Ankerd Con: redesigning an event app before its first con
Ankerd Con is the app my group will use to plan convention trips: rides, meals, rooms and money. Gavin started it, Ayoub built it, and I came in to rework its navigation, its look and its security.
Ankerd Con is the private app of the Ankerd group, the same group behind ankerd.nl. It answers the questions that come up before every convention. Who’s driving and who still needs a seat? Where are we eating on Saturday? Who’s in which hotel room? Who paid for what? Gavin started it in June, and Ayoub built the app itself — the backend, the data model and the screens people actually use. I’m in the group rather than hired by it, and I came to the code in September.
By then the app worked and held real data, but it hadn’t been through a convention yet. Nothing in it had met twenty people sorting out rides at the same time. That made it the right moment to change what’s expensive to change later — what the navigation is built around, what the app looks like, and who is allowed to do what — and the wrong moment to break anything people had already entered.
The complaint: it looked like every other app
Gavin’s feedback was blunt. On a desktop the app looked generic: a narrow phone-width column in the middle of a wide screen, a tab bar at the bottom, and the soft blues, blur and glass that half the internet uses right now. On phones the layout worked fine, so the problem was the look and the desktop layout, not the app as a whole.
Before touching any code I made a proposal: one page with mock-ups of the key screens in a new style. The design got settled between Gavin and me, going back and forth over that page rather than by asking the whole group. The first version came back “too busy”, with avatar stacks and buttons on every row. We tried a raised 3D style and cards that lifted on hover, and dropped both. What came out of that round was clear:
- The look comes from Ankerd’s own mascot, an anchor: ink outlines, one brand colour for the main action, and status colours only for status.
- Flat surfaces. Only a primary button gets a small hard shadow that presses in when you click it. Nothing moves on hover.
- Calm screens. Each screen has one bold element, and everything else steps back.
- A real desktop layout: a sidebar on wide screens, an icon rail on tablets, and the tab bar kept on phones.
I then restyled every screen to match, including onboarding, error pages and the admin portal, and wrote the rules into a short design-system document so that screens added later come out the same way.
The real problem was the navigation
The tabs were Transport, Food, Finance and More. That sounds reasonable until you have two conventions coming up. The Transport tab then showed every ride for every event in one list, and you had to work out which ones belonged to which trip.
Nobody in the group thinks in features. We think in trips. So I rebuilt the navigation around that: Hub · Event · Agenda · Finance · Crew. The Event tab opens one trip, with rides, meals, rooms, cosplays and photos for that trip only. Its overview page became a ticket with a tile for each part of the trip, in the order you need them: before, during, after. Each tile gives the answer straight away (“two people still need a ride back”) and opens the page where you fix it.
The calendar became Agenda. Upcoming trips are a stack of tickets you can swipe through and sign up for in one tap. Past trips turn into a look-back, with who went, the photos from each day, and the rides and dinners. Planning is only half of what the app is for; the other half is looking back at the trip once it’s over.
Every old link redirects to its new home, so nothing anyone had bookmarked or shared in the group chat broke.
Two security audits
An app that holds names, room numbers, travel plans and who-owes-whom deserves more care than “it’s just for friends”. I ran a security audit on the app and fixed what it found, then ran a second audit on the result and fixed that too. The most important fixes:
- Identity comes only from what the login provider has verified. Before this, part of it came from profile data a user could edit themselves.
- Members act only for themselves; admins can act for anyone. Every place that takes a name, from meals and rides to rooms and expenses, now checks that rule on the server, and the app only offers the names you’re allowed to pick.
- Browsers no longer talk to the database directly. Everything goes through the backend, which checks who you are first.
- Uploaded photos are checked, re-encoded, and stripped of location data before they’re stored.
- The rest: security headers, rate limiting, a secret link for the calendar feed, and up-to-date dependencies.
None of this is visible to the people using the app, which is how it should be.
What I took from it
Reworking an app someone else built is a different job from starting one. The fastest route was to make the change visible early, as a page of mock-ups, and to let the objections to it decide the details. It didn’t matter whether the redesign looked good to me. It mattered whether Gavin and the rest of the group recognised the app in it, and whether planning a trip got easier.
The real verdict comes from the first convention we take it to, which hasn’t happened yet.
The app is private, but its code is open on Gavin’s GitHub, as ankerdcon.
Have an app your team uses but has outgrown? Let’s talk.