Some Thoughts on Vertical Ecosystems
or why legacy vertical solutions are in trouble
It’s really fun to see more and more people start to think about verticals. And it’s especially fun to see more and more of the vertical crew coalescing around this newsletter. So thanks for reading. Lots more to come.
Every vertical is going to get killer software. And as different verticals get better software, in my view vertical ecosystems start to form. That’s what this week’s edition is all about.
As far as I can tell, there isn’t a lot of writing around how vertical software companies will interact with each other. Partly, I think this is because the entire category of vertical software is still relatively nascent and much of the current technology is finally able to capture flows previously unexplored. Partly, it’s because everyone is focused on building a great business and it’s hard to zoom out.
But sometimes, it’s helpful to look at what is being built in the aggregate. And in my opinion, that’s unique ecosystems for each vertical. So this week, we take a look.
Every biological ecosystem is characterized by how creatures interact with each other and their environment. Vertical ecosystems are no different.
Simply put, vertical ecosystems are a way to describe the interactions between vertical marketplaces, fintechs, and SaaS products that map the entire value chain of an industry. Like any ecosystem, vertical ones are highly emergent and full of collaboration and competition. Over time, the entire industry experiences end-to-end digitization with high levels of interoperability.
End to End
Vertical ecosystems begin to emerge and characterize every participant in the vertical. This historically wasn’t possible. For instance, there’s a common trope that it was pretty hard to take payment on the consumer internet prior to Stripe. But it was nigh impossible to do B2B payments. Now lots of new infrastructure has been developed and technology is able to service every single business activity in theory with digital workflows. Roughly these workflows fall into 3 different categories and correspond to three different types of vertical businesses.
Vertical Marketplaces enable the exchange of goods and services
Vertical SaaS companies enable the exchange of data
Vertical Fintechs enable the exchange of value
All business activity maps to one of these three exchanges and enterprising founders gravitate to find workflows in parts of the industry not served by technology. Over time, vertical companies begin to add workflows. Marketplaces become fintechs and fintechs become SaaS platforms.
Ultimately, this leads to various companies competing with each other to best serve workflows in the vertical. And like any ecosystem, this fight for territory, itself shapes the vertical. But it’s not all competition. Like Procore, companies decide that they can best shape the industry by enabling additional companies to thrive and becoming enmeshed in every single workflow. Thus, platforms start to emerge. This fosters both competition and collaboration across the vertical and an ecosystem starts to form.
Vertical ecosystems are highly emergent. By emergent, I simply mean that there is no central planning team dictating how the vertical ecosystem should work. Within each vertical, there will be an evolution. However, it does appear that there are some patterns.
For most verticals, a vertical SaaS solution will bring online critical workflows for the industry. Verticals have a natural flow. For construction, everything flows towards the jobsite. For the food industry, things flow towards restaurants. These tend to be the platform for these key industry participants to integrate other flows.
I’ll touch on this later, but this is one of the big distinguishers from how legacy vertical companies and newer ones operate. With legacy solutions, ecosystems didn’t really emerge. In part this is because there wasn’t a willingness to integrate and a desire to become platforms. It’s far different now as companies are now built with API-first approaches and desire to be the source of truth for their stakeholders across all their workflows.
To become the source of truth, you have to be able to interact with every single aspect of your customer’s data. Integrations into other sources of data become table stakes. And as data exchange becomes the norm, digitization pressure is put on the whole ecosystem. As a result, other startups start to emerge also with an eye on integrating data relevant to their stakeholders and with serving as many workflows as possible.
Typically, different parts of the vertical will have different needs. And so companies emerge with different initial workflows. But startups aren’t static and eventually companies start to expand into additional workflows that they can capture. In the process, vSaaS become vFintechs, and vMarketplaces become vSaaS.
Marketplaces are especially interesting in this respect. It’s not easy to aggregate both supply and demand on a platform and it usually involves quickly jumping into additional workflows in order build a moat. Every marketplace has to figure out how to make the buying journey seamless both and simultaneously convince suppliers to use your platform. The problem that these marketplaces often face is that there is an existing analog operation that has a whole host of implicit operations around it. And often these operations are integral to the operation and without which, the whole thing falls apart.
And in most of these markets, the suppliers are going to be the primary obstacle. After all, they already have a network that they transact with. You have to convince them that they should bring their network online or give them enough reason to think they can expand their business. There are various approaches to solving this but it usually means not strictly being a marketplace and more closely building software or fintech for the suppliers. Choco, an ordering system for the food industry, now builds restaurant supplier software.1 Felux, a steel marketplace, offers seller financing and shipping.2 In fact, these adjacent workflows may ultimately be where marketplaces accrue their value.
So different workflows get aggregated, different parts of the industry start to digitize, and an ecosystem begins to emerge.
Over time, I think vertical ecosystems will become highly interoperable. Some of this has been laid out prior. The bet is that API connectivity will come to characterize every aspect of software and this will be of course be true of vertical software as well.
However, this is probably the most speculative aspect of vertical ecosystems, partly because we are in the early stages of two developments. First, frankly many verticals that could become ecosystems are still in the early innings of bringing workflows online. Many verticals are in need of any modern SaaS technology before the whole industry can even start to be talked about as an ecosystem.
If there are established players, often they don’t have the sort of API-first approach that would lead to easy integrations. I think you will need several mature vertical participants before companies begin to consider what other aspects of the vertical they should more readily integrate into.
Second, we are still in the very early innings of data clouds. Data clouds pose an interesting solution for large industries with massive geographical spreads and large sums of data. Historically it has been really hard to do stuff with the large sums of data that are being generated, much less to develop cohesion and unify data silos. Snowflake and others have changed that and now have a vision to enable seamless data exchange. Snowflake in particular has expressed a desire to build industry specific clouds and I fully expect this to be a valuable addition for vertical companies over time. As Snowflake continues to further verticalize its strategy, vertical SaaS companies will pioneer industry data clouds either with Snowflake or in a similar vein in order to enable further data exchange amongst the ecosystem.3 There's a lot left to be discovered on exactly how this shakes out, but over a long enough time horizon this will become the norm.4
At this point, a valid objection could be raised. So what? After all, no entrepreneur is building a vertical ecosystem, they are building a company. What’s the reason to think about vertical ecosystems and how they will develop?
Beginning with the End in Mind
The answer lies in a Stephen Covey principle: Begin with the end in mind. It is only by beginning with the end in mind, that vSaaS companies will fend off the biggest vice of legacy vertical tech.
Point Solution vs. Partner
The best vertical SaaS companies not only generate high amounts of second-order insights but they tend to grow into platforms upon which the rest of the vertical ecosystem operates. Thus for vertical founders it becomes crucial to develop a vision for what the industry will look like fully digitized and then build in such a way that you can both create value for the entire ecosystem as an industry partner and benefit from future vertical companies wanting to build alongside you.
Conversely, many of the legacy vertical technologies are highly insular. They function more as point solutions for various workflow problems rather than attempting to be an industry partner to serve the vertical.
More importantly having a vision for how your vertical ecosystem will form also gives founders an important insight into why they can compete with existing solutions. After all, there are plenty of verticals with legacy vertical software. They beat you to the punch, should you go compete?
You definitely should. Lots of these have huge moats but little innovation. In most cases, if an ecosystem around their vertical develops, they are not set up to capitalize.
Why?They lack the platform vision central to so many modern vertical SaaS companies. Thus, a vision from the outset of what type of ecosystem could develop amongst the industry gives founders a reason to go tackle these companies despite the moats. To hearken back to my favorite Peter Gassner quip: the thing you should be focused on is if the current software provider is really providing value. In many cases, these companies are providing workflows but have ceased to think about how to help the industry continue to innovate digitally.
As a result, these legacy companies are structurally not set up to take advantage of fintech innovation, marketplaces, data sharing, and much more. If vertical ecosystems are inevitable, this gives you a reason to believe you can go win a war with them. At stake is your chosen vertical’s ability to adopt modern technology and benefit from interoperability. These stakes then allow founders to mount a valid claim around becoming the true industry partner. After all, you have a vision for the industry and are willing to fight for it. That’s valuable.
Fortunately by developing a vision, founders are also preserved from the mistakes of their predecessors. Build a platform and benefit from the explosion of technology in the vertical. Build a system designed to integrate data, and you will be set up to integrate non-obvious future sources of data that can’t be fully predicted. This staves off disruption risk from building an insular platform. These are the kinds of things that also allow you to future proof your customers too. And after all, that’s what industries are looking for.
There are some limitation of this approach.
First, it’s fairly clear to me that certain verticals may not have ecosystems. One should never overfit the model. However, this isn’t a good reason not to look hard for one. Some ecosystems may simply be non-obvious. It might require zooming out to see all of the value chain. Hypothetically, a valet management SaaS could operate as a standalone vertical solution. After all, it’s usually a distinct business from the locations at which the company operates. At the same time, an ecosystem approach might eventually yield a more direct integrate into various hospitality management systems and restaurant software in order to more readily gauge surges and payments data.
Lastly, there may be some verticals that can’t develop these types of ecosystems because it doesn’t appear feasible for players to emerge across the vertical. Perhaps the vertical is too small and it’s not clear if a software business can sustain itself. This is definitely true and a entirely valid concern. Some verticals it just won’t make sense. But in most, my hunch is that the lack of feasibility is mostly an infrastructure problem.
There’s a lot of building blocks should be built that are currently not serviced by infrastructure companies. I won’t name all of these because I’m saving them for a later piece but suffice to say if you start to dig into what common features are offered across different vertical companies, you would see some great businesses to be built around developing primitives for vertical use. As these, primitives get built, software products can be developed far more cost-effectively without needing to build the basics in workflows that are time consuming and undifferentiated. This may not mean these are venture-backable but there are various strategies to build these solutions for different verticals and generate a return.
So there you have it. A theory of vertical ecosystems and why the concept is useful. I fully expect my thoughts to evolve on this. The next edition will focus on a pretty interesting player within the vertical scene. Not to give too much away, but there are some massive businesses that have been built around acquiring vertical software businesses.
As always, thanks for giving this a read and for continuing to be a part of this newsletter. It’s been a pleasure watching it grow.
At some point, I think Felux or an enterprising company will begin to make software for steel suppliers/producers.
One notable example is the recent creation of the SEI Data Cloud. SEI is essentially a vertical platform for the value chain around financial services. The team up of horizontal solutions with vertical players is something to watch. Just like Veeva teamed up with Salesforce, it’s not hard to imagine Snowflake teaming up with various vertical companies.
Rise of the Data Cloud was highly influential in developing this thesis.