BarHub

The problem

Local pubs and bars face intense pressure from larger national chains with greater marketing budgets and brand recognition. Smaller establishments struggle to reach new customers, often relying on word-of-mouth while potential visitors default to familiar or well-advertised venues.

For people going out, discovery was just as broken: it was frustratingly fragmented, relying on generic tools like Google Maps that lack any niche focus, personalised recommendations, or the social, planning dimension central to going out in groups.

Research & insights

I ran a mixed-methods research phase: an online survey of 30 potential users, three personas with mapped journeys, and a competitive SWOT. The survey was clear about what actually drives a choice of venue: atmosphere and pricing come first, ahead of distance or brand. People wanted to find places that fit a mood, then plan a night around them.

To find the gap in the market, I ran a SWOT on the two tools people were already using:

The wedge: combine the strengths of both (map-first discovery, a genuine niche focus, personalised recommendations and crawl planning) while fixing the weaknesses neither had solved.

Personas

Three personas kept the work honest, each a different reason to open the app.

Design process

I worked from structure outwards. A priority guide ranked what mattered most, a critical flow chart mapped the journeys, and wireframes turned both into layouts, all before any visual design.

From there I built a high-fidelity, interactive prototype in Figma, using iOS templates and SF Symbols for a native feel. Then I designed a brand to match.

The wireframes became a high-fidelity prototype: rich venue profiles, a map-first home, and a news and offers feed.

One small change came straight from user feedback. The app was originally called “Boozer”; testers felt the word was dated and off-putting, so it became “BarHub”, a central hub for discovering local pubs and bars. Minor on its own, but a clear sign of letting users steer the design.

The build

Most UX projects stop at a prototype. I took BarHub all the way to a fully functional iOS app, rebuilding the Figma designs in Xcode and SwiftUI.

The real engineering challenge was the map. Rendering live pub markers meant learning MapKit, using MKLocalSearch to query for pubs around the user’s location and drop them onto the map.

func searchPubs() {
    let request = MKLocalSearch.Request()
    request.naturalLanguageQuery = "pubs"
    request.region = MKCoordinateRegion(
        center: CLLocationCoordinate2D(latitude: 51.481230,
                                       longitude: -2.580748),
        latitudinalMeters: 5000,
        longitudinalMeters: 5000)
}

Coming to Xcode and SwiftUI as a new iOS developer meant a steep learning curve. I closed the gap with self-directed learning and an agile build-test-refine loop, shipping the app in small working iterations rather than all at once.

I also made a deliberate cut: no login screen. Going out is spontaneous, so forcing an account would only add friction. Skipping it keeps the experience instant while protecting user privacy. You open the app and you’re straight on the map.

The pub crawl planner is BarHub’s signature feature. Per the research, no existing app offered anything like it. You name a crawl, add venues in order or auto-generate a route, and the app plots the path between stops so a group can move through the night without breaking stride.

The pub crawl planner: name a crawl, add venues in order or auto-generate a route between them.

Testing & iteration

I ran small, focused user-testing sessions throughout the build, with a handful of target-demographic users doing real tasks: find a venue, view a review, plan a crawl. Quick to run, and rich in detail.

The most important finding came early: the initial navigation confused people and stopped them completing tasks. They couldn’t reliably move from discovery, to detail, to planning.

So I redesigned the navigation, streamlining the flows down to the three-tab Map / Planner / News structure from the flow chart and clearing the paths between them. Task completion improved in the next round of testing.

Test, learn, fix, then test again. The navigation only worked because users told me the first version didn’t.

Outcomes & reflection

Read the full report