An Algorithm for
B2B Products

3 Aug 2020

London, UK

by Matthew Eric Bassett

There's an algorithm for product discovery, at least for business-to-business products. If you combine that algorithm with decent technical execution and a SaaS business model then you have a recipe for a company. And if your market is big enough then you may even be looking at a billion-dollar startup 1.

The algorithm is this:

  1. Work in, or otherwise accumulate, deep knowledge about the day-to-day work in a particular industry.
  2. In that day-to-day work, identify a time-and-labor intensive process that is lacking automation or software assistance. It's even better if that process is error-prone.
  3. Write software to improve that process, and measure its success by the amount of labor saved, amount of time saved, or reduction in error rate (or all three.)

Each step is necessary.

The first step explicitly states that business-to-business problems are opaque to outsiders. As alluded to in a previous article 2 each industry has complex, time-and-labor-intensive processes that could be automated or improved. But those processes are not being addressed because the people with the skills to improve them are unaware that the problem exists. This opacity is an opportunity. Only those few people who can combine knowledge of the business problem with an ability to execute are in a position to turn that problem into a software solution and then into a revenue-generating company. In today's economy, discovering one of these problems is like finding a gold mine.

The second step is all about what problems are worth solving. Labor-intensive and time-intensive processes are a sign of low productivity and perhaps under-investment in technology. And an error-prone process means that you have more room to increase productivity with your solution. It is an easy sell if your humans make fewer errors and take less time once they are using your solution.

The third step both emphasizes the importance of technical execution while dismissing focus on any particular implementation. Let me explain: One needs the ability to deliver a solution. But many people need to be reminded of what is essential: customers care that a solution exists. They do not care how you have implemented that solution. Too often, business leaders (and some technologists) try to deploy some technology after hearing that someone else is realizing success from it. Once a leader hears that competitor Acme, Inc is using Hexagonal technology, she starts forming teams to use this technology in her company. Those teams are then required to find some hexagonal-shaped problem somewhere in the business. Yet this is backward. The algorithm starts with knowledge and experience, proceeds to a problem, and then to a solution. Your customers will not care whether you are using a generative adversarial network or a clever Excel macro. But they will be delighted if you can build a piece of software that turns a dull, 2 weeks long, error-prone process into an engaging, 2-minute long task.