value |
|---|---|
| Parts & diagrams online | 500K+ |
| Diagram images served from CDN | 546K |
| CLAAS machines added automatically | 500+ |
Frequently asked questions.
Can WooCommerce handle 500,000 products?
Yes, with the right data model and search layer. We de-duplicated diagrams so one drawing serves many models, moved search to Typesense, served images from S3 and CloudFront, and chunked bulk operations to stay within hosting limits. Scholten's store runs healthily at more than 500,000 products.
How does the diagram click-to-buy work?
Each exploded-view diagram is shown in an OpenSeadragon viewer with a custom overlay of numbered pins. Clicking a pin opens the matching part and lets the customer add it to the cart directly from the drawing.
How is inventory kept current across two stores?
Inventory is synced from the dealer's HBS business system, and the store shows combined stock for the Lynden and Burlington locations so customers can see availability before ordering or choosing pickup.
How do you get live UPS and USPS rates?
Checkout calls the UPS and USPS APIs, including USPS Ground Advantage, to quote real carrier rates for each order instead of hand-calculated shipping.
How did you import a manufacturer catalogue with no export?
CLAAS parts only existed in a session-based web portal. We built a Node and Playwright pipeline that walks the catalogue, captures every drawing with its part positions, and publishes it through WooCommerce using prices from the dealer's price file. More than 500 machines were added this way.
What keeps search fast at this scale?
A dedicated Typesense index of roughly 700K lean documents, built after de-duplicating the catalogue, returns results as customers type.
Can other dealers reuse this?
Yes. The catalogue model, diagram viewer, inventory sync pattern and import pipeline are built to be repeatable for other equipment dealers with large parts catalogues.
How long did the build take?
The engagement ran across 2025 and 2026, covering discovery, data engineering, the store build, launch and the ongoing catalogue pipeline.