Side Project2015
Turnout logo

The problem

Having been a long-standing student at the University of Kent, many new students would ask me for suggestions about the nearby nightlife. TurnOut allowed users to discover the best events tailored to their preferences.

User research

University students were asked what was most important to them when searching for a night out. The results included cost, music, busyness, vibe and dating.

Recommendation algorithm

The proprietary recommendation algorithm incorporated these findings to provide a personalised event ranking for each user.

TurnOut required users to login using their Facebook accounts in order to access the Search and Page Event APIs.

The APIs were used to automatically extract nightlife-related pages and events:

  1. Search for all business pages within a radius of the city centre.
  2. Score each page based on a set of nightlife-related keywords to identify relevant pages.
  3. For each event hosted by each page, request the title, description, location, timings and guest list.

Login screen with the TurnOut logo and a 'Continue with Facebook' button.

Recommended events were displayed in a ranked list, providing information about the event, its location, opening times, music genres and the attendees. Events up to a week in advance were available.

The guest list data would include a list of names and their RSVP status. I contacted each venue to gather the capacity figures and used these to rank the busyness of the event.

I also used a hardcoded list of male and female names to render an approximate gender split* (see blue/pink background).

* Please excuse the fact that this feature is not inclusive in current times.
UI displaying a list of events. Each event contains the title, location, times, music genres, description and attendees. At the top, there are tabs to view events for the next seven days.

Many event descriptions would mention genres of music. I matched these against a database of genres and alternative spellings (e.g. Drum & Bass and DnB) to extract a concise set of genres for each event.

Mockup of the music genre selection screen, with text: 'Tell TurnOut which music genres you enjoy to improve the ranking'.

A common deciding factor for attending an event was whether user's friends are also going, so TurnOut would show which of their friends were also attending.

Mockup of the component showing how many and which of the user's friends are attending, with text 'See which events your friends are attending'.

On the settings screen, users could set their preferred music genres and dating preferences which would contribute to the recommendation algorithm.

Sadly, in response to the Facebook–Cambridge Analytica scandal, Facebook shut down their Page Event API leading to the deprecation of TurnOut.

I regularly check for new avenues that can bring TurnOut back to life. Please get in touch if you have any leads.

Settings screen displaying user's profile picture, name and selected preferences for location, genres and dating.


The design for this project was done by Jacob Hammond.