How I build.
Everything below comes from my own companies and from other people's. There are no case studies on this site; clients stay confidential. The method doesn't have to.
The rules
01
Validate before you build.
A roadmap written before anyone has talked to the market is a wish list with dates. At GutWise the order was: prospect survey, B2C and B2B validation, then the roadmap. The roadmap came out shorter, and it survived contact with the first quarter.
Where it comes from: GutWise, Head of Product.
02
One chair.
Product decisions, the words that explain them, the landing page they land on and the roadmap they belong to are one job. Split across a product lead, a copywriter and an agency, every seam costs a week and a little meaning. I sit in one chair and do all four.
Where it comes from: GutWise; every company I've run.
03
Working software, not decks.
If an idea can be touched, it should be. I build prototypes in days, not quarters: a clinical protocol becomes a guided in-app wizard rather than a PDF; a programme step becomes a tool people can use before the product exists. A prototype ends arguments a deck starts.
Where it comes from: GutWise; SHÄP; AcademicAI.
04
Measure what the product does, not what the deck says.
A measurement catalogue with twenty-eight metrics and no product telemetry is a dashboard, not a feedback loop. Which screens people open, which tasks they finish, where they stop: that's the product talking. Everything else is reporting.
Where it comes from: GutWise.
05
Content is the front door, not the decoration.
Relocate.World pairs people who are moving countries with immigration advisors. What brings them in is original, practical writing by professionals, not a hero banner. The same is true for a founder: the essays are the product's front door, and they have to sound like you.
Where it comes from: Relocate.World, co-founder; twelve years of ghostwriting.
06
Marketplaces start on the supply side. [check against your experience]
Nobody comes to an empty marketplace, so you build the side that has something to offer first, and you build it by hand: advisors at Relocate, items and owners at SHÄP. Automate the second thousand, not the first hundred.
Where it comes from: Relocate.World; SHÄP.
07
Research is an asset, and someone else will pay for it.
If your product rests on a question (does the intervention work, does the sensor detect the thing, does behaviour change), that question is fundable. I've spent years inside Austrian government-funded security research, coordinating social science work packages alongside JOANNEUM RESEARCH and the Ministry of Defence, and I've built a ten-partner Horizon Europe consortium. The money is non-dilutive; the data becomes your moat and your next pitch.
Where it comes from: ERC Experience Research & Consulting; Horizon Europe consortium work.
08
Proposals die in the implementation section.
Evaluators forgive an ambitious idea; they don't forgive a work plan that doesn't add up. Person-months that aren't defended, milestones that only mark phase boundaries, partners bolted on for dissemination. Funding is won on structure, and structure is where I start.
Where it comes from: Horizon Europe and FFG proposal work.
The thread
The thread through all eight: voice, product and money are one decision made three times. When three vendors make it, the company pays for the seams. When one person makes it, it holds.
Two weeks. Four documents. A decision.