A machine dealer sells from a catalogue that was already out of date the day it went to print, and from a sales agent who answers "I will check and come back to you" three times in one meeting. The third demo on my automation page is now live, and it is built for exactly that conversation: a sales catalogue for a machine dealer, running in the browser, with seven machines in it.
It is deliberately fictional. The brand is invented, the machines are invented, the photos are generated, and the demo says so on its own banner. That way I can show the whole thing to anybody without touching a real dealer's range, prices or customer list.
What a buyer does in it
The catalogue opens as an application, not as a page you scroll. The machine photo takes the whole stage, a strip along the bottom switches between the models, and everything else sits in panels around it.
- Add-ons behave like add-ons. They are grouped the way a salesperson talks about them: clamping and handling, tools, software and service. Switch one on and it appears on the machine photo, and the weight and the floor space the machine needs are recalculated with it.
- Numbered points instead of a data sheet. The main parts of the machine are marked on the photo, each with a short explanation and a close-up, so a buyer learns the machine by looking at it rather than by reading a table.
- "Will it fit?" answered on the spot. Type in the length and the width of the hall and the catalogue draws a plan to scale: the machine, the add-ons that were switched on, the service clearance around it and a person for scale. The verdict is a sentence, not a diagram to interpret: it fits, or it is short by this much, service clearance included.
- One request for several machines. Machines go onto a request list with the options already chosen, and that list leaves as a single enquiry. Each machine also has a full page of its own, with a gallery, the technical data and a form.
Why this is not just a prettier PDF
The point is not the animation. It is that a dealer keeps one source of data, and everything the customer sees is built from it: the catalogue on the website, a PDF generated on demand that is current the moment it is downloaded, and an offline version on a tablet for a fair stand where the Wi-Fi never works. A price or a specification changes in one place and all three follow.
The honest part of the pitch is the bottleneck. It is never the software, it is the data: specifications in spreadsheets, photos in a folder nobody has opened since the last rebrand, translations that exist for half the range. Getting that into one shape is most of the job, and it is work the dealer has to help with. In exchange, with the data in one shape a full catalogue is seven days of work, not the season-long website project dealers are afraid of, and that matters when the real deadline is a trade fair with a date on it.
Two versions I threw away
The first build was a single beautiful page about one machine: hero shot, specifications, options, a quote form. It was good work and it was the wrong thing. A dealer does not have one machine, and a page about one machine is an advertisement, not a catalogue.
The second build was a grid of cards with filters, sorting and a list view. That one was a catalogue in the dictionary sense and it read as an online shop, which is exactly the wrong signal for equipment that is bought after a conversation, a visit and a negotiation.
The third build is the one that is live. It boots like a piece of software, keeps the machine in front of you, and moves the browsing, comparing and requesting into panels that never take the machine off the stage. The lesson I would carry into the next catalogue I build: for a dealer, "catalogue" means a product explorer their sales agent can drive in front of a customer, not a page and not a shop.
There was a quieter problem underneath. If an add-on has to appear on the machine when it is switched on, it cannot be a separate photo of a different machine. Every option has to exist as its own layer over the same base shot, from the same angle, in the same light, and the picture set has to be planned that way before the first image is made. That is the part that takes the time, and the part a customer never thinks about.
If you sell machines
Take the demo as a stand-in for your own catalogue: the structure, the interactions and the request flow are the deliverable, while the brand, the machines and the photographs would be yours. Nothing in it is tied to the fictional brand it wears now — the whole thing can be customised to any dealer's range, down to the machines, the add-on groups, the photographs and the languages. The same approach works for any range where a buyer picks options and then asks whether the result fits the space. The rest of what I build around it is on the automation page, and the demo is a click away on it.
