Illustrative scenario · Case study
An $800M bicycle manufacturer is told to grow ten percent. Six executives have six answers, and every one of them is credible. This is what happens when all six are priced against what the factory can actually build.
Northstar Bikes is a digital twin built for this case study — a complete working model of a bicycle manufacturer, with its plants, product lines, capacity ceilings and people all constructed, so the method can be shown end to end on a business we can show you every number of. The method is real. The company is modelled, and there is no real business behind it.
The business
Northstar Bikes started in a Portland garage in 1985. Forty years on it ships to four continents on three product lines.
Every unit crosses an ocean, lead times wander, lithium and aluminium prices swing, the euro moves, subsidies change by country, and three factories each have their own ceiling. The product is simple. Running it is not.
The board meeting
Grow ten percent. Take last year's $731M to $804.5M. The plan locks in four months, and the organic trend — what happens if Northstar simply keeps doing what it does — lands around $747M.
$731M → $804.5M · +10%, board-approved
$73M to find, on purpose, before the plan locksMonday
Go all-in on EU e-bikes. Subsidies are expanding and it is our fastest growth market at +22% a year.
Add European marketing spend. Demand is clearly there; we are leaving it on the table.
Sign B2B fleet deals with the last-mile delivery operators. Volume we can close this year.
Raise commuter prices. We did +4% in 2024 and barely lost volume.
Launch the kids' SKU. The aluminium supply is already there.
Open Brazil and Mexico through the Chile distributor we already have.
Every idea is plausible. Every leader is experienced. None can prove their path beats the others, or whether the factory can even build it.
The answer nobody expected
What the room wants
What the data says
The demand is real. The factory cannot build it. Chasing it costs capital and two quarters of lead time to capture a sliver of the upside.
The answer
Not a recommendation. A ranked portfolio, every lever costed against capacity, lead time and price elasticity.
Pure-play e-bikes — the answer the room was most sure of — ranks third, capped by what Hannover can actually build. The CEO does not pick the answer. She picks how aggressive to be.
Why the answer holds up
You win over people who have been burned by black-box forecasts by showing the work, not by asking for faith. Every number on the previous screen can be opened.
ERP, WMS, TMS, CRM and APS, and the spreadsheets and documents around them. Questions are answered in SQL against that data — typed columns, declared relationships, an origin on every row — so any figure can be walked back to what it was calculated from.
Hannover being nearly full is not an opinion — it is a binding constraint in an optimisation, which is why the e-bike path is capped rather than dismissed. Every driver is swung ±20% to show what would have to change to reverse the ranking.
The simulations are seeded, so identical inputs reproduce identical numbers — you can re-run the board's question and get the board's answer. The models are back-tested against history and report how far off they were.
What this is
Northstar Bikes is a digital twin built for this case study. All figures are illustrative: the $800M revenue, the $73M gap, the 87% line utilisation and the four priced paths were modelled, as were the company, its plants and its executives, to demonstrate how the platform reasons.
The method is the product. The questions asked, the way capacity enters as a constraint, the ±20% sensitivity sweep, the seeded runs and the back-testing are how AthenaMind actually works. Point it at your own numbers and it does the same work: ranks the paths, prices each one against what you can actually build, and shows the rows behind every figure it puts on screen.
AthenaMind