The Bowery Capital team is launching a new interview series featuring some of the leading thinkers in B2B SaaS to get their view on the state of business software and how they see the category evolving. This week, Jared Sleeper, Partner at Avenir sat down with us to answer some of our questions. We will be continuing to publish these interviews over the coming months.
How do the public markets influence your approach towards investing in vertical SaaS? I know you are a student of some of the v1.0 public vSaaS businesses - what about them should early-stage companies attempt to replicate?
I think the v1.0 vertical SaaS businesses fall into two buckets and there are two ways I have seen them succeed. One path is that they built incredibly complicated technical tools for creators or engineering specialists - many of the more under-the-radar companies like Autodesk, PTC, Dassault Systems, Bentley Systems, Aspen Technologies, etc. would fall into this bucket. Many of these are multi-billion dollar companies today. The second path is where companies built software to solve a key problem in their industry. Examples of this approach include Veeva tackling the selling problem in pharmaceuticals, Guidewire solving for insurance workflows, Shopify solving the storefront problem, Toast solving the POS problem, and so on. These companies became the de facto drivers of innovation for their category. Tidemark has written a lot about “control points” - the key points from which other vertical SaaS products can emanate out from. This second group of companies went out and dominated the relevant control point for their industry.
When we think about vertical SaaS startups, it is important to remember that there's nothing inherently magical about vertical SaaS - the success has come from either building a tool that is incredibly complicated and hard to replicate and entrenched or from building a core platform that can own a control point from which you go out and add additional innovation over time. For early stage companies, going after one of these two paths is advisable. It’s relatively easy to get your start selling a point solution into a vertical, but ultimately if you don't own the relevant control point or you are not building an incredibly complicated point solution, I don't know that over time you will be able to benefit from the positive vertical software effects (high margins, growth duration, etc.) that everyone knows and loves so much.
Other than payments, which is often the control point many founders/investors default to, what other control points do you think are relevant?
We think a lot about this - it really comes down to looking at the key problems industry by industry. For example, in pharmaceuticals it is less the R&D that’s a big problem from a financial perspective - the problem is on the sales side; for pharma companies there is an incredible benefit to selling incrementally more of a given drug - they are +90% margin so it is no shocker that the dominant vertical software company in the category is one that provides a tool that helps you sell more of the drugs you have already developed.
In the case of Guidewire, if you look at the insurance category, billings, claims, and policy underwriting are each independent control points. You would be amazed how important just having a really good billing system is to an insurance company. For Toast, I would not characterize payments as the control point - the control point is the POS system. If you own the POS system, you will own payments but for a restaurant the POS is also the crux of the entire operation. In banking software, companies like Ellie Mae discovered that the control point for mortgage origination was sharing data across vendors. A typical mortgage application involves a series of vendors working together, so the real challenge is ‘How do you move data across them in a seamless way without people sending each other PDFs.’ I think Bloomberg has found that the control point for their financial data platform is the chat function. Being the system that people use to chat with each other is actually a killer app and they have a ton of other products and data that they sell on top of that. But nothing is probably as valuable or incremental to their business as Bloomberg Chat which gives them a license to do the rest.
Control points really do end up being vertical specific, it depends on what the biggest problems are in those verticals. Sometimes it's intra-company communication, sometimes it's selling, sometimes it’s billing.
In your recent report on the state of software one theme that jumped out at me was the fact that no public companies are growing revenue more than 35% YoY - do you think there are any factors that could reverse this revenue deceleration or do you think it’s a maturation of the category where there is nothing to be done to reverse it?
In short, no - I don’t see it reversing. I don’t think that is too surprising as most software companies tend to see their revenue growth decelerate over time. Revenue re-acceleration stories, even on an individual company basis, are pretty rare. If you took a cohort of companies, which is basically what today's public software companies are, to assume that they would re-accelerate overall seems unwise.
To be clear, there is still a universe of private companies that are growing faster than 35% YoY. Dozens of them, even scaled ones, and many of those are big enough that they could be public. We can get into why they aren’t public later in the discussion - but if they IPO’ed there could easily be 15+ public software companies growing at 35+% YoY. These are companies that are earlier in their lifecycle, but they too will likely decelerate over time. So, at least part of the reason why we have this dynamic in the public market is that we have not had many IPOs since 2021, and the public companies that came before have just naturally decelerated.
The other reason we have this dynamic is that the recent revenue deceleration in SaaS has been faster than in the past. That can be attributed to a wider sector maturation that we are seeing, where there was a pull-forward during the pandemic which probably jumped us as a sector over the steepest part of the S-curve. There is still a lot of growth to come - it is just going to be slower than it was before.
Stock-based compensation has been the bane of public market SaaS investors for a number of years now, why do you think these levels of SBC have persisted (i.e., the trend of >3% SBC dilution among your report's basket of SaaS companies) - do you see it changing in the next chapter of SaaS?
Obviously it makes sense when you are compensating early employees at a startup to give them significant equity, a sense of ownership, and a chance to achieve a massive outcome if the startup does well. This equity can compensate early-stage employees for taking lower cash comp and arguably more career risk by joining a startup. The question is, how has that spread to the point where at big tech companies, many software engineers get 50% of their total compensation in the form of stock.
I think there are two big reasons why SBC has proliferated. The first is that investors have historically discounted equity compensation as an expense relative to cash compensation. They often operate off of non-GAAP numbers and I would guess that most investors are not modeling in as much dilution as they typically end up facing. Certainly private market investors backing late-stage companies almost never attempt to estimate the ongoing rate of dilution and rarely do the math to convert that to a GAAP figure. So you might hear that a late stage private company has a negative 20% EBITDA margin. But that's a non-GAAP EBITDA margin - if you actually did the accounting to measure what their margin was on a GAAP basis, that equity comp may put them closer to a negative 100% EBITDA margin. I think there is a structural psychological fact that these costs are a little bit hidden. Your non-GAAP margin is kind of like the gas price that’s posted and easy to compare across time periods, and GAAP margins are a little bit more volatile and hard to understand - so they just aren't treated with the same seriousness. By using SBC to compensate their team, companies are paying employees using a cheaper form of currency in the eyes of investors than if their total comp package was cash-based.
The other factor which I think has more economic weight to it is that by using SBC as a compensation lever, companies can compensate their employees without burning cash, which gets them to cash flow break-even faster. It’s important to understand that for a lot of startup founders and management teams, that moment when they never have to raise money again is a massive sigh of relief. The higher a percentage of stock that employees are willing to accept as compensation, the faster the company gets to cash flow break-even. Most of today’s growth-stage companies which are roughly cash flow break-even would be nowhere near that if they had to compensate their employees solely in cash.
So, there is a logic to companies using SBC that will probably persist but there are ways this could break down. If investors started to really penalize companies for having high levels of stock comp and dilution, which would probably take some sort of recession to reset investor mindsets, that could change how a lot of the public companies and late-stage private companies think about stock comp. The other way would be if employees simply started putting a lot less weight on stock comp vs. cash comp. There’s probably some ratio where employees are willing to accept X dollars of stock comp in exchange for Y dollars of cash, and perhaps over time those preferences will change. For now, I think this dynamic is fairly stable and software companies will continue relying heavily on stock comp. It's certainly something that we should be aware of as investors, but I don't expect it's going to change dramatically in the next chapter.
With your view that SaaS has rapidly matured during COVID and we are now in a low-growth macro environment, you point to sales efficiency as something these companies need to prioritize - what does that look like?
I have been pretty open that I am disappointed in the financial management of software companies over the last few years of this most recent cycle. In terms of sales efficiency, I think management teams need to have a really crisp awareness and framework for what is the right level of go-to-market spend for any given company. And they need to clearly communicate this framework to investors.
In the report we put out in June, the cohort of public software companies we looked at increased their collective spend on sales and marketing from $29B to $40B between 2021 to 2024. And incremental revenue actually decreased over that same time from $22B to $18B. While there is a chance that this level of spend might make sense, what I don't see from most software management teams is an articulation of why that degree of S&M spend is justified. In my view, the right answer for many of these companies is dramatically lower sales and marketing investment. My strong intuition is that the curve charting incremental revenue vs. incremental S&M spend is logarithmic - these companies could cut a decent chunk of their sales and marketing spend and see less impact than you might think.
If a given software company is going to generate one thousand super high quality leads per year and today they are spreading those thousand leads out across a hundred sales reps - if they then decide to spread those leads out across 60 sales reps, my guess is the 60 sales reps just end up being that much more productive because they were still high quality leads to begin with right? I wish there was more language and articulation and experimentation to really make the case for today’s investment levels to investors. If the math works, and that incremental spend is justified, that's great and the company should be making that investment. But I think we are very far away from having that conversation as a sector. As we get closer to the top of the S-curve - we need to be having that conversation more and more.
How do you think about the tension between “getting fit” for late stage SaaS companies and the need to sustain growth to entice future public market investors? The PE model of driving efficiency at SaaS companies is often portrayed as having a negative impact on the growth of the business.
I am not sure that getting fit does kill the growth engine. I have a mental model where the way returns to a sales motion scale varies company to company. There are some companies where it is very linear and an incremental $10MM dollars into a sales and marketing motion generates about the same amount of revenue as the first $10MM in a nice, linear fashion. There are public companies like Smartsheet that are a great example of this - Anaplan was also like this, they were never particularly efficient, but the sales motion scaled really beautifully.
Then you have companies where, for whatever reason, there are only so many opportunities per year that are up for grabs and so you don't have that linear trajectory. And if the company is dominant in their sector, it will probably get a look at all of the qualified opportunities. For example, if you are Workday in the Enterprise Human Capital Management market, I am not sure what value an incremental sales rep really adds - at this point you are not educating the buyer - they already know Workday. Any company in that space that's looking for vendors is immediately going to shortlist Workday, Oracle, and SAP.
And there are many smaller software companies that face this same dynamic - especially in vertical SaaS - so it really varies company to company. The decision around sales and marketing investment is more about the company’s understanding where they sit and articulating it to investors, and then having some general view of how to go forward. If their go-to-market produces a linear revenue output, then incremental S&M investments are more likely than not to be worthwhile. If it’s logarithmic - which I suspect it is for many companies - then cutting S&M investments shouldn’t hurt growth too much and will help the market perception of the business when it goes public.
You have stated that you believe SaaS will remain in a GUI rather than a conversational interface - what informs that belief?
I am far from an expert on UI but people can generally read faster than they can listen, and charts are often a much better way to visualize something than having it read to you. When I say graphical user interface, I am really contrasting it with a pure chat interface.This is a silly example, but if you are looking for restaurant reservations - you can scan them super quickly in milliseconds and see what catches your eye vs. having a chat conversation and waiting for it to come back. The idea that we are gonna go back to all text does not strike me as logical. There are other ways to demonstrate information that are equally effective and better at getting certain concepts across. I am actually pretty positive on conversational interfaces - I use ChatGPT everyday like everyone else - but I don’t see chat interfaces replacing something like Excel or a good BI tool as much as augmenting them.
You published your viral piece on “What Gone Wrong in Software and Why We’re Optimistic” in June of this year and it was widely read; how have the developments of this summer informed your thinking since that piece came out? Any beliefs you feel more or less strongly about?
Since we published that piece, there has been a wave of public software company earnings reports and one really striking trend is that new logos added by these companies have been particularly weak so far this year. You would be hard pressed to find a public software company, where new logos are not down ~50% from 2021. And that’s on top of the dynamic of net expansion rates also coming down. So, if anything, that makes me feel more strongly that we are in a phase where we have reached the flatter part of the S-curve. If you are a software company and your new logos are not at least flat over time, it is very hard to maintain compounding exponential growth and the model starts to break. No public companies come to mind where logo growth has been inflecting positively and new logos are up meaningfully. Without that, I think the predictions in the slide deck are that much more likely to come true.
One positive change is the recent advances in AI. I grew up in potato country in Maine and I visualize the SaaS disruption of on-prem software like a plow going through a field - just turning all the potatoes over and pushing them to the surface, so people can pick them off the top. That is kind of what happened when on-prem moved to SaaS, every software category was churned out of the earth and there was room for a new business to be built by copying the on-prem player but in a cloud native, browser-based, subscription model. I am not sure I see that happening in SaaS because of AI, but there has been so much excitement and development on the Enterprise AI side of things that I think there will be a new wave of AI infrastructure and AI application companies. In the same way, when cloud came along, there are a lot of public cloud companies that map to an antecedent company that got disrupted. There is also a large cohort of companies like MongoDB, Elastic, DataDog, etc. that emerged because they were supporting cloud. My perception of how broad the AI shift will be has grown over the past few months, which is really exciting because companies supporting that will be a great investment area.
It may look a little bit different from what we are used to but it’s a good thing for there to be a new wave of companies. Even so, it won’t be as easy as last time around - with AI there is a lot more business model uncertainty and technical uncertainty to wade through. Overall, I think it’s a good thing that there is so much interest in the tech and so many genuine AI use cases popping up everywhere and it will almost definitely lead to some great new software companies.
What areas are you excited about as an investor over the next few years? Any areas you are actively avoiding?
I’m not sure there is any area we’re actively avoiding. I’d say we are keenly aware that in software there can be an ‘easy come, easy go’ dynamic where if you are able to blitz scale from $0 to $10MM or $20MM in a year or two, that’s often because your product is very easy to deploy. And often, if it's easy to deploy, it's also easy to rip out. There are exceptions and there are companies which run this playbook and then become stickier over time. But as an investor, I do not automatically light up just from hearing that a company grew really quickly in today’s world. There are alot of very valuable, easy-to-sell consumer AI tools that will not end up becoming hard to rip out, sticky businesses. It’s not so much an area I avoid, as an area where I put up a little yellow flag and just want to make sure I really understand the path toward building a defensible business.
There are also still pockets where core functions in certain industries have not yet adopted cloud software and there is room for that migration to continue. I have done a lot of work lately in the computer-aided simulation space, which is still a predominantly on-prem market but should in theory benefit from being in the cloud. We are very excited about a few startups there. Warehouse management is another great category with a similar dynamic. Anything where there really has not been that change of the guard in the last 30 years, strikes me as an interesting place for innovation. Sometimes it's because the incumbents are just too entrenched and have already made that transition themselves, but other times, it's just because the buyers move slowly and there is a chance now to capture that shift.
Two areas where hyper growth is possible today are cybersecurity and the cloud data stack. Cybersecurity still has a long secular growth trajectory as a sector and incremental products will come out of that. I think there is also a cohort of new businesses that are waiting to be built on top of cloud data warehouses like how Snowflake and Databricks were. In terms of SaaS applications more generally, I think there is going to be a wave of consolidation back to platforms over time. Today, there are companies like Rippling building out suites of bundled software. I think that is an effective strategy if we are at the top of the S-curve for the wider sector. These companies are building a wide set of tools that are integrated and can replace three or four existing vendors in one shot. That's a really interesting model and I think the right approach for the times.
If you liked “SaaS Perspectives: Jared Sleeper (Avenir)” and want to read more content from the Bowery Capital Team, check out other relevant posts from the Bowery Capital Blog.
