Product Diagnosis: A one-week product audit for founders
For founders and product leads about to commit more engineering time to a direction nobody has pressure-tested. Over one week I review the product, the research and usage data you have, the roadmap, and the reasoning behind the decisions already made. You get one written recommendation: what to keep, what to cut, what to change, and what to do next.
To start, email andriy.dsgn@gmail.com with a short note about the product and what you’re unsure of.
Who it is for
Founders and product leaders about to commit more time and money to a direction nobody has pressure-tested. The situations it comes up in:
- Adoption is flat while the team keeps shipping.
- The roadmap holds more than the team can build, and nothing on it is obviously cuttable.
- Two or three directions are on the table and nobody agrees which one is real.
- Complexity has built up to where every change costs more than it should.
- A large piece of engineering work is about to start on an assumption nobody has tested.
- The product was built fast with AI tools, it works, and nobody is sure it holds together.
How it differs from a UX audit
A UX audit checks screens against usability heuristics. Product Diagnosis starts with the decisions behind the screens: which bets the roadmap rests on, what evidence supports them, and which parts of the product absorb effort without paying it back. Interface problems go into the report when they explain one of those.
If the product was built with AI tools
Products built with Lovable, Cursor, Claude Code or similar tools tend to fail in the same places. Each screen looks fine. Together they ship several button styles, more than one way to confirm an action, and flows nobody designed end to end. Before the diagnosis I take an inventory of what the product actually ships, with my own tools, and the findings come back as tasks your team or your agent can act on. How that works: A canvas for reviewing AI-built UI.
If the problem is consistency more than direction, the smaller first step is a Consistency Review, from $700.
What happens in the week
It runs on what already exists: the product itself, whatever research and usage data there is, the roadmap, and the reasoning behind the decisions already made. I work through the user journey as it actually runs, the assumptions holding up the current plan, and the constraints the team builds inside.
Where something cannot be answered from what exists, that gap is itself part of the finding, and it usually turns out to be the expensive one.
What you get
One written document, in four parts.
- Keep. What is working, and what the next round of changes should not be allowed to break.
- Cut. What is absorbing effort and returning nothing.
- Change. What is right in principle and wrong in execution.
- Next. The order the remaining work should happen in, and what to prove before the expensive parts start.
Client reports stay private. The same format, applied from the outside to a product anyone can open: A product diagnosis of n8n.
Why do this before more engineering
Product uncertainty gets more expensive the later it is resolved. A direction that turns out to be wrong costs a quarter of engineering time, the hiring built around it, the design work on top of it, and the founder attention spent defending it. A week spent on the decision is cheaper than a quarter spent on the wrong build.
Not sure it is worth a week yet? The free Product Check takes three minutes and tells you what to look at first.
Where the judgment comes from
Fourteen years inside product teams, plus ventures of my own where a wrong call cost me directly. The startups I have worked with raised more than $200M between them, and Volterra, which I helped take from a rough idea to a platform, was acquired by F5 for $500M. The same work the case studies describe:
- Fully AI: a one-month engagement that began with a full UX audit of an AI agent product, then a design system and a dashboard redesign.
- Equi: zero-to-one design for a digital family office. Onboarding time dropped by 40%.
- TrueFi: principal designer on a DeFi lending protocol that originated more than $1B in loans in its first year.
- Deva: an AI agent platform designed from zero. In the beta, daily active users grew 25% and retention 20%.
- 10Planet: the design architecture under Deva: agents, memory and the design system mapped to protocol logic.
After the diagnosis
If the diagnosis shows the work is worth doing, I can stay on to design it, part-time or as a fixed project, or fix the design system the product is built on. See fractional product architecture & design. If it shows the work is not worth doing, that is the finding.
Redesign is quoted flow by flow from what the diagnosis found, from $700 a flow. The diagnosis fee is credited in full to the next piece of work if it starts within 30 days.
Questions founders ask
Is one week enough to understand our product?
For the decision you are about to make, yes. I don’t need to know everything the team knows. I need the product, the plan and the reasons behind it, and the gaps between them show up quickly. If a week is not enough, I will tell you before we start.
Our product is very technical. Will you understand it?
I have designed an AI agent protocol, a DeFi lending protocol and a cloud infrastructure platform, so technical depth is the normal case for me. I read the documentation and talk to the engineers before I look at the interface. A new domain takes me days to learn, and in any domain the real question is the same: what this product should do, for whom, and what is in the way.
We already have a designer. Why would we need this?
Your designer works inside the decisions the team has already made. A diagnosis looks at those decisions from outside, with no stake in any of them. The report is for the whole team, your designer included.
Can a designer judge product strategy?
Most of my 14 years have been decisions: what to build first, what to drop, what to promise. Design is where those decisions become visible, and where their mistakes show first. I have also run businesses of my own, so I read a roadmap as money being spent.
How much does it cost?
From $900, fixed before we start. The final price depends on the size of the product and how much there is to go through, and you know it before the week begins. If we continue into design work within 30 days, the fee is credited to it in full. Redesign after it is quoted flow by flow, from $700 a flow.
What if we disagree with the recommendation?
Then we talk it through. Every point in the report comes with the evidence behind it, so you can check the reasoning and not only the conclusion. You are paying for a clear opinion. What you do with it is your call.
What if the answer is “don’t build it”?
Then the report says so and explains why. It is the cheapest finding you will get: one week instead of a quarter of engineering.
Can you review a product built with Lovable, Cursor or Claude Code?
Yes. I take an inventory of what the app actually ships: its screens, components and flows. The findings come back as tasks your team or your coding agent can act on. If consistency is the only question, the narrower Consistency Review starts at $700.
Do you work with teams in other countries?
Yes. I work remotely, with teams in different countries and time zones.
Do you do one-off consultations?
Yes. A consultation is $250, up to two hours. Smaller projects get a fixed price once I know what they include.
Not sure what to build next?
Send a short note about the product and what you’re unsure of. If it’s not a fit, I’ll say so.