Web app for planning trips:

how I researched the market, designed the app, and got it live

Try it out

Try it out

Opens in a new tab. No account needed

Planner opens in a new tab.
No account needed

Underhype web app, signed in: a Stockholm trip for December 1–7 with four travelers, a budget, and transport from Arlanda AirportUnderhype on a phone, signed in: a Stockholm trip for December 1–7 with transport from Arlanda Airport

Underhype’s homepage

Idea

The app helps you travel around Sweden and stay on budget.

You see a plan with every move from the moment you arrive in the country — with overhyped places filtered out.

Context

While laptops are great for planning, the phone is still the obvious king once you’re on the trip.

Underhype on a phone: choosing a city, Stockholm, Malmö, or Gothenburg

Homepage for logged-out users

Underhype on a phone: transport from Arlanda Airport with times and prices

Airport transfer options

Underhype on a phone: a map of Stockholm with the offer to get the full app

Map tab for logged-out users

Target audience

Underhype is made for Americans, Brits, and Germans — the biggest groups of international travelers in Sweden. They stay in Sweden for 2–7 days and plan a trip several months ahead.

Context

The idea of Underhype — cut the noise and avoid the mainstream. This idea rules everything in the product.

Expertise in the Swedish market
instead of being a universal app

Hand-picked places and editorial tips.
AI is involved only as a proofreader

One-time payment, no subscription.
Users get the app for a year

Independence. No café or restaurant
can pay to be mentioned

Context

The first version of Underhype was designed at the beginning of 2026 — in parallel with research. Back then it was mostly me testing what AI could do. Once the research and interviews were done, I redesigned the planner from scratch. This time with the financial plan, the user groups, and the marketing channels in mind.

Draft version

The homepage was half marketing, half product. You had to go through three steps before you saw any trip plan.

Underhype on a phone: the first step of planning a trip to SwedenUnderhype web app: the start screen, Plan your trip to Sweden

Dynamic sidebar. On search results, it shrinks and gives more control options

Problems

In the interviews, I asked people to try the interface, and you could see them getting bored. Which city? When? How do you arrive, how do you leave? Clicking through all of that felt like a chore, and they still hadn’t seen a single tip. Besides, the new marketing strategy was to get users just by giving away all the city tips for free — in a way nobody, AI included, can. That was supposed to bring in leads for the paid features — interactive maps, the budgeting tool, offline mode. So the goal was to show the actual results right away, on the homepage, get attention, and then sell extra features to users. The whole marketing part will be done even before people land on the site — on YouTube, Threads, and Instagram.

Release version

The new version shows search results for Stockholm right away — with simple options to change the city, dates, and means of travel.

Underhype web app: Stockholm trip plan with transport from Arlanda Airport and pricesUnderhype on a phone: Stockholm trip plan with transport from Arlanda Airport

Dynamic sidebar. On search results, it shrinks and gives more control options

Context

01 Claude Design

02 Paper

03 Claude Code

04 Localhost

05 Pre-production

06 Production

Production cycle

with AI

01

Claude Design

03

Claude Code

04

Localhost

05

Pre-production

06

Production

02

Paper

Since I was the only human on the project, it was all about art direction, handovers, and documentation for agents.

Claude Design

Generated ideas and kept everything at the design system level, using only color tokens, components, etc.

Paper + Claude Code

I refined things in Paper, handed them back to Claude Design, then to Claude Code to build on localhost.

Pre-production environment

Features behind the paywall got tested in pre-production. I also used it for usability tests.

Context

To “vibe-code” something that actually works, you have to give the AI clear instructions. I used what I learned back at Yandex — an enterprise-level tech company — and documented every step from both the user’s and the system’s perspectives. At the end of the day, it was a lot like a PM job: sending detailed specs to an engineering team.

Context

I follow the Basecamp principle: Fix Time and Budget, Flex Scope. So instead of launching everything at once, I release features one by one to hit deadlines.

Underhype web app: the privacy policy next to a painting of a girl reading in bed

In the pre-MVP version of the app, I didn’t have visuals yet, so the app used Danish paintings.

Underhype is not a pet project. It’s designed to be profitable. The app can generate around 612 € in profit per month. Here is a simplified financial plan for a pessimistic scenario.

Monthly budget

Claude

22 €

Vercel

20 €

Domain

1 €

Supabase

25 €

Paid marketing

0 €

Listed costs / month

68 €

Monthly sales example

New purchases (example)

100

Sales at 9.99 €

999.00 €

VAT reserve (assumed)

199.80 €

Store fee (assumed)

119.88 €

After VAT + store fee

679.32 €

After listed costs

611.32 €

Cover listed costs

Full access price

9.99 €

Purchases / month

11

After VAT + fee / sale

~ 6.79 €

Since I’m solo, I have a plan for a year to make sure I stay on track and keep shipping.

Q1

Q2

Q3

Q4

Research & validation

Building: Claude Design + Claude Code

Release MVP — September

Building iOS version

Today

Drop me a line

work@hugolofquist.com