The Theory of Constraints <> Your Software Business

Now with GIFs!

This is a newsletter about Product-Market Fit, a commonly misunderstood topic that I’m trying to figure out as I build a venture-backed software company.

Last week I wrote about three things I want to hear from customers that tell me I’m heading towards Product-Market Fit.

This week I’m thinking about:

• How the “Theory of Constraints (ToC)” applies to building a business

• Why ToC delivers unpredictable results - and why that’s a good thing

• The implications of unpredictable results on plans and planning

To summarize: It should be obvious which problem in your business deserves your focus, but the implications of solving that problem should surprise you.

Theory of Constraints: Attack the Bottleneck

The third most revered book in the Jewish tradition (aside from the Torah and Tuesdays With Morrie) is The Goal, a business novel by Dr. Eliyahu M. Goldratt.

The book takes place in and around a struggling factory. The protagonist is charged with turning the factory around, and throughout the book uncovers the Theory of Constraints.

Here’s the gist:

Imagine a factory that manufactures Buffalo Bills AFC East Champion paraphernalia.

Different steps in the factory handle different parts of the manufacturing process - sewing, assembly, cutting, etc. etc. etc. If you look at the different steps of the manufacturing process, you’ll notice that inventory piles up before a certain step. This step is the slowest step in the chain, and is called “the bottleneck” for that reason. The entire manufacturing process moves as slow as the slowest step in the chain.

The Theory of Constraints says to focus your improvement efforts at the bottleneck. If you speed up any other part of the process, it won’t matter because, again, the entire manufacturing process moves as slow as the slowest step.

“Cool, Rob, I don’t run a factory. I run a software company.”

Here’s how it applies to software:

At any one point in time, your business has a bottleneck. One specific thing preventing you from “The Goal” - which, I assume, is growing a profitable business. The bottleneck deserves your focus.

You have a thousand priorities, a backlog of tasks, a hundred unread emails. Is there really only one bottleneck? And if so, how can you figure out what it is?

Gary Keller in his book “The ONE Thing” has an elegant solution to this, written in the form of a question:

“What is the ONE Thing you can do such that by doing it everything else will be easier or unnecessary?”

So. Freaking. Good. Simple but not easy to do, as it requires real thinking and tradeoffs.

Every week I ask myself this question. And every day I try to make progress attacking my business’s bottleneck. I assume that if I keep doing this every day, we’ll be successful in the end.

But here’s the surprising thing about attacking bottlenecks, applied to building your business:

Attacking the bottleneck -> unpredictable results

I might be able to guess where the next bottleneck would be in a factory setting. “Well, when Step 1 moves faster, Step 3 is almost as slow as Step 1, so we probably want to start thinking about how we’ll speed up Step 3.”

But even in a factory setting, I’m likely to be surprised at where the next bottleneck emerges. Who knew the input processing for Step 4 would get overwhelmed because of a change we made in Step 1?

When you apply ToC to a software company, the complexity explodes. Software businesses have so many additional variables compared to the (relative) simplicity of a manufacturing line.

When you attack today’s bottleneck, the implications are often surprising and unexpected. Tomorrow’s bottleneck emerges when you solve today’s.

Here’s my company’s experience:

We realized that one part of our franchisee recruiting product was The Bottleneck - If we wanted to grow and increase our customer satisfaction scores, we needed to expand companies’ talent networks instead of relying on their own applicant flows.

Cool! But then as we attacked this problem… strange things started happening. We started getting feedback from potential customers that we hadn’t expected or heard before.

We realized that by attacking one particular bottleneck, our pitch to new customers and our pricing model were no longer coherent to customers and needed to change. What?!

This has become the new bottleneck, one that we’re attacking. It will cause another bottleneck to emerge, which we will also attack. And on.

Can you plan for this?

“Planning is guessing.” - Jason Fried, Basecamp

Each time a new bottleneck emerges, parts of our backlogs and plans become outdated if not irrelevant.

Butt there’s nuance here. Don’t make the mistakes of today’s business ideologues - the Soviet Planners and Blank Slate Agilistas. They have the intellectual nuance of Patrick Star:

• The Soviet Planner assumes that her plans create reality. She spends her time creating detailed forecasts and goals and descriptions of how things will be. If it works in theory, it works in practice. All that matters is reasoning from first principles. And getting that project management / lean six sigma certification.

• The Blank Slate Agilista assumes that you know literally nothing, that everything is learned through experimentation. We only know exactly what we have learned through the narrow confines of our experiments; draw no conclusions, test everything.

You cannot plan for or reason to everything, just as you cannot test everything. Your world is unpredictable and messy; your plans will fail and your experiments will give you (at best) conflicting data that require leaps of faith.

When thinking about the Theory of Constraints and how it applies to business:

• Bottlenecks emerge

• Observing and solving bottlenecks demands thinking, assumptions, and some planning

As you learn, much of your previously written backlogs and plans become irrelevant. But as you attack bottlenecks, new plans and and backlogs become necessary.

How much should you plan? How should you handle backlogs? To what fidelity should you make forecasts? And to what extent should you just get out there and learn by doing?

The answer as I see it: It depends on you being real about how much you actually know and what you can reasonably assume. The implications of what you don’t know could be massive (e.g., “The bottleneck is how to pitch our solution to customers so it makes sense to them.”) or relatively small (e.g., “The bottleneck is increasing the conversion rate on our landing page for a product that’s profitable and delighting customers.”)

In summary: The Buffalo Bills are the AFC East Champions.