Introduction

S7 Compas is a marketplace for author-led tours, similar to Airbnb Experiences or Booking Attractions.
The system has a B2C side for travelers who want to book tours like mountain hikes or yacht trips, and a B2B side for guides who organize and run those tours. There are also internal tools, but in this story I’ll focus only on the B2C and B2B products.

Business Problem

The aviation business is known for extremely low profit margins — around 1%, according to The Washington Post. It’s a tough industry. There’s even a joke: if you want to become a millionaire, start as a billionaire and buy an airline.

Another issue is customer loyalty. Airlines want to keep travelers inside their ecosystem instead of losing them to other services.

The goal for Compas was clear: build a new source of revenue without heavy marketing costs, launch fast, and use the existing S7 customer base to power growth.

Research

I started with empathy. I organized ten in-depth interviews with tour guides from different regions and types of tours.
An interesting finding: many customers book tours without prepayment and then disappear, while the guide has already started preparing and buying supplies.

Together with the UX Lab, we also ran a survey for the B2C audience. I had over 200 pages of reports from research agencies and our internal lab, but I’ll share only the key insights.

The main motivation for joining a tour is convenience — travelers don’t want to plan or organize anything themselves.
The biggest barrier is price: a week-long tour to Kamchatka can cost 100–200 thousand rubles.

I built a customer profile based on all data. The average S7 Compas customer is a middle-aged woman with a moderate income.

Another discovery came from the research: about 30% of all trips are short tours lasting one or two days. The project manager didn’t want to include short tours at first, but after seeing the data we added them immediately.

User Problems

From user interviews we found three main pain points:

1. Traveling is expensive.
2. Planning trips is complicated.
3. There are many scammers and unreliable guides.

To understand the experience better, I went on an author-led tour in Kamchatka myself. I learned that even supermarket clerks from small towns join such trips, and that local tours are often promoted through WhatsApp groups.

Design Process

I defined the scope of the project and worked with the team to prioritize features for the MVP and later versions.
We also set design deadlines and discussed how many designers were needed — though in the end, I still had only one.

To present the idea to the S7 team and the CEO, I made a trailer video for Compas. I wrote the script, edited the footage, and handled the entire production myself.

Product Entry Points

Because we were integrating Compas inside the existing S7 app and had almost no marketing budget, I focused on designing smooth and visible entry points into the new marketplace.

Visual Style

From the beginning, I wanted the visuals to be vibrant and emotional. Guides already had great photos, and I planned to use them as the main attraction.
To make tours slightly cheaper and improve retention, customers earned S7 miles for every booking.

I also wanted Compas to be more than a once-a-year vacation app. It could become a lifestyle product — a place to browse new destinations, read comments, and plan future trips.
That’s why the tour feed was designed like an Instagram feed: with large photos, videos, and comments directly under each tour.

Experiments and Testing

One experimental idea was to show tours in a Reels-like format. It was risky, so I tested it.
We set up a small testing studio with two cameras and phone-screen recording. Researchers helped write the script and recruit respondents.
I led the interviews with my designer, teaching him how to conduct sessions during the process.

The tests showed that Reels didn’t work well for our core audience (age 30+), though younger users liked it. The conclusion: this format could be used as a separate section, but not for the main search experience.

Search and Filters

I designed a focused search system with only essential filters — the ones users really need to find a suitable tour.
Less-important filters were planned for future releases.

During testing, one problem appeared with the calendar: users didn’t understand what dates to enter. We solved it by adding a “tour duration” filter directly on that screen.

Booking and Payments

The payment flow included a required prepayment. This guaranteed that guides could prepare for the trip and that customers were serious about attending.
If a customer canceled too close to the tour date, a small cancellation fee was applied.

Post-Booking Management

After booking, users could manage their tour: cancel, reschedule, or chat directly with the guide.
Bringing communication inside the app was important — it keeps the customer within the ecosystem and allows us to integrate more Compas functions into the chat later.

Results

We built a working marketplace that performed well during user testing.
The B2B part has already launched and guides are actively adding their tours.
The project demonstrated that even with limited resources, it’s possible to design a user-friendly, research-based travel product that extends the airline’s ecosystem and opens a new revenue stream.