Gross Retention Never Lies
Over the last decade there's been a massive proliferation in SaaS metrics that can be used to grade and analyze a cloud business. Arguably there are too many SaaS metrics these days, which opens the door to confusion and mis-use. Try asking any early stage investor what they think about LTV/CAC projections at the Seed stage.
If you ask most people in SaaS what the most important business metric is besides growth rate, you're likely to hear "Net Retention Rate" (NRR) given in response. Now, NRR is incredibly important, particularly at the growth or public company stages where the difference between a revenue base that expands or contracts is gargantuan. A $100m ARR company with 140% NRR gets a $30M additional ARR boost in 12 months compared to one with 110%, a boost that becomes even more important in the following year. Snowflake would obviously not be trading at ~80x ARR without otherworldly NRR (somewhere close to 170%).
With that said, I'll posit the following: For early and mid stage SaaS businesses (Series Seed, A, B, C) there is no more important metric than Gross Retention (or Logo Retention, which is similar though not exactly the same). Whereas Net Retention answers the question "what will my existing ARR base turn into a year from now?", Gross Retention answers the question "how many of our clients like us enough to stick around a year from now?".
One of my partners recently made the statement "Gross Retention basically never lies" and pointed to the fact that the most successful companies in Toba's portfolio (or that we've seen outside the portfolio) had very low Gross Churn (same ideas as high retention) compared to their peer companies, regardless of whether they had super robust NRR or not. Very high Gross Retention is obviously correlated with high NRR, but that's not always the case. You could have a super sticky product with minimal upsell (high Gross Retention but modest NRR), and you could have a very churny product with great upsell (low Gross Retention but high NRR). In our observation, at least for most early stage businesses, the former beats the latter.
Why? I have 4 main reasons for this:
Gross Retention can precede Net Retention. A company with excellent Gross Retention is a company that has captured customer attention and commitment by delivering significant value compared to price. You know that a company in this situation is doing something very right (or is a monopoly). Vendors in this situation tend to have a very receptive customer base who is willing to buy more things from them; in fact they often explicitly ask the software provider if they'd be willing to sell them more stuff. Hence, you can build an NRR story later, even if you don't have a meaningful upsell motion today, as long as the vast majority of your clients want to keep paying you year after year after year.
A high NRR business coupled with low gross retention is hiding a major leaky-buck problem that will only grow as time goes on. I recently saw a SaaS company with 130% NRR and 67% annualized gross/logo retention. This is an awesome way to foolishly pat oneself on the back for your terrific NRR as you literally lose 1/3rd of your entire client base each and every year. A third! Nota bene that the inverse company (high gross retention but modest NRR) does not have a leaky bucket problem, nor is it in a situation where it's likely to deceive itself about how well it's actually doing.
Net expansion cannot continue forever. Let's say you have a company that reports a 150% 12-month NRR. This means that the average $10k account will expand to $15k by the end of that client's first year as a customer. It does not mean that the average account will compound in size at 50% annually forever (to $22.5k next year, to $33.75k the next year, to $50.6k the next, etc etc etc). It certainly could be that expansion continues for many years (usage-based products like Twilio and Snowflake and AWS can see expansion for a very long time) but this is not something that you should assume will continue in perpetuity. In contrast, you could find a nominally strong NRR in a two-product company that has a single upsell product to sell into their base. Once you upgrade everyone onto both products, you have no further upsell story, unless you can find something else to sell them. This is just another instance where NRR can be deceiving.
Net retention can be gamed. Imagine you are a savvy entrepreneur who has noticed that super high NRR companies are valued at insane multiples. You set up a pricing scheme where each client pays you $1,000 for their first year of service but $10,000 to renew for the following year. 50% of your clients say no, which means that 50% of your clients said yes. Voilà, your NRR is 500%, and you just got an iMessage notification from one of the GPs at Tiger Global (just kidding).