Stemly
Redesigning a supply chain forecasting platform
Stemly uses machine learning to help supply chain and finance teams forecast and plan. We replaced a dated enterprise tool with one that people find easy to use.
B2BWeb AppSupply Chain
Product DesignUser ResearchUsability TestingProduct Management
A dated platform
Most enterprise tools in supply chain are old. They are hard to change, slow to react to what happens in the business, and their workflows are so complicated that people struggle to make decisions with them.
No shared picture
- Different mental models: Everyone on the team pictured the product structure differently, and we used industry terms in different ways.
- Complicated workflows: The data science steps behind each forecast were hard to simplify without losing what users needed.
What we did
Three things helped most:
- Fast prototypes: We could change a design the same day, after an internal review or a call with a client.
- A shared glossary: We wrote down the industry terms we used, so everyone meant the same thing.
- Readable charts and tables: We spent the most time on the data views, so people could see what the numbers meant and what to do next.
Result
The new platform did better than the old systems it replaced.
22%
More accurate forecasts than the old systems
40%
Less time spent on each design round
30%
Faster launch than we first planned
45%
Fewer clicks for users
35%
Higher user satisfaction
25%
More productive, as reported by clients