A good net dollar retention rate for a SaaS company

Enterprise accounts expand faster than small-business accounts, so the retention bar sits higher for them. Didi Azaria, CEO of Workiz, benchmarks enterprise net dollar retention well above 150 percent a year, against 110 to 130 percent for small business. The gap makes sense: large customers add seats, upgrade tiers and grow their own usage over a year in a way a single plumber rarely does. A blended target set between the two flatters one segment and lets the other slide.

A startup should start measuring net dollar retention by segment once each segment has enough customers for the number to hold steady. Workiz first sorted its customers by trade at around a thousand accounts, grouping cleaning, HVAC, electrician, and plumbing to see which paid back. Etgar Shpivak, who works with seed and Series A founders, warns that before then the rate swings on single accounts and is too noisy to act on.

Net dollar retention is the number Azaria uses to judge whether customers succeed with the product, and it sits alongside the metrics that measure acquisition. He watches return on ad spend, the revenue back for every dollar spent on ads, and the value of open sales deals. Etgar Shpivak's answer to which matters most is that a founder needs both: a strong retention number can still hide a customer that costs too much to win.

Early-stage startups should not spend on broad brand building, in Azaria's view. Speaking to Etgar Shpivak, he said a startup burns through its runway before a broad brand pays off, so the move is to build a brand in the smallest segment you can lead. Workiz did that in junk removal, a segment nobody else wanted, becoming the name there while selling to forty trades. That is the same logic as winning one boring segment loudly before the rest.

The instinct that matters most in a marketing leader is whether it fits the market you sell into. Azaria, who has hired ten to fifteen marketing leaders, said you cannot move a B2B enterprise marketer into B2B small business, or a B2C marketer into B2B, and expect the same instincts to carry. He also screens for resilience, since in a crowded market the complaints about price arrive daily. See the warning signs when hiring a marketing leader.

A segment that shrinks while you feed it only shrinks faster once you feed it more. Azaria cuts anything under 100 percent, even though it still shows revenue on this quarter's books and still answers the phone. You give up money now because the number says the gap costs far more later.

Before a segment has enough customers, one unhappy account swings its retention rate more than any real gap. You end up reading that account, not the segment. Azaria's cut-immediately rule holds at Workiz's size, with years of data behind each number; earlier, find out whether the account that left was the segment or an outlier first.

Where the next dollar earns the most. Once every segment has its own number, Azaria runs it as a sorting rule: any segment above 110 percent gets more money and more of the roadmap, and any under 100 percent gets cut. The number, not a hunch, decides where investment goes.

A few points off a small revenue base is money a good quarter erases. At $100 million in revenue, Azaria noted, a 10 percent loss is $10 million the company has to earn back on top of gross. The segment that rounded to nothing early becomes a hole you refill every year.

An early-stage SaaS founder deciding what retention to expect, and how to split it by segment, needs someone who has read these numbers across many companies. Etgar Shpivak, a marketing consultant who works with seed and Series A founders, does exactly this: deciding which segments to back and which to wind down. His bio page covers how he works. The target is set per segment, against who you sell to.

Net dollar retention is the revenue an existing customer base keeps and grows year over year, after expansion and churn net out. Above 100 percent means the base grows on its own before any new sales. Azaria argues it should be measured "on almost every vector possible," not as one company-wide figure.