"All models are wrong." George Box, the statistician is supposed to have said. Unfortunately, like he also continued to say, some are useful. Whenever I try to take a step back and think about the broader investing landscape, that remains true.
It's extremely difficult to speak of any broader trend or have an insight into a phenomenon if we get caught up in the specifics and don't step back to see the bigger picture. Forest, trees, and all that.
Which brings me to investing. There's a generic thought process within investing circles, which goes something like this - "in consumer companies, you'll have a ton of competition, but you'll have to outfox them with better experiences and clever marketing. In enterprise companies, you'll have less direct competition, stickier customers, logical sales processes, and predictable revenues.
Investors and media have painted a picture of these as opposing camps, with a fight over who gets to deploy more capital. The same analysis stands for what types of funds returns more capital as well, with entire funds taking the distinction as gospel, and creating portfolios around it.
And yet, in the new cloud dominated world, I wonder if that still holds true in any meaningful sense. Sure, it's easier to make a new casual game app than to make an HCM software, but those are not the relevant units of comparison. And while it's indisputably true that there are categories where the comparisons could be true from a complexity measurement point of view.
There’s a generic framework that investors place themselves in, and there has been a large amount written about what makes for a successful investor in one vs the other. But more recently I’ve been wondering whether we shouldn’t click one level deeper. Companies that are ‘enterprise’ like Slack or monday.com rely on traditional B2C methods to increase penetration with their customer base, blurring the boundaries. While there are obvious extremes, like SAP vs Facebook, those are in the minority.
As someone who thinks about enterprise businesses for the most part, I wanted to think one level deeper about whether this differentiation makes sense anymore. Especially in a world where the boundaries between work and life have blurred, where WhatsApp is a key way to do business, and where Zoom parties are all the rage, it only makes sense!
Similarly there's a few ways to look at enterprise companies too
- The classic enterprise companies - HR, procurement, data analysis: Boring, well understood problems that nonetheless are painful to buy, painful to implement and painful to use. Find whatever is the latest trend, beautiful UI, user controls, AI, and throw it at the established behemoth, and out comes a beautiful start-up. The winners win because they crack the go to market before others, or outdo them through persistence, money or cheating. Find the best team and back them would've been the right advice, but that only works if the "best teams" have scarcity value, and I'm not sure they do any longer.
- The infrastructure layer: Massive changes afoot as we flip flop back and forth between centralisation and decentralisation based on network capability and compute capability. There's a sea change in this every few years purely because of the macro trends, and it's worthwhile therefore to stay on top of it. It's heaven for those who like to geek out, and despite all professed interest, that remains still a smaller community that many would anticipate. The size is enormous, because ultimately everything needs infrastructure. Sad, but true.
- The nameless layer: I can't figure out a pithy name for it now. But obviously how you buy it changes as how you create software changes. Things that used to require 20 engineers a month to implement is now an API call, and guess what, that means you don't need to sweettalk an IT person into doing things for you either. Built on top of this is the fact that in a world with millions of engineers all trying to stand out, building a resume is easier through GitHub than through sanitised performance reviews from an old job or reference calls. Put them together, and you have open source heaven.
And then, do you do B2C investments? I feel when the question is asked it tries to distil a vast array of human experiences into a couple of buckets without really trying to figure out what you're actually saying. So this is an attempt to try and nail down the taxonomy, as it's how I try and look at what a "consumer" company is, and what it does.
- Classic consumer - apps, gaming, social media: This is the toughest area to predict winners a priori. Excellent user experience is essential, but not sufficient, and success comes with scale. Scale comes from a complex interaction of individual needs, that collectively create some "magic" resulting in an outsized win. It's worth saying that a large proportion of giant wins in the Consumer space comes from this, so it's tough to ignore, but it's also incredibly tough to find the right horses to back here. Just try creating a gaming app and launching, it's so much easier to do, but so much harder to do well. Consistent content creation is incredibly important and impossibly difficult unless you have an established fanbase - everyone's willing to give Disney a pass.
- Consumer purchases - Ecommerce and marketplaces: The only question VCs ask is how you can survive Amazon. But regardless, it's an area with enough variety of prior success stories, whether in individual niches like used car sales or DIY tools. The mechanics are understood is not however a reason to invest, in fact it's a good reason not to. It's one that's easy enough to build, technically for sure, and operationally requires the necessary legwork that can also be put in. But winning requires massive amounts of scale, and reaching that scale means paying at the altar of Google and Facebook, which VCs are all too happy to subsidize for you.
- Quasi -enterprise consumer - neo-banks, neo-insurance, and so on: These are consumer facing products that need to have everything that sets consumer software apart - exceptional ease of use, seamless integrations, immediate notifications, amazing UX - but that solve complex enough problems that building one is not trivial. The jury's out still on how large the successes are likely to be, but it's a whole new world of opportunity right there, which is why nobody says no to this bucket of goodies.
As with any attempt at simplification, there’s a bunch of companies that span categories or jump between them, or just plain fall through the cracks. To look at one, messaging sort of fits in category 1, though extends into 2 to make it stickier and more usable. WhatsApp is an example. Some, WeChat, even jump into option 3, because they can.
All this to say something obvious. B2C is not an investment thesis, it’s barely a category. Same for B2B. Every company we interact with on a daily basis, whether it’s Gmail, my supermarket, Amazon, or my public metro, theoretically are B2C companies, but practically that definition is so broad as to be useless. This is my way of trying to reduce that uselessness a little without making it so complicated that I jump back up the simplicity ladder and making dumb pronouncements.
Implications of this product-centric view on the enterprise software world
In the new world, there has been an explosion of solutions that make the task of building a software stack easier. Every inch of what used to be a difficult problem to solve has now been converted to a company who makes it their mission to solve that particular pain point for everyone.
And as it has gotten easier to build a start-up, it's also become harder to stand out in the crowd!
The side effect of making building software companies easier is that now plenty of people want to build them. And the resources that would have overwhelmingly gone towards solving R&D problems get repurposed towards finding and reaching customers. On average a saas company spends north of a third of their ARR on sales and marketing, with often half of that number goes to R&D.
Individually, so far so good. The customers have an extraordinary amount of choice in the type of software they would like to use. The founders and employees have a wonderful ability to build innovative products and get them to the market fast. And the large incumbent companies can afford to let them all fight it out until a winner is crowned, and then just buy them.
The difficulty with the increasing amount of competition though is that unlike some parts of history, its not becoming much harder to create a truly large outcome. Large numbers of companies are all battling for the same customers, and unless there is a clear network effect, there is no reason why there would be a way for one company to get to a meaningful size in the market. And that shrinking of potential exit sizes also comes to play against a backdrop of a larger number of competing players.
And that's the new landscape within which enterprise companies have to operate now. Customer experience, ease of use, plug and play operations and an endless squeeze from the entire market to outcompete your product on price, usability, features and reputation. Every individual innovation, and the success story that results from it, become a clear signal to the larger software giants that they should invest, build or buy a solution in that same segment. Companies that have fantastic products like DataDog or Splunk, seen in this light, look like they are one strong product team in Microsoft away from becoming middling outcomes. They will still be great venture capital investments, but they won't be the equivalent of backing Microsoft in the '80s or Google in the '00s.
And when you think about the recent successes in this new world — they have found ways to back their go to market methods to solve exactly this problem. This isn’t the Microsoft or Oracle way of selling. It’s bottom up like Slack, open source led like Databricks, user experience led like Zoom or developer experience led like Mulesoft — all examples of how you can build a large business by reaching users first and companies second. This can be through Product Led Growth, Open Source led adoption, or UX driven through relentless focus on the reduction of even the tiniest bit of friction that you might not even have noticed before (less than 100ms for anything email related, ahem). Anything to bypass the Byzantine world of corporate procurement and get an extended sales force, or at least a fanbase, inside your prospective clients. I'm not sure this works as a roadmap if you're aspiring to build something new, but rather is probably an elaboration of a trend that seems to be ongoing, but in either case it's a paradigm shift that's worth paying attention to, if only to watch history made as it's made.