Pre-PMF
Before anything repeats. The founder is still working out who buys, for what reason and at what price.
Held together by the founder
The commercial model that worked at one stage becomes the problem at the next. The leaks are the same kinds everywhere; which one costs you most moves with each stage.
Six situations. What holds the company back moves as it grows, and each page says where the money tends to go.
Before anything repeats. The founder is still working out who buys, for what reason and at what price.
Held together by the founder
The deals repeat; how they are won is known to two people and written nowhere a new hire could read it.
Held together by conversation
Qualifying splits from closing, managers are appointed, and the job one person did becomes a handoff nobody has defined.
Held together by role design
A working model is copied into new regions, segments or teams. The playbook travels; the decisions behind it have to as well.
Held together by the system
The company becomes a set of departments, each running its own process against its own target. All of them reach it; the plan is missed between them.
Held together by structure
The model is mature and the trend is flat. No crisis, so the loss stays off the agenda for years.
Held together by governance
Each stage hands what holds the company together to the next, and the gap in that handover is where the leak sits. These four figures show how far the ground shifts between one ARR band and the next.
ARR per full-time employee in the top quartile, by ARR band: $90K under $10M, $160K at $10-25M, $215K at $25-50M, $220K at $50-100M, $250K at $100-250M, $280K at $250-500M and $380K above $500M. What one head carries is the first number a stage changes, and the one every headcount plan assumes will hold.
ICONIQ, State of Software 2025, quarterly operating data from 127 software companies, 2013 to Q2 2025, top quartile rounded, September 2025
Total operating expense as a share of revenue at the median, by band: 291% under $10M ARR, 162% at $10-25M, 122% at $25-50M, 114% at $50-100M, 93% at $100-250M, 82% at $250-500M and 76% above $500M. Revenue starts to outpace the cost of the organisation around $100M-$200M ARR, and until then every stage is paying for a structure the last one did not have.
ICONIQ, State of Software 2025, 127 software companies, 314 to 133 company-quarters per band, median, September 2025
Gross dollar retention in the top quartile: 100% under $10M ARR, 96% at $10-25M, 95% at $25-50M, 94% at $50-100M and at $100-250M, 93% at $250-500M, 97% above $500M. The best quartile starts losing revenue it already had the moment a second renewal cycle exists, and the median loses more at every band.
ICONIQ, State of Software 2025, 127 software companies, 351 to 62 company-quarters per band, top quartile, September 2025
CAC payback in the top quartile, by band: 12.9 months under $50M ARR, 12.3 at $50-100M, 13.3 at $100-200M and 14.8 above $200M. Every band past $100M adds time before a new customer has paid for its own acquisition, and that is the quartile with the shortest wait.
ICONIQ, State of Software 2025, 127 software companies, 410, 114, 138 and 115 company-quarters per band, top quartile, September 2025
The five steps, from finding the leak to handing the fixes over, are the same at every stage. What the stage changes is where the first step looks and what the third one finds.
Growth stage says which seam to look at first. It does not decide whether a finding will stand up.
Not sure? Book a strategy call →
Six sectors, six different disguises for the same leak, whatever stage you sit at.
The six leaks, what gives each away, and which line of the accounts it lands in.
Two months in your records and with your people, ending in where the profit leaks, what it costs in your own figures and what to fix first.
Thirty minutes on the stage you behave like, the records on hand, and whether there is anything to price. A proposal only follows if there is.