The Tale of Two Startups

I am going to make a bold statement 👇

It tends to not be the best product that wins in the market.

The product that tends to win is the one that the customer can integrate into his or her workflow the easiest.

Workflows, at their core, are a path to achieve an outcome. Your customer is repeating workflows over and over. Routine.

Yet, many startup products tends to tamper with existing customer workflows and behavior.

A strong startup product ought to have an “a ha” moment. But he more hurdles on that new behavioral path, the less likely the user is to go through the loop and get to the "a ha' moment. In product workflows, it's all about shrinking the time to get to the "a ha" moment.

In all of my startup investing, my working hypothesis is that the startup whose product changes the least about the existing customer workflow and behavior, wins.

And if 10x increase in value occurs with a very similar if not identical workflow, then a mega market winner is afoot.

This is far from the “disruption” that Silicon Valley likes to talk about. It’s really about complete accommodation. 😂 

The Tale of Two Startups

Let me share a story with you that left an indelible mark on me on why designing around workflows and behavioral friction are so important. The details are anonymized but the stories are similar. 

I once met two different startups years apart from each other that were solving the exact same problem, selling to the exact same user, and yet diametrically opposed in their market traction. Let’s say that both startups existed in the legal space selling to paralegals on staff at legal firms.

Each of the startups’ value propositions were very similar. It involved vastly shortening the time to complete a very repeatable task - getting the customer to the intended outcome much faster.

Both startups had also identified that the paralegals did not want to be replaced.

✅ But what was also a material difference was this - one CEO (Ryan) thought about the product design through the lens of behaviors and workflows. The other CEO (Craig), didn’t.

Both were even using the same essential technology. But, Ryan had mapped out a more robust and clear work process accounting for behavioral tendencies, where Craig had not.

This difference led them to different contexts for their solution. ✅

So what was the difference?

Ryan observed that the paralegals did not want to be replaced from their jobs using AI.

These paralegals worked in a particular industry. In their workflow, there was a very common PDF document that was routine for their jobs that came up over and over again.

Ryan’s startup developed a V1 of the product that created a summary of this PDF document.

Problem solved? No, it turns out that this summary was not something that the paralegals were behaviorally used to.

In his observation, Ryan identified that paralegals said that they read the summary, but behaviorally they read the full PDF document anyway, just to be sure that the summary was accurately representing the document.

Specifically, Ryan observed that it was routine behavior for paralegals to read through this PDF document.

In essence, the summary that his startup developed wasn’t cutting it from a value perspective because they still took the same on the task by reading the PDF.

Ryan realized that this “quality assurance” action was a critical behavior that the paralegals saw as the value that they believed to be contributing. Thus, changing this behavior too much would make the paralegals nervous to adopt his startups’ new technology.

So, Ryan created a V2 solution - the paralegals start with a PDF, but through Ryan’s startup’s technology, the paralegals now read a smart PDF which shortens the time of their existing process. Essentially, they execute same process but it’s a much shortened process with shortcuts. 

✅ Importantly, by good product design, the paralegals are still reading the PDF in the same way they had in the past with existing behaviors. But instead of a summary which created friction due to disrupting their workflow, the solution became a smart PDF which was much less onerous than it was before. ✅

It’s augmenting the paralegal’s skills and empowering them to be able to do better. 

Unsurprisingly, Ryan’s startup is doing very well.

On the other hand, Craig had also observed that the paralegals did not want to be replaced from their jobs using AI.

But Craig was actually was led to a different conclusion - that paralegals DO want a summary. This advice was seeded to him by an external party, not a user in the market.

Additionally, Craig never acknowledged customer workflows and the friction they possessed. 

So, Craig’s startup build a product that created summaries for this long document. But, when Craig and I talked, he couldn’t understand why customers were not adopting his product.

Craig didn’t get to the granularity of workflow behaviors.

In so doing, Craig never got to the key insight that Ryan did - that paralegals want to read the entire PDF document. 

I want to be very clear with these two anecdotes:

The underlying technology? Exactly the same.

The use case? Exactly the same.

The market user(s)? Exactly the same.

The major difference? 👇

✅ One CEO honored the embedded friction in workflows and behavior and what changing them could mean. He designed a solution in step with those workflows, choosing to work *within those constraints* and not develop something outside of them…

…whereas the other CEO had designed their solution completely unaware of workflows and customer behavior. ✅

Not seeing this caused both CEOs to develop different solution workflows.

✅ Note also that these two solutions are not, behaviorally speaking, 180 degree differences off. They are say, behaviorally 15 degrees off from each other. But until a startup’s solution points to the true north of the market, 15 degrees off is the same thing as 180 degrees off. ✅