Daikin · Tools for the people who move the machines
My role: Product design and research, embedded with product owners, engineers, and researchers.
Daikin sells climate systems through people who never appear in the ads: distributors who order equipment by the pallet, and installers who bolt it to walls. Both groups needed software. Neither wanted software that tried to impress them. I started in the field with the distributors and installers themselves, and designed both tools around how they work.
This work is under NDA. This is a text case study; I walk through the full detail in live conversations.
The distributor platform
A distributor’s relationship with a catalog is nothing like a consumer’s. They retrieve rather than browse. They already know what they need, or they know what the discontinued unit was and need its current equivalent. They do this dozens of times a day, and each extra click is a tax multiplied across a working year.
The design position: this is a retrieval instrument. Filters mirror the profession (series, capacity, compatibility) rather than the database schema; reference data lives inside the list so nobody loses their place; and the ordering path assumes history: repeat orders, known quantities, no ceremony.
The installer app
The second tool served installers inside Daikin’s incentive program: install a unit, prove it, get paid by Daikin directly. The design problem here was trust, in both directions.
Watching installers work on site made the fix obvious: the capture flow had to guide toward what verification needed, what to scan and what the photos had to show, before the installer left the job. Getting those requirements into the moment of capture, instead of into a rejection email, was the highest-leverage decision in the product. Payout status stayed transparent at every stage: submitted, verified, paid.
What happened
Within five months of launch, the program paid out several million euros directly through the app. Real money moving through a capture-and-verify flow, to users who have no patience for bad software: adoption at that pace is the metric I trust most.