Power equipment, industrial tools, parts and attachments — catalogs where a wrong answer means a returned machine. HOLLOU answers spec, fitment and parts questions from your own catalog, says so plainly when your catalog can't answer, and turns the buyer into a lead or a checkout instead of a bounce.
Operators and trade buyers ask precise questions and distrust vague answers — because for them a wrong answer isn't a bad experience, it's a freight return and a job on hold. A search box turns “does this fit my model?” into forty SKUs and no answer. A generic chatbot is worse: it improvises a spec, and you eat the return.
The number exists — on a spec table halfway down a product page, or in a PDF. The buyer won't guess, and shouldn't have to dig.
Compatibility lives in tables and footnotes. Get it wrong and the dealer pays twice: the return, and the buyer's trust.
The buyer knows their machine, not your SKU scheme. If they can't get from machine to part in one step, they call — or leave.
HOLLOU answers only from your own site and catalog — product pages, spec tables, compatibility data, stock and price. It doesn't fill gaps with general knowledge, because in this category a plausible guess is the most expensive kind of wrong.
Working pressure, flow, capacity, weight, power draw — retrieved from your catalog at the moment of reply, not written by the model.
It asks the machine and model year before recommending, answers from your compatibility tables, and flags conflicts instead of selling around them.
When your catalog doesn't carry the answer, it doesn't improvise. It says so — and offers to get the answer from your team, capturing the buyer's contact, machine and question as a lead.
High-value units and volume orders become structured quote requests for your sales team. Stocked, priced parts go straight to add-to-cart and your own checkout.
Modeled on a composite equipment catalog to show the behavior — the products and figures are illustrative, not a customer's results.
The agent quotes the figure from the product's spec table — 150 lb, as published — and links the product page it came from, so the buyer can verify it themselves.
It asks the make and model first, then answers from the attachment's compatibility data: universal quick-attach, requires 21+ GPM hydraulic flow — and tells the buyer to confirm their machine's flow against its spec plate, because that number isn't in the dealer's catalog.
It resolves the machine to the part your catalog cross-references, confirms in stock, and puts it in the cart — one exchange instead of a phone call.
Volume and freight pricing aren't published in the catalog, so it doesn't invent them. It says so, and captures a structured quote request — quantity, destination, contact — for your sales team to price.
The agent installs as a single script tag on any website and runs alongside your existing store — your product pages, your cart, your checkout all stay exactly where they are. Removing it is one line too. Equipment is one of the categories where we think this earns its place; the others are on use cases, and how engagement is priced is on plans.
From your own site and catalog only — product pages, spec tables, compatibility data, stock and price. The agent is catalog-grounded: it retrieves before it answers, and it doesn't fall back on the model's general knowledge when your catalog is silent.
It tells the buyer plainly that the spec isn't published, and offers to get the answer from your team — capturing their contact, machine and question as a lead. Every one of those refusals is also logged for you, so you can see exactly which specs your catalog is missing.
Yes. High-value units and volume orders become structured quote requests routed to your sales team, with quantity, destination and contact details attached. Stocked, priced parts and accessories go to add-to-cart and your own checkout.
No. The agent works from your site as it stands — Shopify has a direct sync, and other platforms work through feed or API ingestion. The honest caveat: answer quality tracks what you've published. If a spec isn't on your site anywhere, the agent will refuse rather than guess, and the gap report tells you what to add.
A 30–60 day pilot on a controlled slice of your traffic, with success criteria agreed in writing before launch and no long-term contract. If it doesn't show what we both hoped, it ends — removing the agent is one line.
Send us your store URL. We'll index a slice of your catalog and show you the agent answering real spec and fitment questions about your actual products — before you commit to anything.
No catalog credentials needed for the first preview