By industry
Seventy eight percent gross margin, 1.5 percent churn a month, sixty two percent of revenue under contract and eleven months of a customer's revenue spent winning them. Margin is not the constraint here. The interaction between retention, contract length and the cost of acquisition is.
Every number on this page is a default. The first document you upload that contradicts one replaces it.
At seventy eight percent margin, a lost customer costs you almost the whole invoice.
That is the thing high margin businesses forget. When cost of sales is small, revenue and contribution are nearly the same number, so churn is not a percentage, it is a direct subtraction from profit. It is also why the elasticity of 1.1 is the mildest in the table and still matters: you can move price without losing much volume, and the volume you do lose is expensive.
| Starting point | Default | What it drives in a run |
|---|---|---|
| Gross margin | 78 percent | Nearly all of a lost invoice is lost profit, and nearly all of a won one is won profit. |
| Monthly customer churn | 1.5 percent | About sixteen and a half percent of accounts in a year. Compounds, which is why small changes here move the twelve month answer more than anything else. |
| Price elasticity of demand | 1.1 | The mildest of the four. A ten percent rise is modelled as about eleven percent less volume before contracts and switching costs soften it. |
| Revenue under contract | 62 percent | The highest share of the four. Most of your book cannot leave this month even if it wants to. |
| Average months left on contract | 9 | A price change reaches roughly a ninth of the locked book each month, so the full effect lands three quarters into the year. |
| Share of cost from suppliers | 25 percent | Hosting, data, payments and the tools underneath. Small enough that supplier scenarios rarely decide the year. |
| How hard competitors match a move | 0.45 | Under half your move, once. Packaging differences make direct matching harder than in retail. |
| Months before competitors react | 3 | A quarter of clear air, which is roughly one release cycle. |
| Revenue a head can serve per month | 16,000 dollars | A blunt proxy. See the weaknesses below, because this is the assumption that fits software worst. |
| Cost to a customer of switching | 0.55 | Migration, integrations, retraining and the data that lives in you. This is what carries a price rise. |
| Monthly market growth | 0.9 percent | About eleven percent a year, the fastest of the four. Some of the answer is the tide rather than you. |
| Monthly cost per head | 9,500 dollars | The highest of the four. Hiring scenarios are expensive quickly. |
| Regulatory exposure | 0.15 | The lowest of the four. A price move is unlikely to attract a regulator agent. |
The cost to win a customer starts at eleven months of that customer's revenue, the longest payback in the table. 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
Almost every software question is a retention question wearing different clothes. Because churn compounds and margin is high, a third of a point on monthly churn moves the twelve month answer more than most pricing decisions do, and the sensitivity ranking on a run will usually say so plainly.
Pricing is the second shape, and it is unusually favourable here: a mild elasticity, a high switching cost and a long contract book mean the immediate damage is small and delayed. The catch is in the same sentence. The damage is delayed, so a price rise that looked fine at month three can still be arriving at month nine.
The third is the cost of growth. At eleven months of payback and 9,500 dollars a head, a sales hiring plan is a long cash commitment before anything comes back, and the run carries the ramp, the onboarding months and the attrition that takes one of them away again.
Uploads
In order of how much band they remove.
The highest value upload in this industry by a distance. It replaces churn, revenue per customer, new customers per month and both concentration figures, and it lets the twin build named accounts instead of segment averages.
Replaces revenue under contract and average months remaining. These two set the timing of every customer response, so a wrong value here shifts the whole path rather than its level.
Replaces price level and gives the elasticity something real to work against. Actual realised discounting is usually the gap between the plan and the accounts.
Replaces cost to win a customer, quota per rep, ramp months and sales attrition. Without it, every growth scenario runs on four defaults at once.
Honestly
Capacity as revenue a head can serve per month is the worst fit in the whole table. In software the serving is done by a machine, and headcount constrains how fast you build and support rather than how much you can deliver. Treat the 16,000 dollar figure as a rough support and success ratio, replace it with your own, and do not read a capacity constraint in a software twin too literally.
There is no expansion revenue. A customer is a customer at a monthly value; there is no seat growth, no usage tier, no upsell path and no net revenue retention above one hundred percent. For a business whose whole model is land and expand, that is a material omission, and the run will understate growth from the existing base.
Pricing is one price level, not a packaging structure. Real software pricing decisions are usually about tiers, bundling and what moves between them, and an elasticity applied to a single price index cannot represent that. It can tell you the direction and roughly the size of a general increase, which is still more than most companies have.
There is no funnel and no free tier. Trials, conversion rates, product led signups and activation do not exist. New customers per month is one assumption, not a pipeline.
And the three customer segments, price led, mainstream and anchored, are a behavioural split rather than a cohort model. If your question is genuinely about cohort decay by signup month, this is not the tool.
Start here
Mild elasticity, high switching cost, long contracts. The most favourable pricing setup in the table, and the delay is the part to look at.
Raise priceEleven month payback at 9,500 dollars a head. Cash out for several quarters with a ramp in front of it, which is the part plans leave out.
Hire a teamAttention comes off the first one. The run puts the cannibalisation and the new revenue in the same twelve months instead of separate decks.
Launch a productIn a high margin subscription business, churn is the assumption that moves the year, and a customer list with dates replaces it along with four other defaults in one upload.