Rowm, designing in a category where everyone sells the same thing

A commodity market

Rowm is a travel eSIM app. Pick a destination, buy a data plan, install it, land with internet already working. It sells to individual travellers and is packaged as bundles for companies whose staff travel often.

There are dozens of apps doing exactly this. They cover the same countries, resell from the same operators, and describe themselves in almost the same words: 200 destinations, instant activation, no roaming fees.

Rowm has no structural advantage over any of them. It competes on price and on the markets it can reach. That was clear from the first conversation, and it set the brief. If the product is a commodity, the only thing left to compete on is whether it works without friction at the moment it is needed.

For a travel eSIM, that moment is narrow and stressful. You have just landed. You do not have internet yet. If something goes wrong, you cannot easily search for help.

Thirteen apps, annotated

Before any screens, we took the market apart. Around thirteen competitors, captured through Mobbin and by hand, laid out flow by flow on one board and annotated with what worked and what did not.

Then the reviews. Store ratings for each app, and forum threads where people described what had actually gone wrong for them.

Most complaints were not about price or coverage. They were about not knowing whether the thing was working. Installed but not connected. Connected but no data. The wrong plan for the wrong country. A plan activated too early, burning days of validity before the trip started.

We kept a second board of patterns from outside the category, from products with nothing to do with eSIM, for the parts where copying a competitor would only have reproduced their problems.

Thirteen competitors, laid out flow by flow and annotated. The complaints in their reviews pointed at the same place: the gap between paying and being online.

The gap between paying and being online

Buying is the easy part. Every app in this category has a serviceable checkout.

The hard part is the stretch between payment and working internet. Install the profile. Have a device that supports it. Turn on data for a line that is not your usual one. Wait for a network at the other end to pick you up. Any of those can fail, and each fails differently.

So most of the work went into states rather than screens.

• A compatibility check before purchase, so nobody buys a plan their phone cannot use.
• Every install and activation state drawn, including the ones nobody enjoys designing: pending, failed, partly complete, expired before first use.
• A way out of each failure, rather than a dead end with a support address on it.
• Plain confirmation after install, telling people the plan activates by itself once they connect to a network at their destination, so that silence does not read as something broken.

In a product people rely on while stranded, an unhandled state is a support ticket at best and a refund at worst.

A compatibility check before anyone pays, guidance after install so that waiting for a signal does not read as a failure, and an empty view that points somewhere instead of sitting blank.

A home screen for someone already travelling

The original plan for the home screen put discovery and promotion first. Destinations, offers, campaigns.

That serves someone who has not bought anything yet. It does not serve the person the business actually keeps, who has already bought and is now abroad, checking how much data is left before opening a map.

We pushed for active eSIM cards on the home screen, showing remaining data and status without a single tap. The owner's objection was build cost, and that he had not seen the pattern in the apps he used himself. Two arguments settled it: people open this app mid-trip far more often than mid-purchase, and the components already existed in the design system, so it was cheap to add.

Campaigns moved down, still visible without scrolling but no longer the first thing anyone sees. In their place, a header with the person's name and how many plans are active. An app people open while they are far from home should greet them, not sell to them.

The header and the cards change with the person's state: nothing yet, one plan running, two plans with data left and data spent. Campaigns stay visible without pushing the traveller's own plans below the fold.

What it runs on

Rowm ships in English, Arabic, Spanish and French. It is built on the same multilingual system architecture from the banking work, adapted to this brand, so adding a language does not mean rebuilding screens. I cover how that system works in a separate case study.

I hired and directed one designer on this. We worked out the concepts together, he owned building components into the system, wiring prototypes and gathering the competitor screens, and I owned the flows, the harder screens and the relationship with the founders.

The app launched in early summer 2026 on iOS and Android. A referral programme and a dark theme are specced and handed off for the next release.

Nobody wins this category on destinations. It gets won on having the fewest moments where someone abroad cannot tell what is happening.