Service and Rivalry Traps

Or: how to accidentally ruin the fundamental benefits of software.

Rival and non-rival goods.

Imagine we are out on a hike. The trail washes out on us and we and both fall down. We’re both injured. But we only have one walking stick between the two of us. That walking stick is rivalrous. If you’re using the stick, I can’t. If I’m using it, you can’t.

Rival goods are the default. I can’t copy+paste your walking stick. I have to go make my own.

But ideas are non-rival. If I have a great idea I can tell it to you and now we both have it.

Many ideas go beyond non-rivalry and get to anti-rivalry. I get more out of some ideas if you also have them. It’s easier to talk to someone about that book you read if they read it too. It’s easier to talk politics with someone who has the same basic ideas about politics as you.

Software

One of the great thing about software is that pure software is non-rivalrous. If you have a photo editing app on your phone, I can have that same photo editing app without taking yours.

Many of the biggest software businesses though rely on anti-rival software. Usually these are multi-user software like social media. The more other people use it, the better it gets for you. The more people are on facebook, the more likely you will find your friends on facebook. The more websites Google indexes, the better search becomes. The more users search with Google and click on results, the better Google knows which results are the best ones.

But not all software businesses are anti-rivalrous. Many of the most powerful pieces of software fall into a trap that makes them rival goods.

How?

If you have a product so complicated it needs a lot of help, the software is non-rivalrous but using the product is not. To start using the product you need lots of direct help. Direct help takes someone’s time. Someone’s time is always limited [1].

Time is limited. Time is rival.

Where can you find these kinds of companies? Look for large enterprise software companies that have large customer support or implementation teams. Electronic medical records is a great example. It can take a year or two for a hospital to install and start using medical records software.

When a business with software is not a software business.

If your products require lots of human interaction to use, you have a consultancy. Consultancies are rivalrous. If each new user takes some time from the core product team, you will not scale the way you expect. Imagine if each new user of facebook had needed an engineer to take ten minutes to explain how to use it. Facebook’s 2.7 billion users would mean 51 thousand years of customer support, just to get that user to sign up.

The pain comes from the gap between thinking you are an anti-rivalrous software company but you are actually a software-enabled consultancy. Consultancies plan and price very, very differently.

• Consultancy: small # of clients, big $ per client

• Software: big # of clients, small $ per client.

Middle Grounds and Shifting The Work to Your Customers

Some software products will always be complicated because they are solving very complex user-facing problems. Google Search is an example of a very complex problem but a simple user-facing-surface-area: the search text box.

This usually applies to software businesses that’s inherently complicated, like financial risk analysis, or medical records, or any kind of flexible process automation (dev ops, Machine Learning pipelines, etc.).

There are two main strategies to address this start-using-the-product-gap: Customer Service and Compliance.

Customer Service

Customer Service is a middle ground way to shift these rivalrous time costs, but it’s just a shift. It’s certainly less costly to not have your expensive Software Engineers doing customer support. Having some software engineers do some customer support can be a way to let the product builders (the software engineers) learn ways to make the product not need customer support though!

Compliance

Discover or create a carrot/stick that makes users need to pay the upfront cost of learning how to use you, and they will. Strict governmental regulations are one form of this. But there can be softer forms of “I can only get this good thing by figuring out how to use your complicated piece of software”. Maybe you want to be a professional photographer or movie editor. Each user will have to put in the time to learn your piece of software. The rivalrous time cost still exists, you’ve just shifted it to your users. This is an economic win for you, but make sure your software is worth it in the long run. Otherwise you’re just arbitraging your customer’s time, selfishly.

Know Which Business You have

In summary: pure software businesses are magical only when you don’t have to spend your rivalrous time to get people to use it. The two extremes are “real” software businesses and software-enabled consultancies. The planning and pricing for these two extremes are very, very different. There are mitigation strategies to find middle grounds with customer service or compliance [2].

[1] If you’ve figured out a way to have more hours in the day, I have investors to introduce you to.

[2] “compliance” isn’t the best word here, open to suggestions.