Product strategy: understanding chain-link logic

This work-in-progress post is part of a set on the topic of mission, vision, and strategy.

“A system has chain-link logic,” says Richard Rumelt in Good Strategy, Bad Strategy, “when its performance is limited by its weakest subunit, or link.” Rumelt’s example: the space shuttle Challenger‘s performance was limited by the rubber o-rings that sealed the joints between sections of its fuel tank. When the o-rings failed, the rocket exploded. Improvements to other parts of the system, such as the boosters or avionics, wouldn’t have made much difference unless the o-ring problem was fixed.

Other examples of systems whose performance is limited to that of their weakest link:

  • An assembly line can move no faster than the slowest step on the assembly line
  • A project can be completed in no less time than it takes to complete the “long pole
  • A literal chain is only as strong as its literal weakest link

Not all systems have their performance limited to that of their weakest link. Some systems have performance limited to that of the strongest component. An example would be humanity’s current effort to develop a vaccine for the virus behind Covid-19 — only one lab need accomplish the goal.

And some systems have performance limited by the combined contributions of many components. For example, consider the cables suspending the Golden Gate Bridge:

There are many cables. The amount of weight that the cables can suspend (assuming the towers don’t fail) is limited by the combined performance of all the cables.

Examples of strongest-contributor systems:

In chain-link systems, the way to improve performance is to identify the weakest link and to improve its performance or to reduce reliance on it. This is straightforward.

The tricky thing is when multiple components are below the desired performance bar. In that case, fixing one component won’t be enough to right the ship. It might not even make a noticeable difference. The system can then get stuck in a low-performance state.

Rumelt gives an example. Marco Tinelli is the GM of a machine company near Milan. Over lunch, Tinelli told Rubio the story of his company.

When my uncle passed away, the responsibility for the company passed to me. Things were not good. The quality of the machines had declined, especially compared with our best competitors. Costs were too high and the sales personnel were not technically sophisticated. To sell a sophisticated machine with microprocessor controls takes a sophisticated salesperson. If we didn’t change, we would slowly go out of business. But it seemed as if everything had to change. Where to start?

As he spoke, I realized that Marco Tinelli’s diagnosis was that his machinery company had chain-link logic and that it was stuck. Any payoff from better-quality machines was diluted because the sales force could not accurately represent their qualities and performance. A better sales force, by itself, would have added little value without better machines. And improvements in quality and sales would not save the firm unless costs were reduced.

To fix the problem, Tinelli had to see it clearly. He had to identify the three performance-limiting factors (quality, cost, and sales) and improve them all, knowing that overall company performance might not improve until all three are fixed.