I didn't post last week for personal reasons, but I had been meaning to write about this since I started this blog, so I'm happy to finally share my thoughts.

Today's post is inspired by the experience I had at a startup where I worked for a few years. Its story is long and filled with ups and downs. During my time there, I saw it transition from a technology-backed service provider startup to an enterprise SaaS company. I worked there for almost three years under the title of Product Engineer, witnessing the company's pivot from its final attempts to become profitable as a service provider, to selling its software to competitors.

The company's business idea was very similar to Uber or Cabify, but applied to a different industry. It tried many things for about 6 years until the founders started showing the technology they'd built to competitors in the industry and received initial positive feedback.

Since then, it has gone down a very slow rabbit hole. Why? Because it began targeting the biggest names in the industry, meaning enterprise sales.

So what's wrong with Enterprise Sales?

Slowly (emphasis on slow), we learned that it takes a long time for these sales to happen. First, you must find your internal champion—someone with some degree of agency, decision-making power, or influence within the company. Then you must sit through many rounds of demos with multiple big names at the company you're selling to. You have to make sure the head of operations agrees, that the IT department approves, that it doesn't exceed the finance department's budget, that the decision is approved by a Leprechaun at St. Patrick's parade (only available every 17th of March), etc.

Such a long and tedious process means a few things:

• The buyer is most likely not the actual user, so you're not really selling solutions to their pains. Instead, you're selling ideas that help your buyer climb the corporate ladder by claiming their latest acquisition brought X amount of revenue—while the real users might be forced to use something they may even hate.
• The time to close a lead is measured in months (I think the company's average was 8 months). You're going to need money to pay expenses while you close your next customer next year.
• Whatever you're selling will most likely require modifications because your product might not fully meet requirement Y. Your product will very likely become a Frankenstein.
• If you have these modifications, they'll become part of the contract irrefutably, so you know this will become a task in someone's SCRUM/Kanban/Kangaroo board.

But in all fairness, it's not all bad. People do make a lot of money selling to enterprise, and once you're in, you're quite likely not going anywhere. If you have a well-defined and established product, it might not even be much of a hassle—some of the points above may simply be "occupational hazards."

What does that mean for a startup like this one?

After about half a year of trying to sell the software it already had—with minor tweaks—and receiving further feedback, management realised they needed to sell features we did not truly have in order to close some customers. This gave them a more compelling narrative to raise funds for this new company (now branded with AI to ride the wave).

This meant prototyping solutions to showcase around, then deeply understanding the customer's business need to ensure the prototype addressed it, then prototyping even more—and before you know it, you have a complete design already approved by all buyers and sellers. It's ready to be developed or "parametrised" in short bursts of engineering witchcraft, also known as a waterfall of sprints. Time is of the essence, and hypotheses are very high-level, so you just develop because it's become part of the contract, and you must obey the contract!

As a Product Engineer, your role is now focused more on challenging already-made decisions, hoping you'll be able to convince stakeholders to narrow the scope and cut some corners. This means finding compromises that the buyer understands—arguing that we most likely don't yet know how users will actually use feature X, so it doesn't make sense to dedicate much time to it.

But okay, you have something that's part of your product offering, and you must be able to repeat its implementation easily next time, right? Yeah, right. If only it were that easy. It turns out the next customer needs something completely different, so get ready for round two!

Repeat this enough times, and you might end up actually reselling some of the custom things you've built for other customers, but let's not fool ourselves—we've been doing good old consulting in the meantime. Sure, we might have been architecting our software to be reusable and customisable because we're SaaSsy, but we've been operating as a regular software shop since then.

Does this apply to everyone?

There are many product companies selling to enterprise that have very solid products, so clearly they don't face the issues mentioned above.

Do all startups suffer the same fate? Probably not. There must be startups out there selling to enterprise with a very clear product offering that almost undoubtedly covers the business need. They're probably more likely to focus on end users and iterate their product organically.

But if the team is small, the product is not well-defined, the market is relatively small and old-fashioned, and/or your company does not have enough money to maintain focus on something specific, then I'd be afraid it could start behaving more like a software shop and less like a startup.

What does that mean for me?

I'm not entirely happy with this way of building software. I worked in consultancy before, and I know it can be a lot more stressful than this, but I've since read much more about startups and building products. I feel this sort of environment is only slowing me down. Companies that match this description are therefore a no-go for me.