StoreGround floorG07 · Personal project
What does the next store cost?
Opening one flagship is a project. Opening thirty stores in three sizes takes a model: every store teaches the crews something, bigger orders buy cheaper, and the street outside decides which size can pay its way. This is my store rollout model, rebuilt from scratch with anonymized numbers. Play the short version right here, or download the full workbook.
Play the store rollout model
Build each store
Every size comes in three recipes. Lean skips the showpieces, Balanced is the default, and Showcase adds premium finishes and the big screens. Then flip any line on or off.
Flagship
3,400 ft²First store–
Standard
1,650 ft²First store–
Small
900 ft²First store–
Switchboard
The lines the recipes change. A switch you flip beats the recipe.
Grow the chain
Open stores over five years and watch each one cost less than the last. Labour gets cheaper with practice, and things you buy get cheaper in bulk.
2, 4, 6, 8 and 10 a year
How fast costs fall
Supplier deal
See the numbers
Send it to the streets
A store needs a street that can pay it back. The 78 Melbourne streets from my Tableau dashboard take the same test here, and the ones that pass decide how many stores of each size you open.
Who picks the store sizes
Send to Tableau
Type these into the dashboard's build cost sliders, which move in $50,000 steps:
Streets that pay back, out of 78
See the numbers
Does it pay?
The whole rollout over ten years: building and stocking the stores and running the program, against what the stores earn.
See the numbers
Six challenges 0 of 6 done
The same six as on the workbook's Start sheet. Every one of them can be done with the levers above.
This is the short version: recipes, switches, pace and streets. The full workbook has every lever, from store sizes, rent and staff to margins and a typed amount for any line, plus all four street tests side by side, every store's P&L, a sheet to compare runs, and notes on every assumption.
What makes the next store cheaper
Every line of a store's build sits in one of three groups, and each group gets cheaper in its own way as the chain grows.
- Learning. Mostly labour: site preparation, walls and joinery, cabling, fitting rooms, systems integration, local design work, hiring and training. Each time the number of stores of a size doubles, these lines cost 10% less, down to 70% of the first store. A crew that fitted out a flagship has learned something for a small store too, so stores of the other sizes count for half.
- Price breaks. Things you buy: flooring, lighting, fixtures and the screens. The discounts start at the 3rd, 8th, 15th and 30th store: 3%, 7%, 10% and 14%. Commit to the whole plan up front and every store gets the discount for the full order from day one.
- No scale effect. Air conditioning and fire protection, storefronts, permits and launch events cost the same every time.
The Cautious and Aggressive outlooks move those numbers both ways. Running costs scale too: the content platform licence gets cheaper per store as the chain grows, and one field technician looks after eight stores.
Where the streets come in
The model and my Tableau dashboard use the same 78 streets, the same rule for sizing them and the same store economics. The model hands the dashboard each size's first-store build, rounded to the dashboard's $50,000 steps: $1.4M for a flagship, $800,000 for a standard store and $450,000 for a small one. At those numbers 59 of the 78 streets can repay a store in three years.
The dashboard hands back the mix those streets add up to: 24% flagship, 49% standard and 27% small, against the illustrative 30/50/20 I started with. That mix sets the sizes in the rollout: 7 flagships, 15 standard stores and 8 small ones out of 30.
Then the loop closes. Once the chain is built, its stores cost 13 to 15% less on average than the first of each size, and at those costs 5 more streets pay back. A full store P&L, with rent that rises with traffic, staff, energy and screens to keep fresh, is stricter: 32 pass. Same store, two streets puts the two halves side by side.
What's in the full workbook
Ten sheets and 8,137 live formulas, with no macros. Yellow cells are yours to change, and everything else follows.
- StartHow to play, the six challenges and your results so far.
- Control PanelEvery lever and the headline results.
- SwitchboardAll 32 lines of a store's build for each size, with on and off switches and your own amounts.
- ScaleHow learning and price breaks cut the cost of each new store.
- Store P&LWhat one store earns: sales, margins, running costs and payback.
- MelbourneThe 78 streets, four payback tests, and the values to type into the Tableau dashboard.
- RolloutEvery store in opening order: its size, build, street and profit.
- ProgramThe whole rollout year by year: investment, profit, cash, payback and value.
- CompareRecipes and outlooks side by side, and a table to save your runs.
- NotesHow it works, every assumption, and a glossary.
Download the workbook (.xlsx, 198 KB)
Where it comes from
At Under Armour I built a scenario model in Excel to size a chain-wide rollout across store tiers: how cost per store and cost per square foot fall as you open more, against break-even benchmarks. This model is built on that thinking, from scratch, with new line items, new store sizes and new numbers. Every cost in it is illustrative, and none of Under Armour's numbers are here.
The version on this page runs the workbook's rules in JavaScript and gives the same numbers, so you can play without downloading anything.
What's next
The Python that cleaned the Melbourne counts goes up on GitHub, with every data decision.