Software Sector Primer

Software is worth the cash its recurring revenue throws off after the staff are paid, including the part paid in shares, and most of that value sits more than ten years out. Code costs a great deal to write and very little to sell again, so gross margins are high. Value turns on how long customers stay and expand, and on what it costs to win new ones, a cost that arrives years before the revenue does. So a software valuation rolls revenue forward from retention, charges stock pay as a cost even though no cash leaves, and asks what growth a price requires.

Worked example: a DCF on a fictional seat SaaS grower

Take a fictional subscription company with $2.0B of opening recurring revenue, priced per seat and billed a year ahead. It keeps 92% of its revenue base before expansion, and net retention fades from 115% to 102% over ten years while new-customer revenue fades from 12% to 1%. Its margin before stock pay widens from 8% to 32% as sales and marketing falls back, and stock pay falls from 18% to 8% of revenue. The primer runs ten years at a 10.00% WACC with 3.00% terminal growth.

The company burns cash in year 1, turns cash-positive in year 2 and earns most of its value after year 10. The year-10 margin is the strongest lever: ten points either side moves value from $25.46 to $55.56. One point of terminal net retention is worth $4.77 a share up and $3.70 down. The return on new capital barely matters, $38.16 to $41.52 across a wide range, because at 3% growth little profit is reinvested.

The same model values three more fictional companies, one for each of the other business models. A one-year consumption shock that takes the usage-priced platform's net retention from 118% to 103% costs 9.2% of its value. The licence-to-cloud platform is worth $83.71 a share when each dollar of support converts into 2.50 dollars of annual cloud revenue, and $54.51 when it converts one for one. Switching off the acquirer's deals takes it from $106.43 to $69.29.

Stock pay, free cash flow and the Rule of 40: filed figures

Reported free cash flow, operating cash flow less capex, adds stock pay back because no cash left the company. The table takes it off again. The Rule of 40 sum is revenue growth plus free cash flow margin, shown before and after stock pay. These figures are as filed, each for the fiscal year shown; SAP and Constellation report under IFRS.

Stock pay moves the order. Snowflake's sum falls from 53.1 to 19.0 once its stock pay is taken off, while Microsoft's moves only from 38.0 to 34.2. The sums are filed arithmetic, not a ranking of the businesses: a sum says nothing about how long the growth will last or what margin the company will reach.

On the fictional companies the 40 line can be drawn, and it shows the rule's blind spot. The seat SaaS grower's sum after stock pay runs from 25.7 in year 2 to 24.3 in year 10, below 40 in every year of its forecast, yet the model values it at $40.51 against an illustrative price of $40.50. The rule screens one year; value depends on the whole path.

Retention and backlog: what filers disclose

There is no standard definition of retention, so two companies printing the same number may be measuring different things. The basis, the window, the customers left out and whether expansion counts all vary.

Where retention is missing, remaining performance obligations are the next disclosure to read, and the part due within twelve months matters most for next year's revenue. Salesforce's current RPO grew 16% to $35.1B in FY2026 against revenue growth of 9.6%. Microsoft's total RPO rose from $375B to $684B in FY2026, but only about 30% falls due within twelve months, so most of the increase is revenue for later years.

Which model values which software company

What the full primer adds

The primer builds from how software makes money and the four business models to the recurring-revenue roll-forward, retention and backlog disclosure, the cost lines and sales efficiency, stock pay and the Rule of 40, then the discount rate. Four fictional companies are valued from first input to value per share: a seat SaaS grower, a usage-priced platform, a licence-to-cloud platform and a vertical-software acquirer, with sensitivities and a reverse DCF. Filed figures for six listed companies follow, each with its fiscal year, then ten screens, the red flags and two contrasts: Snowflake against Constellation on how growth is paid for, and Salesforce against Microsoft on capital.

Free guides cover the individual pieces: Net Revenue Retention: How Filers Define It, Gross vs Net Revenue Retention, Usage-Based Pricing and Net Retention, Rule of 40: Which Margin Goes In, Stock Pay as a Share of Revenue in Software, Software FCF Before vs After Stock Pay, RPO and cRPO in Software, Current Share of RPO by Software Company, Licence-to-Cloud Transitions and the Revenue Dip, Reverse DCF for Software Stocks and Vertical Software Acquirers: Return on Acquired Capital. Research profiles for Microsoft, Salesforce, ServiceNow, SAP, Snowflake and Constellation Software read the same measures from filed results.

The companion Excel model spans thirteen sheets. It runs the four fictional companies, and a column for a company of your own in any of the four business models, through the same ten-year schedule: retention and new-customer revenue roll recurring revenue forward and fade to the terminal levels that set perpetual growth, margins widen in a straight line, and stock pay is charged as a cost. Change an input and the valuation rebuilds from the first year; the reverse DCF reads the growth a typed price requires, and the sensitivity grids re-run the model in every cell.

Sheets: Quick Start, Instructions, Dashboard, Assumptions, Inputs, Engine, Value, Reverse DCF, Sensitivity, Efficiency, Licence to Cloud, Acquirer, Grid Runs.