Programming note: we’ll be off next week, returning the 10th - Happy 4th of July, and enjoy the slightly more action packed roundup to make up for the gap in coverage next weekend!
Let’s say I sell you SPF Infinity sunscreen. You apply it liberally during your summer vacation but still end up redder than the lobster on your dinner table. You’ll of course complain that my sunscreen can’t be SPF Infinity. In this hypothetical, I’d respond that of course it isn’t and point you to my own opaque definition of SPF footnoted on the bottle. Fraud is defined as “wrongful or criminal deception intended to result in financial or personal gain”. While fraud in a case like Wirecard - the German payments company in the news this week that “lost” $2B in cash - is straightforward and criminal, how are we to interpret wrongful deception in the case of my SPF definition… or opaque definitions for SaaS metrics?
Let’s revisit our call for accounting-like standards for SaaS metrics. This week we reviewed the most recent annual filing for each of the approximately 65 US-based public SaaS companies we track and discovered 35 different “names” for retention. Parsing through the 35 “names” we found there are really just 3 types of retention (obviously) - net & gross dollar and user retention - with net dollar being the most popular disclosure. But there are 23 different “names” for net dollar retention alone. Worse yet, even the companies that agree on which retention metric to report do not agree on how to calculate that metric.
On the first point, companies call net dollar retention everything from "average quarterly net expansion rate" to "dollar-based net retention rate for organizational domain-based customers". Derivatives like these are harmless until you confuse Benefitfocus' >95% "software services revenue retention rate" for gross dollar retention when in fact the definition is net dollar retention. The word choice can’t be accidental - just say net dollar retention! On the second point, almost 1 in 6 companies exclude from their retention calculations either cancelled customers, all churn or a specific customer segment (e.g. SMBs). For example, Fastly - which has traded up from approximately 7.0x NTM revenue in May to ~30x now and boasts 130% net dollar retention - excludes churned customers from their calculation of NDR. Is that multiple really deserved when we don’t really know what the long term cash flow profile could be? Looking at Fastly you’d think it was better than comp Cloudflare which has 112% net dollar retention… but the latter includes all customers (especially churned)! Finally, 25% of companies choose not to share any retention metrics (they’re not required, of course).
To be clear we’re neither indicting any public software company for fraud nor are we positioning for any type of investment thesis. We do believe though that it is important to highlight the fact that public companies have incredible leeway when telling a story to investors. If Wirecard was able to manufacture its cumulative net income for the prior decade, other companies are hiding things too. There is significant research to substantiate the claim that in strong markets companies have much more flexibility with respect to “pro forma” metrics and fraud increases. For now though - in the case of public SaaS v. OpenView - the disclosures are at worst incredibly annoying.
The takeaway here is don’t buy my sunscreen for summer vacations - but more importantly, don’t take a metric at face value unless you are 100% sure the data and calculation is what you’re led to believe. A wise person once noted everyone carries bias into evaluating a business, the key is to acknowledge and correct for that bias in analysis and interpretation. If pattern recognition for SaaS metrics dominates our perspective on a company and we don’t conduct due diligence we’re exposing ourselves to unnecessary risk. This also holds for operators reviewing metrics internally - the sales team’s view of their pipeline is always going to look the best it can for them - so double check each and every footnote!
Huge thanks to my colleague Dan Knight for his help with the analysis in this post.
Financial & Business Planning Resources
OpenView Product Benchmarks
OpenView’s benchmarking thought leadership continues! Congrats to colleagues Sam Richard and Kyle Poyar who launched the first ever Product Benchmarks Report. Product-minded founders and product leaders already have plenty on their plates. They’re expected to make hard (and often unpopular) decisions and create MVP product offerings while still watching the market, the competition and, of course, their roadmap. Now they’re also expected to implement new metrics and go-to-market tactics. To make the best strategic decisions and be held accountable in their roles, product and growth professionals need objective data (i.e. how much conversion is good conversion). OpenView launched our product benchmarks report to help - these findings allow operators to compare themselves to their peers across the metrics they use most. Think of this report as your go-to resource for building better products.
This week Blake was joined by Sahir Azam, Chief Product Officer at MongoDB. In the episode they discuss how to go from individual adoption to org-wide adoption, MongoDB’s philosophy for monetizing open source software, and how they scaled their open source product using both bottoms up and top down sales motions. You can listen to this and all prior BUILD episodes here.
Capital Markets Resources
The public SaaS companies we track closed Friday trading at a median EV / 2020 expected revenue multiple of 12.1x. This compares with 10.1x at all-time highs in February and 6.3x at March lows. Since February, of course, companies have reported earnings and grown (at the median, the SaaS set grew TTM revenue by 6% between Q4 and Q1) so we’d expect to see growth rewarded in stock price - but not multiples per se. A higher stock price increases the numerator and revenue growth increases the denominator of the EV / Revenue equation. For multiples to expand, investors are paying more than ever before for $1 of revenue, and this expansion (melt up) is even more pronounced for some of the higher growth names (e.g. Zoom). Between the pandemic and recession there is no compelling explanation for why multiples are up ~20% over what were already hot prices in February, but increased liquidity in the sector is a long term positive.
As we near the end of H1 there will be some portfolio rebalancing given the equity performance over the last quarter that could result in some downward pricing pressure (per BofA). This, statements on Trade from Trump advisors, Fed limits on bank dividends, and US virus case concerns all contributed to a volatile in the public markets (SaaS +3.27%, Dow -3.31%, NASDAQ -1.59%, S&P 500 -2.86%). While the NASDAQ has passed its pre-COVID high (up 12.86% YTD), the Dow and S&P still have yet to regain lost ground (-12.19% and -6.77% YTD respectively). The NASDAQ in particular benefits from tech stocks driving the gains, so more broadly the markets (companies in the indices) show impact from the recession.
As Peter Lynch said, “It is consumers who lead the way out of a recession with their car buying and home buying and replacing of old appliances, all of which then creates a revival in the businesses up the line - lumber mills, steel plants, capital goods manufacturers, the advertisers, the newspapers that sell space to the advertisers, and so on." With recession now official and consumers making up ~70% of US GDP annually - individuals are critical to supporting economic growth and recovery in United States and elsewhere. This week we saw European PMI data suggest reopening (which is ahead of the US) is enabling economic recovery. In the US though, as case counts accelerate in certain geographies and mortgage delinquencies increase… a second shutdown would be devastating to the US consumer, prolong the current recession, and throw any forecast for the shape of recovery (V, U, or W) out the door.
What Else We’re Reading
How to Lead Through a Crisis. Coronavirus Dashboard from Tableau. We’re still in a bear market. The Expert Problem (like how the experts at Gartner forecast -11% growth in enterprise software expense, yet a survey from JPM suggests purchasers are saying something entirely different).