Gross margin and NRR matter more than revenue.
Software is sold once and billed forever. The cost to serve the next customer is close to zero, so a good SaaS business compounds on top of a base that keeps paying. The single most powerful signal is whether existing customers spend more each year without any new sales, because that means the product is growing itself. The enemy is churn: a leaky bucket kills compounding no matter how fast the top of the funnel fills. Judge it on retention and gross margin long before you judge it on revenue.
The product is written once, sold as a subscription, and then the real business begins: the customer renews because leaving hurts, spends more as they grow, and that expanding base of near-costless revenue turns into cash that compounds. Every box after the first is about retention. Break the renewal box and the whole chain collapses, however good the product looked at the sale.
Recurring, retaining, high-margin, unit-economic, sticky, self-funding, and sanely priced. Miss the first three and you are not looking at a SaaS business at all, whatever the company calls itself.
Answer these six and you have understood the industry. The rest of the page is the detail behind them.
Software has the best economics of almost any business, for one simple reason.
Software is written once and then sold to customer after customer at almost no extra cost. Better still, customers pay every month or year to keep using it. So a good software business compounds on top of a base of customers who keep paying, and each new customer is close to pure profit. Compare that with every other business on this site: a carmaker buys steel again for every car, a bank risks capital on every loan, an airline burns fuel on every flight. Software's next sale carries almost no cost and no risk. That is why investors pay more for a rupee of software revenue than for a rupee of almost anything else.
The whole model rests on customers staying and, ideally, spending more each year. The magic is when existing customers grow their own spending without any new sales. The enemy is churn, customers leaving. A leaky bucket destroys the compounding no matter how many new customers pour in the top. Notice what this means for how you judge a SaaS company: the sales the company announces matter less than the renewals it does not have to announce. The quiet base is the business.
A business that runs its billing, payroll or sales pipeline on a piece of software has woven that software into its daily work. Its staff are trained on it, its data lives in it, its other tools connect to it. Leaving means retraining people, migrating data and risking breakage, all to save a little money, so almost nobody leaves over a price rise. This is switching cost, and it is the moat of the entire industry. The deeper the product sits in the customer's workflow, the more the subscription behaves like rent on a building the customer cannot afford to vacate.
Three numbers do most of the work: the recurring base, whether it retains, and whether growth actually pays for itself.
The heartbeat is annual recurring revenue, ARR, the subscription base that renews on its own. The single best quality signal is net revenue retention, NRR: take last year's customers only, ignore every new sale, and ask what they pay this year after expansion and churn. Above 110% means the business grows even with zero new customers; below 100% means the base itself is shrinking and new sales are filling a hole. One warning: only genuinely recurring revenue counts. One-time setup fees and project work dressed up as ARR is the oldest trick in the sector, so check what actually renews.
Growth costs money to acquire customers. The test is whether a customer is worth far more over their life than they cost to win, and the arithmetic hangs on retention: a customer who stays eight years repays their acquisition cost many times over, one who leaves in eighteen months may never repay it at all. This is why retention is not just a quality signal but the input that decides whether the whole growth engine creates or destroys money. Cheap growth that never pays back is just burning cash with better vocabulary.
Beginners see an 80% gross margin and assume an 80% profitable business. The gap between those two numbers is where most SaaS disappointment lives.
Gross margin measures the cost of serving customers: servers, support, hosting. In software that is genuinely tiny, hence margins of 70 to 80%. But business software is sold, not bought: salespeople run months-long pursuits of each corporate customer, marketing feeds the funnel, and engineers keep building or the product falls behind. Those costs sit below the gross margin line and routinely swallow most of it. So a SaaS company can be a magnificent 80% gross margin machine and still lose money for years. The gross margin tells you what the business could earn at maturity, once sales spending relaxes. The cash flow tells you what it earns now.
Read the number in reverse too. When a self-described software company reports gross margin below 60%, something that is not software is hiding inside: usually people. Implementation teams, customisation work, consultants billed by the hour. There is nothing wrong with that business, but it does not deserve software economics or a software price, because serving the next customer costs real money and the revenue does not scale independently of headcount. Gross margin is the single fastest test of whether the software story is true.
This distinction matters more for Indian investors than any other idea on this page, because India's listed technology sector is overwhelmingly one and not the other.
TCS, Infosys and the companies in this sector's own examples list, Persistent, Mphasis, KPIT, Tata Elxsi, mostly sell engineering and IT work billed by the person, by the hour or by the project. It is a fine business: clients stay for years, cash generation is excellent, dividends are real. But its engine is the opposite of software's. To grow revenue it must add people, so revenue and headcount rise together, and margin is capped by salaries, competition and the client's procurement department. Software revenue scales without new costs; services revenue is manufactured, person by person. One is a machine that copies itself, the other is a workforce that must be fed.
Three checks. Revenue per employee: a services firm sits near the salary line, true software towers above it. Gross margin: services runs 25 to 35%, software 70 plus. And growth mechanics: does revenue guidance come with hiring guidance? If yes, it is services. The reason this matters is valuation: paying a SaaS multiple for a people-hour business means paying for operating leverage that does not exist. India has very few genuinely listed SaaS companies; most trade abroad or remain private. So when an Indian technology stock is pitched to you as SaaS, run the three checks before accepting the label, because the label is where the premium hides.
Every failure mode in software eventually shows up in one number. This one.
Churn is the share of customers, or revenue, that leaves each year, and its damage compounds quietly. A business losing 15% of revenue a year must replace that before growing at all: to report 20% growth it must actually sell 35% worth of new business, every year, forever, against rivals doing the same. Meanwhile a competitor at 3% churn grows at a stroll. Worse, churn is a verdict on the product: customers leaving in numbers means the software was nice but not necessary, and no marketing budget fixes necessary. High churn does not just slow a SaaS company. It reclassifies it as a treadmill.
Companies rarely print a churn number in bold, so look for it sideways. NRR below 100% is churn wearing a suit. Deferred revenue flattening while reported revenue grows suggests renewals are weakening. A rising share of revenue from new customers rather than existing ones means the base is quietly thinning. And watch the customer mix: small businesses churn far more than enterprises, because a five-person firm abandons tools easily while a bank never does. Two companies with identical growth can have opposite futures depending on who their customers are and how long they stay.
Software valuations look absurd from the outside: multiples of revenue for companies without profits. There is a logic to it. There is also a trap in it.
Investors price SaaS on multiples of revenue because high-retention subscription revenue behaves like an annuity: if NRR is 115% and churn is tiny, this year's revenue is not one sale, it is the floor of every future year's revenue. Combined with 75% gross margins, each recurring rupee is worth a genuine multiple of itself over time. But every part of that logic leans on retention. Revenue that does not renew is just revenue, and deserves no such treatment. The multiple is not a price for revenue, it is a price for the durability of revenue, which is why the same rupee of sales can honestly command 15 times at one company and 2 times at another.
Read a SaaS company on recurring-revenue growth, net retention, gross margin, and free cash flow, because accounting profit is easy to massage but cash is not. The Rule of 40, revenue growth plus profit margin above 40, is a quick health check that catches both failure modes: growing fast while burning uncontrollably, and profitable but stalled. And apply the discipline this site's valuation question demands everywhere else: at a rich revenue multiple, ask what has to go right. Usually the answer is years of high growth and retention, with no stumble. The business model is forgiving; those prices are not.
Every sector reduces to a demand metric, a pricing metric, an efficiency metric, a capital metric, and a risk metric. For saas / software, these are the ones that matter.
What each number tells you, and how to read it.
| Metric | Why it matters |
|---|---|
| ARR (Annual Recurring Revenue) | The recurring, predictable revenue base. The heartbeat metric; high-quality SaaS grows it 30%+. |
| Net Revenue Retention (NRR) | Revenue from existing customers including expansion minus churn. Above 110% means growth without new sales. The best moat signal. |
| CAC (Customer Acquisition Cost) | The cost to win one customer. Rising CAC with flat NRR means deteriorating unit economics. |
| LTV / CAC | The unit-economics ratio. Above 3x is healthy; below 2x means burning money to grow. |
| Gross Margin | Software efficiency. 70-80% is normal. Below 60% points to a services drag. |
| Churn Rate | Below 5% annual churn is a sticky product. Above 15% is a leaky bucket that kills compounding. |
| Rule of 40 | Revenue growth plus FCF margin. A score above 40 is a healthy balance of growth and profitability. |
| Free Cash Flow Margin | Real profitability. Accounting profit can be massaged; FCF cannot. Positive FCF means self-funding. |
| Revenue per Employee | The fastest SaaS-versus-services test. Near the salary line means people-hours; far above it means the software is doing the work. |
Software costs almost nothing to serve one more customer and gets paid every month, so sticky, recurring revenue compounds beautifully.