Soup Tasting

Ray Dalio’s Principles is required reading at Bridgewater, and contains plenty of wisdom that resonates well beyond its original context. Far down on the list, at #139, we find: 

... 139) “Taste the soup.” A good restaurateur constantly tastes the food that is coming out of his kitchen and judges it against his vision of what is excellent. A good manager needs to do the same.

Soup tasting is hard and requires you to pierce comfortable levels of abstraction.  Often where there are bad outcomes, there is a gross lack of soup tasting, both because of inertial unwillingness to take a bite and because of ineffective gustation.

Amazon’s Jeff Bezos is the archetypal soup taster (among many other outstanding talents).  Bezos is renowned for the depth of his knowledge and the clarity of his insights (especially when making snap judgments), but equally important is his ability to get to the crux of seemingly complex matters in five questions or less.  It’s easy to forget how many complex decisions Amazons has faced over the years, and the fact that their success is often taken for granted is largely a tribute to Bezos’ ability to ask the right questions so incisively and consistently.

The importance of soup tasting seems intuitive enough, but how you develop the ability to taste soup well is one of the more underrated challenges of leadership for a number of reasons.  To begin with, there is never just one kind of soup.  The metaphor applies equally well to the commercial success of your business and the view from inside.  At the same time, not all soup is equally important, and even the most astute taster’s capacity is limited, so you need a focal point. As Bezos has often described, “We start with the customer and we work backwards.”

More fundamentally, soup tasting is largely about overcoming bias, which is generally a very difficult process.  It needs to be about fearless inquiry, not seeking reassurances. Anyone who has done any actual cooking has probably had the experience of asking someone else if a dish tastes funny, while silently convincing himself that the answer is no. Of course, if it does taste funny, being polite does the aspiring chef no favors.  For soup tasting to have any value as an exercise, you can’t be afraid of what you might discover. 

Soup tasting is as much art as science, and as such it is hard to turn it into a predictable framework.  Still, some basic principles apply:

  • It all starts with probing.  Any time you are presented with an assertion, whether it’s a project plan, forecast, or report, review it tenaciously.  If something isn't clear to you, probe down. If something strikes you as particularly important, probe down deeper. If there are implicit assumptions, challenge them.  Think of the annoying little kid who responds to everything by simply asking “why?” It seems repetitive, but if you proceed from the right starting questions you will quickly get to the heart of the matter.
  • Get closer to the problem.  Something about the soup seems off.  Now you need to taste it some more.  The first step in getting close to the problem is simply a more thorough probing.  If that doesn’t do the trick, you need to go down two or three levels, either by honing in on the most important things in your area of credibility, or by asking someone who is credible.  By the way, assessing who has credibility in what areas, beyond just being aware of their reputations, is its own important form of soup tasting.
  • Measure.  Soup-making, both literal and figurative, requires experimentation, and it’s one of the hallmarks of the Amazon approach.  Bezos places a premium on experiments that are measurable and produce hard data.  As he explained in Fast Company, “The great thing about fact-based decisions is that they overrule the hierarchy. The most junior person in the company can win an argument with the most senior person with a fact-based decision.”  At the same time, as Bezos will quickly tell you, “there’s this whole other set of decisions that you can’t ultimately boil down to a math problem” – hence you need to master the art as well as the science.

It’s also well worth considering what soup tasting is not:

  • It’s not micromanagement.  This means telling people how to do something without tasting the soup for yourself, or telling them how to do something in an area where you lack credibility.  
  • It’s not distrust.  Distrust is not a productive default position, but neither is blind trust. Real trust is developed by consistent soup tasting – as the old saying goes, “trust, but verify”.  Knowing which issues to escalate as priorities, and how to escalate them as a team, is also an art form, honed through soup tasting interactions.
  • It’s not indefinite, nor is it an end in itself.  You need to find the middle ground between an excessively laissez-faire approach and never-ending inspection.

The more soup you start to taste, the more you'll want to taste, but as with anything, you can overdo it – just as you can proofread too long, and you’re bound to miss something obvious. It is critical to cultivate credible soup tasters throughout the organization, but the transition from soup taster to meta-soup taster is a tough one. It only works if your trust has been validated, and requires a great deal of intellectual honesty to avoid indulging in wishful thinking, feel-good exercises, or just shedding responsibility. 

In the end, soup tasting is how you know what is true – “overcoming bias” and “intellectual honesty” are really just fancier ways of expressing this.  And the truth matters more than anything else.  In his introduction to Principles, Dalio states,

“I also believe that those principles that are most valuable to each of us come from our own encounters with reality and our reflections on these encounters – not from being taught and simply accepting someone else’s principles…. So, when digesting each principle, please…

…ask yourself: “Is it true?”


All soup tasting, ultimately, is a variation on this one simple yet profound question.