The SaaS metrics bankers actually use

Why definitions matter more than usual here

Every technical interview rewards precision, but SaaS metrics punish imprecision in a specific way: several of them sound similar, get confused constantly, and lead to a different conclusion about the same company depending on which one you actually mean. Confusing gross retention with net revenue retention, for instance, is not a minor slip; it can flip your read on whether a company's existing customer base is shrinking or growing. This article defines the handful of metrics that actually come up in a TMT interview and, more importantly, connects each one to why it moves a valuation multiple.

The core metrics, defined

Annual recurring revenue (ARR)

ARR is the annualized value of a company's active subscription contracts at a single point in time. If a customer signs a contract worth a fixed amount per year, that amount is straightforward ARR; if a customer is on a monthly contract, ARR is typically the monthly recurring revenue (MRR) multiplied by twelve. ARR is a snapshot, not a period of recognized revenue, which is exactly why it can diverge from the revenue line on the income statement. A large multi-year prepayment, a significant one-time implementation fee, or a meaningful usage-based component that varies month to month can all cause reported revenue and ARR to tell slightly different stories about the same business. Interviewers care about this distinction because ARR is the metric that revenue multiples in how software companies are valued are usually built on, so understanding exactly what is and is not included in it matters for getting that multiple right.

Net revenue retention (NRR)

Net revenue retention measures what happens to a fixed cohort of existing customers' revenue over a period, typically a year, including upgrades, downgrades, and cancellations, but excluding any revenue from new customers acquired during that period. Take the ARR that a set of customers represented a year ago, track what that same set of customers is paying now (some churned entirely, some downgraded, some upgraded or bought more), and divide the current figure by the starting figure. A result above 100% means the existing customer base is growing in dollar terms even before counting a single new customer, which is the single most prized characteristic in software investing, because it means the business would keep growing from its existing base alone.

NRR above 100% is achievable specifically because SaaS businesses can expand within an existing account: a customer starts on a smaller plan and later upgrades, adds more users, or buys an additional product, and that expansion revenue can outweigh the revenue lost from customers who churn or downgrade. A company with NRR meaningfully above 100% is telling investors its product gets more valuable to customers over time, which is a structurally different and generally more attractive story than a company that has to replace every dollar of churn with a new customer just to stay flat.

Gross revenue retention

Gross retention measures the same cohort over the same period but strips out any positive contribution from upsells and expansion, capturing only the downside: cancellations and downgrades. Because it excludes upside, gross retention can never exceed 100%, unlike net retention. The gap between gross and net retention is informative on its own. A company with 90% gross retention and 115% net retention is losing some customers or seeing some downgrades but more than making it up through expansion within the accounts that stay. A company with 90% gross retention and 95% net retention has very little expansion motion happening, meaning growth has to come almost entirely from new customer acquisition rather than growing existing accounts, which is a more expensive and less durable growth engine.

CAC payback period

Customer acquisition cost (CAC) payback measures how long it takes a company to recover, through gross profit generated by a new customer, what it spent to acquire that customer in the first place. It is typically expressed in months: total sales and marketing spend attributable to acquiring a cohort of customers, divided by the gross profit that cohort generates per month. A shorter payback period means the company gets its acquisition spend back faster and can reinvest that capital into acquiring the next customer sooner, which is why a company with a short CAC payback can often afford to grow faster (spend more aggressively on acquisition) without straining its cash position as much as a company with a long payback period.

CAC payback connects directly to the reinvestment logic in how software companies are valued: a company suppressing its own EBITDA to fund customer acquisition is making an implicit bet that the CAC payback period is short enough, and the resulting customer retention long enough, for that spending to be a good use of capital rather than a value-destroying one. A company with a long CAC payback and mediocre retention is spending heavily to acquire customers who may not stick around long enough to make the acquisition worthwhile, which is one of the more serious red flags an interviewer might build into a case study.

The magic number

The magic number is a rougher, faster efficiency check: net new ARR generated in a quarter, annualized, divided by the sales and marketing spend from the prior quarter (the spend that presumably generated this quarter's new business). It answers a simpler version of the CAC payback question: for every dollar spent on sales and marketing, roughly how many dollars of new annualized revenue is the company generating. A higher magic number suggests the company's go-to-market engine is efficient and can likely support continued or increased investment in growth; a low magic number suggests either the market is saturating, the sales motion has gotten less efficient, or competition has intensified enough to raise the cost of winning each new customer.

None of these efficiency metrics exist in isolation from each other, and interviewers like to test whether you can hold two of them in your head at once. A company could show an excellent magic number (efficient at winning new logos) while quietly having weak net revenue retention (losing existing customers out the back door), which nets out to unimpressive overall growth despite one metric looking great. Reading the metrics together, not in isolation, is what separates a candidate who has memorized definitions from one who understands what actually drives a software business's growth.

Billings versus revenue

One more pair of terms gets confused often enough to be worth a direct comparison: billings and revenue. Billings is the amount a company invoices customers in a period, a cash-oriented figure driven by contract terms (a customer who prepays for a full year in one invoice generates a large billings number in that quarter, even though the revenue from that contract gets recognized ratably over the following twelve months). Revenue is what accounting recognizes as earned in that period under the applicable revenue recognition rules, regardless of when the cash was actually invoiced or collected. Because of this timing gap, billings is often watched as a leading indicator of revenue growth for a subscription business, since a pickup or slowdown in billings today tends to show up in recognized revenue over the following quarters as those contracts amortize into the income statement. This is also the mechanism behind deferred revenue, the liability a company carries for cash it has been paid but not yet earned; a growing deferred revenue balance is generally a healthy sign for a subscription business, since it means the company is collecting cash for services it has not yet had to deliver.

How these metrics vary by customer segment

None of these metrics behave identically across every type of software company, and interviewers sometimes probe whether you know that a headline number needs context about who the customers actually are. A company selling primarily to small and mid-sized businesses (SMB) typically has faster, cheaper sales cycles, which tends to produce a shorter CAC payback period, but SMB customers are also more likely to go out of business, get acquired, or churn simply because a smaller company is more fragile, which tends to pull gross retention lower than an enterprise-focused peer. A company selling primarily to large enterprises typically faces longer, more expensive sales cycles (multiple stakeholders, procurement processes, sometimes a multi-quarter evaluation before signing), which raises CAC and lengthens payback, but enterprise contracts tend to be stickier once signed, since ripping out a deeply embedded enterprise system is expensive and disruptive for the customer, which generally supports both higher gross retention and more room for the expansion revenue that drives net revenue retention above 100%. Neither profile is inherently superior; a candidate who can explain why a given metric looks the way it does given the company's customer mix is demonstrating real judgment rather than reciting a threshold.

Connecting the metrics back to the multiple

Every metric in this article ultimately answers one question for an investor: how much can I trust this company's current growth rate to continue, and at what cost? High net revenue retention says growth has a durable floor even without new customers. A short CAC payback and a strong magic number say the growth that does come from new customers is being bought efficiently. A company that scores well across all of these tends to earn a premium revenue or ARR multiple relative to a peer growing at the same headline rate but scoring worse on quality, because the market is effectively paying for confidence that the growth will persist. This is also exactly the vocabulary you need to sound credible pitching a software company, covered in pitching a tech stock in a TMT interview, and it is worth cross-referencing against internet marketplaces and network effects to see how a transaction-based internet business substitutes take rate and GMV for some of these subscription-specific metrics.

Practice question

A company has 130% net revenue retention and 85% gross revenue retention. What does that combination tell you?

It tells me the company is losing a meaningful amount of revenue from customers who churn out entirely or downgrade, since 85% gross retention means 15% of the prior period's revenue from that cohort disappeared through cancellations and downgrades alone. But net retention of 130% means that whatever expansion is happening within the accounts that stayed, upsells, seat additions, or customers buying more product, is more than making up for that loss, and then some, pushing the existing customer base's revenue up by 30% overall. That's actually a pretty attractive combination in some ways: the expansion motion is clearly working extremely well. But I'd want to understand why gross retention is only 85%, because that's on the lower end, and I'd want to know whether the same customers churning out are being replaced by better-fit customers, or whether the company has a real retention problem that strong expansion revenue is currently masking. If the churn is concentrated in a specific customer segment, say, smaller accounts that were never a great fit for the product, that's a very different story than churn spread evenly across the customer base. I'd also check whether the expansion revenue driving that 130% net number is concentrated in just a few large accounts, because that would make the growth less durable than it looks at first glance.

What the interviewer is listening for: Whether you can read gross and net retention together rather than citing one number as good news, and whether you instinctively ask what is driving the gap instead of taking the net number at face value.