I have been thinking through this post for some time but haven’t quite gotten to where I want to be on a point of view so figured taking a version one shot at writing it out might help me refine my thinking as well as generate some feedback (both positive and negative).

For as long as I have been working in and around enterprise software there was always a prevailing tailwind for startups around the enterprise customer’s build vs. buy conundrum. The company has felt a problem and knows it exists but can’t get a solution from any existing vendor and is working to sort prioritizing and marshaling the resources internally to address it.

Along comes a startup whose singular purpose and passion is to solve that thing and voila even though the startup product is early and the team untested, they could get a shot at it because the alternative was try to force prioritize it from an existing vendor on an unknown timeline or to navigate internal politics and procedures with the hope of getting something that sorta fixes the problem 12-18 months from now and with the additional burden of having to maintain and administer this new thing that is now part of the proprietary internal technology stack.

With the latest and fast moving developments around AI enabled and assisted software development, that tried and true equation is coming under pressure.  For a realized and immediate need, a company can now build something quickly to address it without needing to take on the additional risk of a new vendor with a janky product. They can now influence the iterations internally in the same way they could previously influence the startup product roadmap with the same or better turnaround time for results.  

Yes, there is still the question of insourcing or outsourcing key parts of the enterprise technology stack and required management of it but the equation has changed.  Yes also the culture of that company has to be one of innovation and speed but those were the early buyers of startup technology along the way.

The window an untested startup could get to solve a problem that no current vendor or trusted provider could is shrinking as the desperation required for that situation can be now addressed internally more easily than ever before should the company choose to do so.

This same dynamic puts pressure on how startups have traditionally exited and created liquidity and returns for their investors, founders, and employees.  Large platform vendors would do a similar build vs. buy calculation when deciding to acquire early stage companies and the elevated valuations they were willing to pay were indicative of the get it now, get something differentiated, and immediately be present in a new market or product category.

The sweet spot of being in the ~12 month timeframe for that vendor’s product roadmap where they want to have it but have to evaluate the cost of waiting to get it along with the cost of building it increased exit multiples for startups driven by the “right now” value of being able to put a somewhat proven product into the distribution system of that platform vendor.

This doesn’t mean that you can vibe code your way to enterprise grade software but it does redefine the landscape of opportunities for startups and startup investors.

As a former management consultant, quadrant visuals always come to mind so here is an initial and am sure to be refined way to look at this.

Y-axis: how simple or complex the need/opportunity is to solve

X-axis: how scalable and repeatable solving that need/opportunity can be

The underlying thinking here is that things that are easily solved will be with internal tooling or the defensible position of any vendor trying to address more easily solved needs even at scale will be perilous as switching costs are low and new entrants prevalent and constantly arriving.  These incrementally better products could get traction in a build once, deploy everywhere SaaS world where even just better UI and reporting could win the day.  Thus the replacement of client/server based software for web-based software but still solving in the same categories – CRM, procurement, supply chain, accounting, payments, etc.

So that leaves the more gnarly and complex problems that are either so unique that there is no repeatability in the solution or have the potential to be solved now in those established categories in a way never before possible because of the application of new technology in the form of AI and its acronym laden cousins. 

So for each quadrant, a few themes:

High Complexity, Low Scalability – this remains the domain of service providers, consultants or internal teams building bespoke solutions but is also where new opportunities emerge as once highly complex and custom solutions can be delivered at scale as they move to High Complexity/High Scalability needs.

Low Complexity, Low Scalability – quickly built and deployed to address a unique point need with limited expense and easily replaced and enhanced. Unlikely to get movement from this quadrant as the appeal and value of executing on Low Complexity/High Scalability opportunities is lacking.

Low Complexity, High Scalability – I would argue the traditional zone of workflow-centric SaaS that could solve for relatively generic and similar business processes but do so with standard functionality and limited customization. 

High Complexity, High Scalability – where I think the actual enduring opportunity exists for now where problems previously deemed too hard or too expensive to solve can be solved but in a capital efficient, timely way that creates a somewhat repeatable approach from customer to customer. Taking business processes from workflows and task execution to full on experiences and outcomes exponentially better than was available with software, spreadsheets, and manual steps.  There is also a dynamic where as the technology reduces the complexity, the opportunity begins to take on characteristics of Low Complexity/High Scalability markets.

So where are the opportunities for startups and their investors? Good question and an initial somewhat poorly formed answer would say where the benefits of technology have not been fully realized and adoption has been stymied – stodgy & legacy industries, highly regulated industries, skilled trades. Places where conventional wisdom and the SaaS landscape simply avoided or relegated to legacy “good enough” installed solutions too hard to displace but no longer innovating or growing much (often due to private equity ownership but that is a topic for another day). 

If we are entering a world where being able to build the software is no longer an obstacle to entry or sustained competitive advantage, then what is?  A unique and proprietary distribution edge? A highly specialized and differentiated amount of expertise and domain knowledge among the founding team? A brand and reputation of being able to solve these highly complex problems in a scalable and dependable way? 

To be determined but speed will continue to be the only real competitive advantage a startup has over incumbents and competitors but where that speed is directed will be critical for founders and increasingly important to understand and use to evaluate potential investments as an investor.