The Secret User Psychology Behind 100M Users’ App

Launched the casual dating app in 2013 as 'Bang With Friends', a deliberately provocative name that evoked strong emotion. When told the name, people reacted 'That's hilarious, really? I want to try it.' Colin frames it as one way to be 'outrageous' to get noticed; not every name needs to be provocative. The app was later renamed DOWN.

Mapped competitor icons in dating search results (Tinder, Bumble, HER in reds/oranges/yellow; Hinge, Feeld, Pure in dark, clean typeset styles) and designed to contrast: a deep purple no competitor uses, the D and O of the logo personified as two flirty characters, and hints at the product (devil horns on the O, flirtatious eyes, tongue out). The icon matters most because it appears everywhere in search results, while screenshots need a tap; DOWN's traffic comes mostly from its top ~20 keyword searches. Has A/B tested hundreds of icons and keeps testing.

Doesn't force screenshots to match the icon color; tests them independently. Has tested yellow, purple, blue, red, multi-color and tie-dye backgrounds; a deeper red currently wins and stands out. Captions speak to the user's mood rather than features: 'Thirsty for something a little different?' instead of 'match and date with people'; first screenshot shows two people kissing, partly covered by the logo to add mystery; 'Open-minded for whatever you desire, casual or serious.' Expects the set to change again within a month.

Went against the 'shorter onboarding' wisdom: grew a 12-step onboarding (one question per step) to 15 steps. Added a threesome question: 'Are you open to threesomes? Yes/no'; if yes, 'Do you have a partner already?', then 'What is your threesome fantasy?'. The answers go on the profile. Goal: 'good friction' that signals DOWN is open-minded and different from Bumble/Hinge, sets the expectation you can meet that kind of person, and makes users invest effort (IKEA effect). Contrast with bad friction such as search-radius or 'bars or clubs' questions that can wait. Run as an experiment on a 13-year-old app; paywall untouched.

In onboarding, set a high bar for profile quality using a social norm: to get men to add at least three real photos, show copy like '90% of men on our app have at least three photos', which prompts them to keep up with the competition.

Treats App Store optimization as rent already paid (30% of sales go to Apple/Google), so the store page should be the best storefront possible. Competes on roughly the top 20 keyword searches (dating app, hookup app, casual, relationships, matching) since most traffic comes from keyword search, not category browsing; continuously A/B tests icon and screenshots.

Because dating users are used to bots and fake profiles, DOWN goes out of its way to feel authentic: store imagery should show people who look like real target users, not stock or obviously AI-generated photos, and every user is photo-verified to confirm they are real and match their profile photos.

Worked with the data team to find the 'magic moment' when live interactive video clicks for viewers. Gave viewers a free gift in the app: a small animation they could send to the streamer. The streamer could poke the viewer and the viewer could poke back, with the viewer's name appearing in chat. This moved viewers from passive watching to feeling recognized and part of the show (reciprocity), after which they sent more paid gifts (buying points), watched more streams and stayed longer. Then drove more users to that moment.

Seeded with founders' friends, high net-worth people and celebrities for social proof; each user got only two invites, creating scarcity and exclusivity, and people passed on unused invites (urgency). Rooms were live and ephemeral, so missing them meant missing out, which fed more live sharing. Colin reads this as a FOMO cycle of social proof, scarcity, exclusivity and urgency.