migulko
.eu
Get a free review

9 November 2026 · 3 min read

From 3 free fixes to a full redesign: how my product review works

How a product review goes step by step: 3 free fixes first, then a paid review with analytics, then targeted fixes or a full redesign, whichever the data supports.

Taras Migulko · Product Designer & Product Owner
You send a linkA website, dashboard or mobile appI go through the whole productIs it clear in 3 seconds, even without the language?WEBSITE · DASHBOARD · SAASMOBILE APP · APP STORE · GOOGLE PLAY01Look and functionEvery page and every key flow, as a new user02It all goes into FigmaI mark the problem areas and comment on what to change1The button gets lost.Make it the one main action03A 5-minute LoomI walk you through what I would change01I download the appBest when it is live and available in Czechia02Search keywordsWhich searches it shows up for, which it missesMost important03Store pageDo the title and text match those searches?04LookScreenshots and screens: is everything clear?05FunctionDo the first 3 screens work for a new user?06A 5-minute LoomI walk you through what I would changeYou get the 3 things I would fix firstFigma comments and a 5-minute Loom, free, no call

Most products I look at do not need a redesign on day one. They need someone to go through them the way a new user does and point at the few places where people get stuck. So I split the work into steps. Each step is small enough to judge on its own, and after each one you decide whether to go further.

A free review: the 3 things I would fix first

You send a link: a website, a dashboard or SaaS product, or an app on the App Store or Google Play. I go through the whole product the way a new user would. From there the review splits in two, depending on what the product is.

A website, dashboard or SaaS

I review how it looks and how it works. I go through every page and every key flow: the first screen, sign-up, onboarding, the first real task and the moment it asks for money.

Everything goes into Figma. I put the screens there, mark the problem areas and write a comment next to each one: what to change and why. On top of that I record a 5-minute Loom and walk you through the changes, so you see each problem on the screen instead of reading about it.

A mobile app

Here the review covers the look, the function and the search keywords. It works best when the app is live, so I can download it the way any user would. I am based in Prague, so the app needs to be available in the Czech App Store or Google Play.

  • Search keywords. This is the most important part. I check which searches the app shows up for and which ones it misses. An app that people do not find in search never gets the chance to show how good it is.
  • The store page. Do the title, subtitle and description use the words people actually search for?
  • The look. The screenshots and the app itself: is it clear what the app does and where everything is?
  • The function. The first 3 screens after install. Do they get a new user to the point of the app, or lose them on the way?

In both cases you get the 3 things I would fix first: comments in Figma and a 5-minute Loom, by email. There is no call and no obligation. If the 3 fixes are useful, your team can ship them right away.

Send your product for a free review

What I check first: is it clear in 3 seconds

Whatever the product is, the first question is usability. Is everything clear to a new user, or is something important hidden? I give the first screen about 3 seconds, the same as a visitor does. If it is not clear by then what this is and where to go, people leave, and nothing further down the page matters.

For a landing page this means one thing above all: show the product right away. The product should not hide in a paragraph of description or below three screens of slogans. From the first screen it should be clear what it is, what it does and where everything is.

A test I use: imagine the visitor does not read the language of the site at all. The page should still make sense. Below is the same landing page in a language most people reading this do not read. On the left the product hides behind text. On the right the first screen shows a number, a chart and one button, and you can tell it is a sales dashboard without reading a single word.

Product hidden in textProduct shown firstThe first 3 secondsすべてを一つにまとめる、新しいプラットフォーム私たちは、チームがより良く働けるように、革新的で柔軟なソリューションを提供します。データ、人、そしてアイデアをつなぎます。詳しく見るシンプルで、安全で、すべてのチームのために。今日から始めましょう。The first 3 seconds売上を、ひと目で。無料で試す今月の売上€48,210+18%You read it and still don't know what it doesNumbers and a chart: you get it without the words
The 3-second test: if a page only works when you read every word, most visitors will not get past the first screen.

The same goes for apps. The first 3 screens have to explain themselves without a manual, and the store page has to show what the app does before anyone reads the description.

A full review with analytics

A first look shows what is on the screen. It does not show what people actually do. If the 3 fixes make sense, the next step is a full review of the product with real data.

I check what analytics the product already has and add what is missing: Firebase, Google Analytics, App Store and Google Play analytics, RevenueCat for subscriptions and heatmaps for websites. Then I map the funnel from the first visit to payment and look at where people drop off, which screens they skip and what brings them back.

You get:

  • the funnel with the numbers at every step;
  • the problems ranked by how many users they affect;
  • a short plan: what to fix, what to redesign and what to leave alone.

This part is paid, at a fixed price agreed before I start.

Fix what holds the product back

Often the data points to a few screens: an onboarding that asks too much, a paywall that shows up too early, a form people leave halfway. Then the right move is to fix those parts and leave the rest alone. I design the changes, build them or hand them over to your developers, and after the release I check the same funnel to see whether the numbers moved.

Or redesign it

Sometimes the problems sit deeper: the main flow is in the wrong order, the product tries to do five things at once, or the interface cannot take new features without breaking. Then patching single screens costs more than starting over.

A redesign goes through the usual stages: research and user flows, a clickable prototype to test the new flow, the design system and all screens, then build and launch. The choice between fixes and a redesign comes from the data in step 2, not from taste.

Build here means the whole product, not only the screens. I connect the platforms and APIs, mailboxes and automated emails, set up the database and how the data is split between services, deliver files to the servers and set up the security. The code is Swift for iOS, Kotlin for Android, Node.js, HTML and CSS, kept in GitHub, and every release is tested before it goes live. This site is built the same way, and on some project pages you can try the prototypes right inside the phone and browser screens.

After the launch

The work does not stop at the release. I keep watching the same numbers: retention, trial-to-paid conversion and drop-offs in the funnel. They decide what goes into the next release. This is how I run my own app, Mild, which is live on the App Store and Google Play.

Why it goes in this order

Every step is a small decision. The free review costs you a link. The full review costs a fixed price and gives you a plan based on real numbers. Only then do you choose between fixes and a redesign, so the budget goes where the data says it will change something.

If you are building a product and want a second pair of eyes, send it over for a free review.