Slides · Experiment
First-slide experiment
In Slides, Miro's presentation tool, most people made one slide and never came back to it. I ran an experiment on that first moment to find out what makes people keep building.
Getting more people to try something is not the same as getting them to find value in it. The first slide is the moment people decide, so that is where to show why it is worth continuing.
Why
Almost half of all slides (48%) were created the same way: by turning a section of the whiteboard into a slide. But a week later, only 7% of those slides were still in use, and 78% held just one slide. People tried it once and left.
My hypothesis was that they never saw what slides are for. So I tested two new ways to start, against the old one.
My contributions
- Designed and ran the experiment. Defined the question, the three versions to compare and what counted as success, then measured whether people were still using Slides a week later.
- Checked new features before release. Led the final quality review of two new Slides features, ready-made slide templates and a slide picker. Tested them in light and dark display, in other languages, and next to timelines and interactive items on the board.
Who benefited
First-time users. With the full-screen editor, they were 24% more likely to come back within a week. It did not help people who already used Slides, so I recommended offering it to newcomers only.
The team. The pop-up library got more people to add a slide (16 percentage points more), but each person added a third fewer. That changed what we aimed for next: showing why a second slide is worth making, not just how.
Result
The +24% for first-time users was statistically significant. Across all users, neither new version changed one-week retention in a way we could rule out as chance.
What I'd do differently
I'd test the follow-up sooner: a short prompt after the first slide that explains why a second slide helps. The library we tested had also changed by the time the results came in, so I'd re-run it against the improved version to see if it can match the full-screen editor on total slides and return rate.
Want to talk about this? Get in touch →