Corvane, a fictional vacuum brand, is deciding what goes into its next cordless stick vacuum. 24 people who own a stick vacuum spent four days on a private online board, about 20 to 30 minutes a day, whenever it suited them. Each day opened a new set of activities: photos of real floors and storage spots, a narrated cleaning video, a markup of the new concept and short polls. A moderator read every post and followed up with probes, in the open thread or privately.
Finished all 4 days
24 of 26
2 dropped after day 1; their posts are excluded
Posts and replies
1,148
about 48 per participant
Moderator probes
212
every probe answered within a day
Photos and videos
227
186 photos, 41 short videos
Day by day activity plan As run
Who took part Measured
18 questions by type
10
Group5
Private3
Private firstPrivate first is used for reactions to the concept, so the first answer is not shaped by whoever posted first.
Posts and probes by day Measured
Day 5 is a short follow-up morning: no new activities, only probes on loose ends.
When people posted Measured
45% of posts came after 5 pm. A two-hour group on a weeknight would have missed the people who clean after the kids are in bed.
This is how a board reads from the inside. Participants answer the activity, see each other's posts in group questions, and the moderator steps in with probes: asking for an example, a photo or the reason behind a passing remark. Private questions look the same, but only the participant and moderator see them.
What the probes added Measured
7
of 24 mentioned the battery in their first answer18
of 24 described cleaning around the battery once probedPeople rarely volunteer a habit they have stopped noticing. One question, "what happens on the days it dies early?", turned a minor complaint into the largest theme of the study.
How we probe
On day 3 each participant saw the Glide 2 concept board privately and placed up to six markers on it: a green marker on anything they liked, a red one on anything they disliked, each with a short comment. Nobody could see anyone else's markers until they had finished. Click any area of the board to see what people said.
Green fill: liked by at least half the board. Red fill: more dislikes than likes. Numbers are participants who placed a marker on that area, of 24.
Liked and disliked hotspots, of 24 Measured
130 markers in all, an average of 5.4 per person. Every marker came with a comment.
Private answer vs group answer Measured
"Would the wall dock stop you from buying the Glide 2?"
5 of 24
said yes in private, day 3, before seeing anyone else9 of 24
said yes by the end of the day 4 group thread on where the dock would goFour people changed their answer after reading renters describe drilling into a wall they do not own. Both numbers matter: the private one is the cleaner read of first reaction, and the group one shows how the objection spreads once it is said out loud.
How to read a markup
Every post was coded. A theme counts a participant once, however often they raised it, so the numbers count people out of 24. Quotes are lightly trimmed. Names are invented first names with a last initial.
Biggest frustration today, ranked first · day 2 Measured
Each person ranked eight frustrations privately; this shows who put each one first. n=24.
10 points across six features · day 4 Measured
Average points out of 10, private. With 24 people, treat this as a direction to check. A survey would size it.
A board produces a lot of material quickly. The deliverables are built so the product team can follow along while it runs and act on a short answer the week it ends.
What Corvane took away
Lead with the battery.
Rethink the dock.
Runtime worry shapes how 18 of 24 people clean, and the swap-in battery was the most liked area of the board. The touch-free bin earns its place if the pack shows it does not jam. The wall dock puts off more people than it wins over: make a floor stand the default. Replace "smart suction" with words that say what it does.
What you get, and when
Recommendations, in order
Next step
The board found what to test. The survey sizes it.