Thanks, 2016. It’s Been Weird.

Read the campaign story

A streaming service knows what people play, but a usage chart is not automatically an interesting advertisement. Spotify’s late-2016 outdoor campaign turned listening behaviour into short observations about the year. Instead of leading with catalogue size or interface features, its posters addressed listeners through the tracks and playlists they had chosen. The work treated music as a trace of how people were living through public events and private moods.

Contemporary launch coverage describes a rollout across 14 territories with locally relevant data on billboards. One British execution connected a spike in listening to R.E.M.’s end-of-the-world song with the Brexit vote. D&AD’s entry identifies the work as a celebration of fans and artists and records it in its 2017 archive. The year of that award entry is not the campaign’s launch date: the outdoor work was already reported in November 2016.

The useful mechanism is editorial selection rather than data volume. A behaviour becomes a story only after someone chooses the interesting detail, supplies an audience-relevant context and writes a clear observation. Spotify also made a private-feeling activity public enough to recognise without requiring a passer-by to log in. That separates this execution from a personalised recap: the communication unit is a shared street-level joke, not an individual’s dashboard or share card.

A small team can borrow the method without mining intimate customer behaviour. Use an opt-in poll, genuinely aggregated activity or voluntarily submitted anecdotes to find a pattern that says something about the community. Then test whether the writing feels affectionate and intelligible to people outside the data team. The important limit is trust. An impressive data capability can become an uncomfortable reminder of surveillance if the brand exposes an individual or makes a sensitive inference. No campaign-specific subscriber or revenue gain was verified in the sources used here.

Find the human observation first

Ask what a behaviour reveals about a recognisable moment before deciding which statistic to feature.

Write for someone outside the dashboard

Reduce each execution to one fact, one context and one readable punchline. Test comprehension at billboard viewing speed.

Protect the people behind the pattern

Use consent, aggregation and disclosure controls. Reject jokes that expose a person’s private life or infer sensitive characteristics.

A small test for another brief

Turn an opt-in community survey into three humorous, privacy-safe poster concepts.

Fast comprehension, perceived warmth and voluntary sharing intent; record privacy discomfort separately.

Reject any line that identifies an individual, relies on a sensitive inference or makes the community feel watched.

• Collect one non-sensitive preference and a voluntary explanation without names.
• Choose three sufficiently aggregated patterns and write one observation for each.
• Test the posters with ten people who did not take the survey.