A sales history by month with units and revenue
The highest value upload here, because it is the only thing that can give elasticity something real to stand on. A default elasticity of 2.2 is doing a lot of work in a retail twin until this arrives.
By industry
Retail and ecommerce
An elasticity of 2.2, six percent churn a month, five percent of revenue under contract and a competitor who matches seventy percent of your move within a single month. Everything here happens quickly, which makes the decisions cheap to reverse and expensive to get wrong at scale.
Every number on this page is a default. The first document you upload that contradicts one replaces it.
Nothing is contracted, nothing is sticky, and the competitor answers next month.
Five percent of revenue under contract and a switching cost of 0.12 mean the customer base can respond to a price move immediately, and at an elasticity of 2.2 it responds hard. The competitor reaction of 0.70 at a lag of one month is the fastest in the table. A retail price decision is not a strategy, it is a move in a game where the other player goes next.
| Starting point | Default | What it drives in a run |
|---|---|---|
| Gross margin | 42 percent | What a lost sale costs. At 2.2 elasticity, volume moves enough that this number decides whether a discount ever pays back. |
| Monthly customer churn | 6 percent | About fifty two percent of the base in a year. Retail relationships are short by nature. |
| Price elasticity of demand | 2.2 | The highest of the four. A ten percent rise is modelled as roughly twenty two percent less volume, and almost none of it is delayed. |
| Revenue under contract | 5 percent | Effectively nothing. There is no contract book to delay a customer response. |
| Average months left on contract | 1 | A decision reaches everybody immediately. Month one is the answer, not a preview of it. |
| Share of cost from suppliers | 55 percent | Goods bought in. A supplier move lands on margin fast and there is no contract cover to absorb it. |
| How hard competitors match a move | 0.70 | The hardest match in the table. Prices are public and comparison is one click. |
| Months before competitors react | 1 | One month of clear air. Any plan that assumes a quarter of advantage is wrong here. |
| Revenue a head can serve per month | 18,000 dollars | Store and fulfilment capacity as a single ratio. Crossing it shows as service before it shows as revenue. |
| Cost to a customer of switching | 0.12 | The lowest in the table. Nothing holds a customer except price, habit and delivery. |
| Monthly market growth | 0.4 percent | About five percent a year. The tide is real but small next to a competitor move. |
| Monthly cost per head | 4,600 dollars | The lowest of the four, which makes headcount scenarios less decisive than pricing ones. |
| Regulatory exposure | 0.18 | Low. A price move is unlikely to wake the regulator agent. |
The cost to win a customer starts at three months of that customer's revenue, the shortest payback in the table, which is why marketing spend and price interact so directly here. Every value is a replaceable default, drawn with a band on each replication: gross margin within fifteen percent, churn within thirty five, elasticity within forty, market growth within fifty.
The shape of the question
It is a pricing question, and then it is a competitor question
The first shape is a discount, not a rise. At 2.2 elasticity the volume response to a price cut is large, and at forty two percent margin the arithmetic of whether the volume covers the margin given away is tight enough to be genuinely uncertain. That is exactly the case where a sweep across the range beats an opinion.
The second shape is the competitor. With a 0.70 match at a one month lag, any advantage from a price move is short, and the interesting question becomes what the market looks like after both of you have moved. Running hold, match and undercut against the same baseline on the same seeds is the cleanest way to see it.
The third is growth spending. Three month payback means customer acquisition is fast to judge, and the constraint is usually not cash but whether the customers acquired at that price behave like the ones already there. A six percent monthly churn says many of them will not stay long.
- A sweep is more useful than a scenario here, because the profit curve against price is steep
- Month one is most of the answer, which makes the tripwire arrive early and be worth acting on
- Almost no contract cover means a supplier rise has to be passed on or absorbed, quickly
Uploads
The four documents that change a retail twin most
In order of how much band they remove.
Margin by category or product group
Replaces gross margin with something closer to your actual mix. A blended forty two percent hides the categories where a discount can never pay back.
Supplier terms and cost of goods
Fifty five percent of cost is bought in, so supplier cost share, concentration and lead time all matter, and none of them are guessable from a profit and loss alone.
Customer and order data
Replaces churn, revenue per customer, new customers per month and the cost to win one. In a business with six percent monthly churn, repeat behaviour is the whole economics.
Honestly
Where the model is weakest for retail
There is no seasonality. The engine runs a monthly path with no calendar in it, so a business that does a third of its year in one quarter will see a twelve month run that looks nothing like its actual shape. Unless your uploaded history lets the twin pick up a level, read the run as an average month repeated rather than a forecast of any particular month.
There is no catalogue. One price index, one elasticity, one margin, across everything you sell. Real retail pricing is a per category decision and the interesting effects are in the mix. A run can tell you about a general move; it cannot tell you to discount one range and hold another.
There is no inventory and no markdown. No stock position, no reorder, no lead time into availability, no clearance. For a business whose main financial risk is buying the wrong thing, that risk is not in the model.
There are no channels and no baskets. Marketing spend is one cost to win a customer, not a set of channels with different economics. Basket size, attachment and returns do not exist.
And churn of six percent a month is a subscription shaped idea applied to a transactional business. If your customers do not have an identity you track, treat the churn number as a rough proxy for repeat rate and do not build an argument on it.
- No seasonality and no calendar
- No catalogue: one price, one margin, one elasticity for everything
- No inventory, no availability, no markdown
- No channel mix, no basket, no returns
Start here
The three runs worth doing first
Cut price to move volume
The classic retail question, and a genuinely tight one at 2.2 elasticity and forty two percent margin. The sweep gives the whole curve rather than one point.
Reduce priceA competitor undercuts you
Hold, match or split the difference, on one baseline and one seed set, against a rival who matches seventy percent of your move inside a month.
A competitor movesLaunch a range
New revenue against attention taken off what already sells, in the same twelve months, with a competitor who is watching.
Launch a productGive it a sales history before you ask about price
The elasticity default of 2.2 carries more of a retail answer than any other single number. A monthly history with units and revenue is what replaces it, and it is usually the easiest file to produce.