Funnels vs. factories

Bringing manufacturing thinking into software startups

If PMF is the ability to replicate a single “hell yes” case study, then our business is just a system designed to replicate that case study thousands of times.

We can visualize this “case study replication system” as a factory line that brings people from “prospective customer” → “actual customer” → “hell yes” customer. AKA: at the end of the factory line, we produce identical case studies. Our marketing, sales, customer success, and even product/eng activities form this assembly line. Our job, therefore, is to shape each activity to develop an effective and efficient factory line.

When we look at our go-to-market system as a “funnel”… WTF do you do with a funnel? Nobody knows. So we’re prone to pursuing whatever the latest fad is.

Instead, when we look at our GTM system as a factory line, we can leverage 100+ years of production/systems thinking - from Henry Ford to Taiichi Ohno to Eliyahu Goldratt to Don Reinertsen to John Seddon to Stafford Beer - to understand it and improve it.

A few insanely helpful concepts I regularly import from manufacturing:

• System bottlenecks: When we view our GTM system as a factory assembly line, we can find the system’s bottlenecks. If we make improvements anywhere OTHER than the bottleneck in our GTM system, it won’t have much impact on the overall system’s performance. So at any one point in time, we need to find & attack our bottleneck. Nothing else matters.

• Defect rates: Lean thinking pushes us to remove waste. Talking to bad-fit customers is waste. Selling to customers who churn is waste. When we view these things as defects that our system creates, we can then attack the defects and fix our system.

• Variation: The more variation in what we’re producing, the more difficult it is to produce anything efficiently or at high quality. This is why I push for one case study. More case studies require far more moving parts in the production system… which leads to many more defects and a nearly impossible system to master.

• Suboptimization: In a factory, it doesn’t matter if one machine is perfectly optimized if the other machines are defective - you wind up with a finished product that sucks. What matters is the entire-system output. In our case, that means successful, retained customers. We should gear our entire factory towards that. What often happens because we don’t have the whole-system output in mind is that departments suboptimize and fight: Marketing optimizes for leads, sales optimizes for closed-won revenue, CS desperately tries to retain revenue, product is off doing whatever product does, and everybody hates each other because we’ve optimized each subsystem to the detriment of the whole.

• Demand: If we look at our business as a factory production system, it takes a certain amount of energy to move from start to finish in the assembly line. The difference between a physical product factory and a case study factory? Your customers co-create the case study with you. The more intense demand there is, the more energy THEY spend PULLING the case study through the factory (and navigating around imperfectly-designed steps in the factory), and as a result the less effort YOU have to spend PUSHING the case study through the factory.

• Unfolding: This is *not* a production-system concept, but it still fits. As our case study evolves through the unfolding process, our factory tends to need to evolve too. As we make small tweaks, the entire production system needs to reform to become coherent again.

Manufacturing thinking isn’t just for scaling businesses - it actually works really well for early-stage, pre-PMF startups. In the “infinity-to-one” stage, we can’t have a perfect production system. We can either have an empty factory OR a high-volume, high-variation, high-defect-rate production system from which we can eliminate waste and standardize output. Choose the latter - talk to a bunch of customers, sell them stuff, and iterate towards standard output and a low defect rate.

Manufacturing is so damn cool.

—

PS:

• My startup, Waffle, is launching a new firewall you can add to your AWS account that gets most stacks SOC 2 compliant in a few minutes — not months of eng work. See more here & email me rob@reframeb2b.com with any questions / to see if it can work for your stack!