New data shows that net retention at software companies has been halved in recent quarters, partially explaining the slowdown of revenue growth at tech firms.
This isn’t wholly surprising, since net retention forms a core plank of the SaaS economic model and has been under extreme pressure, as we noted last week. This is because software companies are finding themselves trying to meet two seemingly contradictory asks: Tighten costs and stop letting growth slow too much while your existing customer base reins in spending.
The Exchange explores startups, markets and money.
Read it every morning on TechCrunch+ or get The Exchange newsletter every Saturday.
If you need a refresher: Net retention (aka, net dollar retention and net revenue retention) is a measure of how much existing software customers spend on your product over time. The metric is normalized to 100%, which indicates that a software company’s existing customers are spending no more and no less than they did before. Net retention metrics over 100% tell us that existing customers are spending more, while anything less than 100% signifies a fall in total spending.
Enterprise software companies are expected to enjoy net retention comfortably above 100%. The higher this metric, the better, because if you can land customers that continue to spend more on your product over time, your company not only buys revenue with sales and marketing spend, but it also nets future growth. And since software revenue tends to be high margin by nature, that boost to revenue brings with it gobs of gross profit that can offset costs.
In other words, declining net retention not only makes the SaaS economic model dicier than it was before, but it also means software companies will find it harder to lose less money and keep expanding at the same time.
Now to the new data. According to Altimeter investor Jamin Ball, median net retention at public SaaS companies has followed the following curve in recent quarters:
- Q1 2021–Q4 2022: Between 120% and 121%.
- Q1 2023: 116%.
- Q2 2023: 111%.
As we are more interested in how far above 100% these numbers are, this decline from 120% to 111% is not a difference of just 7.5%, but a shocking 45% fall over just two short quarters. It appears the trend we detailed last week was described accurately, and it was uglier than expected.
Worse, as we are discussing median net retention rates, we can assume that at least half of all public software companies were under the 111% mark. We’ll get more data as companies continue to report their quarterly results, so expect the numbers to move a little, but this does not look good.
Lower net retention, slowing growth and lots of SaaS companies are still in the red. Is software really just not that good a business? I think there’s more nuance to what’s happening here.
Maybe software is too cheap
You can get a subscription to Slack for as little as $7.25 per user, per month. Sure, that’s the cheap tier, but still it’s incredibly inexpensive. You can spend more — a stonking $12.50 per month — for the next tier up, or you can get an enterprise plan like my parent company Yahoo for more features, though I presume it’s possible to negotiate a volume discount at that point.
Frankly, I have no idea how many workers Yahoo has in total, but if you do a little mental math, you can come up with an estimate for what my company, which has revenue in the billions, is paying for Slack. To be honest, Slack is software that we could not do business without. As a manager of a small team, at least I could not.
And we’re probably paying Salesforce the equivalent of a few salaries to keep the whole thing running. That’s insanely cheap.
The situation is great for Yahoo, but it’s not that good for Slack or its own parent company, which is facilitating a simply massive amount of commerce in return for what amounts to a few basis points at most. If Slack were a physical good, it would probably consume more of our total expenditure. Hell, if it was a line item like fuel at a trucking company, it would account for a greater portion of our gross margin.
So when I think about Slack’s revenue growth rate — 20% year-over-year in Salesforce’s most recent quarter — I wonder if it is so low because the market expects to pay too slim a fraction of the benefits it gains from software for the hosted code itself.
How did we end up here? That’s a tricky question, but I wonder if it’s partly venture capital’s fault. “Fault” may be too strong a word, but because net retention rates have historically been high, startups could afford to charge less for their products than they otherwise could have, as any customer would eventually end up spending more. Ask for less now and get more later, all while living off the money VCs have invested. In effect, that will teach customers to expect a piece of software to cost less than lunch for every person who uses it for a month.
But declining net retention blows a hole through that math.
So, how do software companies revitalize net retention, drive prices up and generate more cash?
Enter AI
Of all the numbers that have come out in 2023, I think this one from Microsoft is the most important (emphasis ours):
Today at Microsoft Inspire, we’re excited to unveil the next steps in our journey: First, we’re significantly expanding Bing to reach new audiences with Bing Chat Enterprise, delivering AI-powered chat for work, and rolling out today in Preview — which means that more than 160 million people already have access. Second, to help commercial customers plan, we’re sharing that Microsoft 365 Copilot will be priced at $30 per user, per month for Microsoft 365 E3, E5, Business Standard and Business Premium customers, when broadly available; we’ll share more on timing in the coming months. Third, in addition to expanding to more audiences, we continue to build new value in Bing Chat and are announcing Visual Search in Chat, a powerful new way to search, now rolling out broadly in Bing Chat.
Is $30 per month a lot of money? Yes. Microsoft 365 E3 costs $36 per month by itself for a year, so the introduction of an AI tool like Copilot could all but double that figure. That’s a massive, massive shift in value that will be captured by Microsoft’s productivity suite with the addition of a new feature.
Provided that the market will bear the cost of the new service — and is actually interested in paying for it — Microsoft will have set the bar incredibly high for the amount the market must pay for AI-driven features in the LLM era. For startups and other software companies, such a towering bar for AI pricing grants room to greatly expand how much each customer spends in the coming quarters. It could help rebuild net retention rates and get the metric back on track.
That alone would help make SaaS companies more profitable and spur them to grow faster. What’s more, if new customers also buy larger packages with AI-powered tools as part of the same deals, new customer acquisition could become more profitable. It would be a win all ’round for software companies.
We are currently stuck staring at seemingly bedraggled results from enterprise software companies because their customers are spending conservatively. But there could be good news just around the corner, provided services with generative AI become as powerful as expected and are as in demand as hoped.