An internal tool that tracks every style from design to ex-factory, scores its launch risk, and tells merchandising, design and production teams whether to launch, mitigate, delay or drop it, weeks before it becomes a problem.
Teams already had the data: milestone dates in the PLM system, trackers in Excel, updates over email. What they didn't have was one answer to "will this style make it, and if not, what do we do?" StyleSignal gives them that answer.
8 standard milestones for every style, from design approval to ex-factory, each with an owner and a due date.
A 0–100 score built from how late milestones are, how much work is left versus weeks to launch, and how critical each milestone is.
Launch, launch with a fix, move to a later wave, or drop, with the reason and the owner who needs to act.
of styles hit their launch date before StyleSignal
average warning before a missed launch, too late to fix
a week per merchandiser spent chasing and compiling status updates
Baseline from a discovery audit of one season (sample data).
"I find out a style is late when it's already too late to swap it."
Job to be done: protect the launch plan and revenue.
"I update the same status in three different places."
Job to be done: update a milestone once and move on.
"Just tell me what's at risk and what it's worth."
Job to be done: make the call on delays or drops in a 15-minute review.
Discovery: 10 stakeholder interviews, shadowing 2 weekly line reviews, audit of 3 existing trackers (sample).
Pick a style, change a milestone's status or days late, and watch the risk score and recommendation update.
If merchandisers see an explained risk score every week, they'll act on at-risk styles 3+ weeks sooner.
Piloted with one collection team for a season, while a second team kept using the Excel trackers. On-time launches were compared at the end.
The MVP runs on a weekly PLM export with transparent scoring rules. The logic table is in the PRD so anyone can audit it.
| Style | Collection | Vendor | Launch in | Score | Recommendation | Next action · owner |
|---|
All data is illustrative sample data, created for this portfolio.
Early in the pilot, the team wanted an AI model to predict delays. I chose simple, explained scoring rules instead. Merchandisers wouldn't act on a score they couldn't question, and trust was the adoption risk. Weekly use reached 86% in the pilot.
The key product trade-offPDFs open in a new tab; Excel and PowerPoint files download. Portfolio case study; all data is illustrative sample data.
Consumer Goods (FMCG)