At Avenir, we’re optimistic about the future of software- but keenly aware that the sector has struggled lately.
This 45-page deck is our answer to a question echoing through the venture ecosystem and beyond: “What’s Wrong with SaaS?!”
E-commerce in the US is in crisis: penetration peaked in Q2 2020 and has been stagnant since at ~16% of retail spend. Where will it go from here? In this thread, I’ll break down the bull/bear arguments, discuss a potential POV and share a link to the model used to derive it 🧵:
I don't know who needs to read this, but a friendly reminder that the typical scaled SaaS company is probably worth ~8x revenues at "steady-state," which is a decade of >15% growth away for most.🧵👇
I took some time to write up my public SaaS framework and honed it down to three "negative factors" to avoid and two "positive factors" to seek. I'll post a separate thread on each in the weeks to come. For now, here's a working outline of the five things I watch for🧵👇
Weekend research results.
30/72 companies in the
@BessemerVP
Emerging Cloud Index have at least one of:
Indian CEO
Indian CTO/CIO
>10% of headcount in 🇮🇳
No other country is close. The interconnection between the 🇺🇸/🇮🇳 tech ecosystems is deep, growing and crucial for both.
Yesterday, I posted an outline of my public SaaS framework. As promised, I'm going to go through each of the five factors in detail with a mini-case study. First up: what is TAM saturation and what are the warning signs for a SaaS company at scale? 🧵👇
Just built this Hedge Fund name generator in about 10 minutes using
@OpenAI
chat. 🤯 Didn't write a line of code, I just went back and forth with it a bit and pasted the output into
@Replit
. Incredible. Here are the prompts I used, typos and all 🧵:
Last week, I published my public SaaS investing framework. This is the second of five threads diving deeper into each of the factors: when is it right to worry about competition hurting scaled SaaS companies? 🧵👇
My last thread on steady-state SaaS valuations got lots of thoughtful pushback, which I'm going to respond to collectively for efficiency's sake. Here's a thread on what the SaaS doubters have wrong: 🧵👇
It's done! Here's the final thread in my scaled SaaS investing framework series. I saved my favorite topic for last: how to recognize extensible products that will create durable growth, with a few famous case studies. 🧵👇
A few weeks ago, I published my public SaaS framework. Here's the third of five threads diving deeper. When is it right to worry about obsolescence risk? I'll cover the "product model superiority" of SaaS and discuss which categories are vulnerable to obsolescence. 🧵👇
Summary of the seven notable downturns in the BVP Emerging Cloud Index (including this time around).
So far, this downturn is pressing the bounds of normal, but still normal.
Mini-🧵 on four things that are different this time and a thought on how it may end:
Stock comp has been on investors’ minds lately, especially in SaaS. It’s a complex and fraught issue, so I dusted off my blog with my first new post in years.
Here's a quick 🧵 on stock comp in SaaS given this year's downturn, with the link to the full post at the bottom:
I'm seeing lots of comments on LTV:CAC and "unit economics" in SaaS.
Both are helpful mental models- but I see them cited as if they convey some sort of absolute truth.
Models and metrics are to be used, not believed.
Some examples of how investors are using them wrong:
ServiceNow is such a beast of a company. My nine favorite quotes from the earnings call, in no particular order:
"We have 11 individual product lines with north of $250 million in ACV."
"In Q4, our gen AI products drove the largest net new ACV contribution for our first full
It's amazing how stock prices change fundamental narratives.
Just over a year ago, SaaS was eating the world with a bulletproof business model.
Now, it is brutally competitive with no moats, high up the S-curve, will never be profitable, etc.
All because rates went up. 🤔
There's a small public company in Poland that has a lot to say about today's key SaaS debates.
It IPOed in 2014 with $6.2m of ARR (!!) and has since paid out 1.4x its IPO price in dividends (!!).
Text S.A. (aka Livechat) is a huge outlier in public SaaS.
A breakdown 🧵:
This one was fun! Continuing my SaaS framework series: a thread on how TAM tailwinds drive excess growth for scaled SaaS companies, with a case study on $SHOP and a list of the top five trends I see putting wind into the sails of software companies today.🧵👇
Some personal news: I'm pursuing an exciting new opp and on leave this summer. One upshot is I can rejoin fintwit. I've learned so much here (thanks
@AltaFoxCapital
,
@jamesjho_
,
@GavinSBaker
,
@DanRose999
et al) so I'm excited to jump in. My first🧵is on $ZM + $FIVN 👇
SaaS employment update:
+909 jobs for the cloud index in December vs. +2664 last year
19/66 (28%) of companies actively cutting jobs
Fastest growers: $DBX, $ZS, $S, $FRSH, $SNOW
Biggest declines: $EVBG, $TWLO, $ZUO, $MNTV
Raw data:
Sad, tough times.
After staying consistent with historical trends for most of the year, net SaaS hiring has fallen by over 50% in the past few months- a major shift for an industry that employs 376,000 people today. Here are three key takeaways + a link to the raw data:
Takeaways from $SNOW conference/analyst day:
1) Have to credit Ali’s strategic/PR chops. Delta Lake had lost to Iceberg and a defensive move was needed- buying Tabular was probably the best move on the board, and the timing was great.
2) For all the malaise, $SNOW is building a
7/ First, it is striking that outside of electronics/department stores, not a single category of retail has shrunk in real terms over the past ten years, per Census Bureau data. So far, e-commerce has primarily captured incremental spend rather than disrupting physical commerce.
1/ First, to level set: though macro factors (stimulus, inflation) confound absolute numbers, this chart built with Census Bureau data says it all. Since Q2 of 2020, e-commerce penetration of retail has been flat to down:
I've been doing a bunch of enterprise IT buyer survey work, and one result is both unambiguous and a narrative violation.
Buyers think their existing vendors will benefit from generative AI.
3/ What happened instead is now clear: as the world re-opened, e-commerce faced a triple whammy: stimulus abated, wallet-share shifted back to experiences and, most importantly, “foot traffic” shifted back to physical retailers more than many anticipated.
This is the smartest (and only?) bear case on Microsoft- people anchor to P/E but FCF will be suppressed indefinitely. Much more expensive on FCF and uncertain if metal rental is as high quality of a business as O365/Windows long term.
AMZN, GOOG, META & MSFT: are we ignoring that this massive capex will eventually hit the P&L as D&A, or are we just assuming AI revenues will offset it? Just wondering if we're ready to have this conversation or not
1/ Most SaaS products are quite sticky, and some are effectively utilities- during the pandemic, some travel companies defaulted on their debt but continued paying their SaaS contracts because they couldn't operate without them!
21/ My takeaway from all of this is that e-commerce is far from dead- there is a decade of healthy share gains to go. But logarithmic growth will feel different from exponential growth- competition will be fierce and the industry will need to innovate to target late adopters.
Early signs an investor is building moats:
* Expert calls turn into "catch-ups" (bc they want to trade notes)
* Other investors call you as a "go-to" on some topics
* You often make valuable intros w/o much effort
* Relevant analogies from 5+ years ago pop to mind regularly
I spent some time over the weekend taking a quantitative look at a much-discussed question- what is the market signaling to SaaS companies today about growth vs. profitability? Here's a quick 🧵 on the results:
Update on the tech downturn (BVP Cloud Index):
1) 30 Yr yield is up ~1% in two months, though still at '14-'19 levels.
2) Relative declines have moderated as the downturn broadened.
3) Now the longest/largest decline in absolute terms.
4) Multiples at "average" trough levels.
SaaS Hiring Update:
*2561 net hires in November, down 60% from last year
*8% annualized growth vs. 31% in June
*1 in 4 companies have decreased headcount since October
*Fastest growers: $DDOG, $SNOW, $ZS, $GTLB, $S
*Fastest shrinkers: $MNTV, $EVBG, $TWLO, $DOCU, $ZUO
4/ So the base case for a high-quality SaaS business in steady-state is a 40% operating margin (say 20% COGS, 15% R&D, 15% S&M, 10% G&A) with GDPish growth.
This is a very wonky post. I've spent some time lately thinking about go-forward SaaS returns at the sector level and how those relate to interest rates.
On my math, my expected IRR for a typical SaaS company today, is ~16% (more on how I estimate that below).
That's an 11%
One of the scary things about the tech bubble is that it created a generation of investors who don't seem to under the fundamental connection between asset prices and returns. It's hard to write about without seeming to condescend- but I think some could benefit, so here's a 🧵
Some personal news- I'm thrilled to (finally) announce that I've joined
@steadviewcap
as Head of US to open our office in New York (
#longNYC
). A bit about what I'll be up to since the Steadview twitter handle was literally created just a few days ago:
Like everyone, I've been thinking about how LLMs will impact public SaaS cos. Here are five I think *might* benefit (hopefully not all obvious!):
$TWLO, $SNOW, $WIX, $CFLT, $ZI
Explanations in the 🧵 below.
Note: These are not fully baked/researched, just thought bubbles.
17/ For instance, I assumed 30% of today’s in-person apparel sales will move online, 20% of hardware stores, 10% of grocery, 0% of gas stations, etc. The result was a long-run asymptote of ~27% of total retail sales occurring via e-commerce vs. 16% today.
11/ In summary, fundamentals-focused investing in public SaaS is a function of avoiding negative inflections (competition, saturation, obsolescence) while seeking drivers of more durable growth (TAM tailwinds, extensibility).
Just wrapped up a huge (n=50, 5000 datapoints collected!) IT exec survey on how generative AI will impact headcount/software spend at the very largest enterprises.
I'll be unpacking/posting datapoints for the next week at least, but a few interesting takeaways so far:
1) 34/50
The focus on being "cash flow positive" is causing suboptimal decisions for SaaS startups.
It's become such a catchphrase that folks are pursuing it blindly.
For those with solid unit economics and a shot at market leadership, burning right now is the right thing to do.
Another data-backed narrative violation:
Yes, Snowflake and Databricks compete and Databricks has momentum.
But what matters for Snowflake is that it is the center of gravity for a certain kind of data for as many customers as possible- and Microsoft is the greater threat.
SaaS hiring update. Tough out there.
Headcount is still growing overall, but up just +1.4% annualized in February, down from +30% annualized in June '22.
35% of SaaS companies are now seeing headcount decline, and less than a third are on pace to grow
11/ Third, the logistics infrastructure required for e-commerce to immediately jump to its long-term penetration simply didn’t exist in 2020. Building warehouses and last-mile delivery capacity takes time. The pandemic boom almost certainly bumped into an inelastic supply curve.
13/ The 16% estimate doesn’t hold up to even cursory inspection: it includes grocery, gas and auto, together $2.8Tr of spend, in the denominator. Remove them and e-commerce penetration is more like 29%... but to read that as "e-com share can triple from here" is fallacious.
It's pretty amazing how many SaaS companies are citing macroeconomic weakness with unemployment near all-time lows and the market near all-time highs.
Tough to argue we're in an uncharacteristically tough spending environment overall- it's sector-specific.
Zoom reported results earlier this week and grew 3% Y/y with a 39% operating margin. It was a generally uneventful report, except...
Zoom spent $117m on R&D in the quarter, more than it spent in the 10 quarters leading up to the pandemic, combined.
There's a lesson here about
Still a long way to go for the information revolution:
Global annual spend on oil: $3Tr (35bn barrels @ $86/barrel)
Global annual spend on cloud computing: $120bn (AWS + Azure + GCP)
22/ As ever, it's exciting to be investing in the companies/technologies propelling that along. Secular trends facing temporary walls of worry can create great opportunities.
P.S. Feel free to share your own assumptions using the Google Sheet in the🧵!
Just like that, profitability doesn't matter again for public SaaS. At first glance there's actually a slight negative correlation between EV/Rev and profitability (not statistically significant, though).
I generally believe software incumbents will benefit from AI- but nothing I've seen convinces me that usage-based pricing will contribute to that. If anything, I think it is probably bad for the sector:
1. Incredible amount of nuance in describing it to customers/selling in most
12/ I'm grateful to invest in a sector where downturns generate more excitement than fear. It's going to be a privilege to watch this amazing set of companies mature over the coming decades as an investor.
Well, it's official. Total headcount for the Bessemer Cloud Index has rolled over and was down by 2500+ heads in March and by another 5000+ in April. Both probably underestimates.
If anything close to this pace is maintained, SaaS headcount will be down Y/y by June.
At the start of my career, a portfolio manager taught me a lesson that seems more important than ever today.
I advocated selling an investment I'd recommended zealously. Instead, he bought more and imparted wisdom.
A short 🧵 on investing through nadirs in sentiment:
14/ If you dig in, there is another $1.6Tr of retail spend in categories that are hard for traditional e-commerce players to address. Warehouse clubs/superstores, building products, auto parts, pharmacies/drugstores... some of this will move online, but much of it won’t.
5/ “If it didn’t go online then, it never will.” This argument has some intuitive pull- as a consumer, it is hard for me to imagine shopping online more than I did in the summer of 2020, in the same way it is hard to imagine having as many Zoom meetings as I did then.
The narrative elasticity of public companies is fascinating.
There is almost always a "bear case" that gets amplified after a few bad quarters.
I remember being at a public investor group lunch to discuss CrowdStrike where I was the only bull (out of twenty!).
The bear case
18/ My numbers here aren't forecasts, just illustrative- but if they are even close then e-commerce is >50% of the way to steady-state penetration today. I could argue for 45% or as high as 60%. Here's the sheet if you want to download it/try yourself:
After a slew of conversations where I found myself saying similar things, I took some time to write down my career advice for tech investors who are mid-career- i.e. moving beyond entry level roles but not yet leading investments personally/taking risk:
9/ Discount that $3.2bn back at 10% (a solid prospective return, I'd argue) and this hypothetical co is worth $1.2bn today. 12x sales.
Remember, the base-rate of SaaS companies slipping from "growth" to "steady state" is low. Re-acceleration has been just as likely historically.
My top ten quotes from Microsoft's earnings call tonight:
"It was a record quarter, driven by the continued strength of Microsoft Cloud, which surpassed $33 billion in revenue, up 24%."
"Over half of the Fortune 500 use Azure OpenAI today, including Ally Financial, Coca-Cola
On an EV/GP vs. Rule of 40 chart, many of the cos furthest below the trendline (i.e. lowest multiple given the metrics) compete with Microsoft.
"What's scary about Microsoft is that a company so big can develop software at all. They're like a mountain that can walk." -
@paulg
For no particular reason (😔) some thoughts on what a recession would look like for SaaS. To be clear, macro is above my pay grade, but I try to stay aware that recession is well within the fan of outcomes and, in the long run, inevitable.
First (because it’s been so long and
2)/ Yes, some have meaningful gross churn, but that's often the result of the "bathtub effect." Churn is highest in the first few years of a customer's life, but once the pattern of use is established it falls off. Plus there's customer expansion to offset, often at low effort.
@BucknSF
A temporary hit to productivity from a macro freeze isn't indicative of a wholesale change in unit economics, IMO. Might be some ways to mitigate, but otherwise probably best to trim underperforming employees and focus on customer relationships- one day budgets will be back
20/ Here’s an attempt I made at plotting what that might look like. It results in 5-6 years of ~7% growth in inflation-adjusted e-commerce and ~10% in absolute terms. Physical retail sales do not need to shrink for this chart to materialize, while might be conservative.
Spent the morning refining my thoughts on the $DBX S-1- here's a full breakdown/analysis and an updated estimate of how it will be valued:
#Dropbox
#IPO
Finally put pen to paper on a topic I've been thinking about for awhile:
How to identify and avoid building/investing in companies that sell a dollar for ninety cents, and as a result grow quickly with happy customers.
Easier said than done!
3/ As a result, most major costs for SaaS companies are highly leverageable. Cloud costs decrease over time. If R&D is held flat or cut, the product still improves. S&M is >75% growth-oriented. G&A is a function of the others.
Hosting a casual dinner for a set of public SaaS investors in NY tomorrow at my place (NoHo) and have room for one more (ten total). Good mix of Tiger Cub style folks and MMs. If this is your primary investing focus area and you’re interested in joining, shoot me a DM!
It's been a tough few years for e-commerce.
After popping to just under 18% of retail sales during the depths of the pandemic, there is finally a little progress- penetration in Q4 was the highest since Q2 of 2020 at 16.55%.
A quick thread on the likely path from here 🧵:
Appfolio reported Q4 results tonight, is up 9% after hours and... wow. No position here, but this is why we invest in software:
39% Y/y growth, the fastest since at least 2019.
24% NG operating margins, up from -2% three quarters ago (!!)
My top quotes from the call (No Q&A
2/ Obviously, this has disappointed tech investors and companies alike. Many had mental models in which e-commerce could continue to grow at or above trend after the initial benefit from the pandemic. Not just a one-time bump, but an acceleration of adoption.
Spent some time this afternoon thinking through how AI will penetrate CX ticketing use cases.
Depends a lot on the ticket mix for a given customer- not all CX requests are created equal!
Would love pushback/thoughts/feedback on the below:
Thrilled to be investing in Sigma.
As a Tableau user since 2014, the
@sigmacomputing
+
@SnowflakeDB
combo has been a game changer for how we use data at Avenir- we went from buying a dataset to productizing insights in an hour.🤯
DM me if you're interested in a demo sometime!
Today, we’re announcing that we’ve raised $200M in Series D funding.
Read more about our Series D in
@TechCrunch
, on how we’re building a data workspace for everyone:
And, check out our press release announcement here:
Excited to be hiring an analyst to work with me in New York (
#longNY
). It'll be an inclusive/open process. If you or someone you know is looking for an entry-level/entrepreneurial/fun crossover tech investing role, take a look. Application and details:
I recently spent a week in 🇮🇳 featuring a slew of startup meetings, a ten-mile run in Mumbai, a walk through Chandni Chowk and tons of time with the
@SteadviewCap
team. I came away feeling better than ever about the prospects for Indian tech. My five biggest takeaways: 🧵
4/ Everyone knows that’s where we are, but where from here? The bear case, as stated by various smart investors I’ve spoken to in the past few weeks, is that the pandemic gave us valuable and underwhelming information about the potential long-term penetration of e-commerce.
I've seen lots of anecdotes on used car prices lately and... wow. This has to be the most shocking FRED chart I've ever seen. There are bizarre things going on in the physical goods economy right now.
5/ There aren't too many examples of this in the field today because:
1) Most SaaS companies are still growing way faster than the steady-state 5%.
2) It's most comfortable to get to these margins under PE ownership, not in the public eye.
10/ This is the sort of math that experienced SaaS investors have internalized/accepted to the point where drawdowns like last week don't phase them. Oh, and it doesn't factor in any net cash generation over the next 10 years- that's incremental.
I spent some time this week thinking through the impact of crypto on SaaS. I'm always on the lookout for companies that might disrupt the existing SaaS paradigm I've written about, in which most scaled companies are pretty safe from disruption. Crypto might change that! 🧵👇
SaaS cos that have added headcount this year (ranked by growth):
$SNOW
$S
$AI
$DDOG
$TOST
$NET
$WDAY
$NOW
$NEWR
$ZS
$OLO
$FROG
$ASAN
$DBX
$AMPL
$PCTY
$TENB
$FIVN
$CWAN
$CRWD
$CFLT
$SHOP
$SUMO
$MNDY
All others shrinking.
Also, 7 of the names above stopped growing by March.
6/ For example, consider that $MSFT, the "mountain that codes" is growing 15%+ at 35%+ operating margins and is nowhere near steady-state, decades after its founding.
Starting to add more vertical-specific companies to the dashboards and at first glance, the vertical software premium is alive and well (first chart).
However, it is entirely explained by higher profitability.
Run the same scatter plot using the Rule of 40 (second chart) and
8/ At 40% margins, it generates $160m of operating profit, say $128m of net income after taxes. Apply a 25x multiple for what is, again, a utility-ish business that should still be growing earnings healthily, and the business is worth $3.2bn. That works out to 8x sales.
15/ Even in core e-commerce categories like apparel, it’s tough to imagine anything close to 100% adoption in steady-state. Have you been to an H&M or a Zara recently? It’s not a pleasant experience (at least for me!), but that’s because they are absolutely mobbed with customers.
7/ TL;DR: SaaS stock-based compensation has become a big issue in the current downturn and companies can benefit from addressing it head-on. My thoughts on how they might do that:
Assets that will very likely go up in value if AI is as disruptive as it seems to be in the next few decades:
* Scarce real estate/land
* Consumer brands which can be run with skeleton crews
* Collectibles/art
Other ideas for sources of authentic scarcity?
As a thought experiment (not saying I believe this), here's a bear case on Microsoft right now:
* Contrary to market perception, AI model-training and (to a lesser extent) cloud computing are not fantastic businesses relative to the core software franchises- they are
This is so well put. Even talented generalists have to be careful with how they structure their top-of-funnel idea generation.
Generalists with FOMO = hype chasing, not enough conviction to see sectors through downturns, and (often) pain.
@callerycap
Both can be successful if you utilize the advantages of each. Big mistake I see generalists make is going into hyper competitive areas against specialists vs using flexible mandate to find low hurdles.
8/ You can argue this point both ways- on one hand, we may be experiencing the reality that consumers like in-person shopping more than we suspected. On the other, e-commerce innovation continues and it’s hard to imagine physical retail not shrinking over time as a result.
9/ Second, it feels counter-intuitive to suggest that e-commerce penetration is even distributed in the way we’d expect it to be if it were asymptotic. There are still age/geographic differences, and international examples like South Korea where penetration is much higher.
9/ 1) The market the company serves is growing, underpinning its growth rate. For example, it's been a great decade to be selling tools to companies that sell online. As e-commerce GMV grows, so does the addressable market for any company enabling it.
1/ The (now thoroughly understood) reality is that SaaS is a great business model and, at a certain scale, remarkably predictable. Net retention and sales efficiency change slowly in most cases, leading to a fairly narrow cone of uncertainty.