Smoking Snapshot

The idea came out of a team brainstorm. Someone suggested a Spotify Wrapped for smoking, and the note from the room was that the big number should be a shock. I had a working prototype by that evening.

It was called Smoking Wrapped, and it led with borrowed cigarettes, the ones social smokers take from friends and never count. One slider asked how many were borrowed. A toggle asked whether you’d call yourself a smoker, and if you said no, the result answered back. “You said no. Your year says otherwise.”

Borrowed cigarettes didn’t make it past the first review. The yearly total as the shock stayed in every version after.

When the team asked for it for real, I wrote the production plan: hosting, a database for the answers, the domain and a note that “Smoking Wrapped” was a prototype name, not a campaign one. It became Smoking Snapshot in July.

Then I wrote the user journey. Two routes, one for social smokers and one for light smokers, each asking only the questions that fit, with the same result cards at the end.

The biggest decision was the ending. Products like Wrapped end with a share card. The journey I wrote kept the result private. On venue iPads, a QR code gives people a way to carry on with support on their own phone.

Aleks’s design arrived on 9 June, with purple waves, tall condensed type and a colour card for every number. I rebuilt the app to match it, and it went live on 24 June.

The mockup put every question on one form after the result, with the questions about you first and the reaction last. On a phone, that’s a long form after the part people came for. So I split it into one question per screen, one tap each. In July I moved the three reaction questions to the front, while the number was still sinking in.

Five days before Pride, I found out the campaign iPads were on iOS 12, from 2019. Its Safari doesn’t understand clamp(), which the layout used for nearly every size, so the screens broke. I added a fallback for all 156 uses and fixed tap to advance for the older browser.

Then I made it ready for Saturday. Answers save on the iPad first and upload when the signal comes back. The app keeps working offline once it has loaded. Done resets it for the next person. Touch targets went up to 44 pixels, and hover effects that stuck on touchscreens came out.

The team took Snapshot to Liverpool Pride on 25 July, the Wirral Food Festival the day after and Warrington Wolves on 23 August, with iPads at each.

At Pride, 88 people went all the way through, nearly all of them on the iPads. Every one of them answered every question, because at that point a use only saved at the end. That’s nearly a tenth of the campaign, in one day. The next eleven days brought 81.

Warrington Wolves fell in the campaign’s busiest week, 413 uses in seven days, and by then paid social was doing most of the work. 64 of that day’s 83 came through Instagram. The events put Snapshot in front of people face to face, and the ads took it to everyone else.

After Pride, I suspected people were leaving before the last question. The app saved once, after question seven, so anyone who stopped earlier left no trace. Their sliders, their result and their reactions all went with them.

I tested three layouts at phone and iPad sizes. A long page with all seven questions under the result. A permission screen before the first question. And a split that asks the three reactions first, saves, then offers the rest as optional. The split was easiest. It shows 1 of 3 instead of 1 of 7, and each reaction takes one tap. The client wanted a big engagement number on those three questions, so the order suited them too.

The save after the reactions went live on 6 August.

Every figure here comes from the campaign’s own data, with the number of people behind each percentage. The answers were optional, so each question has its own base.

I built a live dashboard for the team. It refreshes every minute, and it has a read-only feed an AI agent uses to update the campaign slide, so updates came straight from the data.

Then I wrote the final report, in September. Eleven slides on what happened, what changed and what to do next, with every percentage next to the number of people it’s based on.