Show Me the Money - a Simple Framework for Better Decision Making

Note: Photo attribution for the 2017 Tesla Model X image to: Michel Curi, Gold Coast Art Fair, Grant Park, Chicago, IL.

This blog marks a series of blogs that I will do on better decision making influenced by the book, How to Measure Everything, by Douglas Hubbard.

decision making tesla2017 was the year I agonized over buying a Tesla Model X. I ran through all the reasons I needed it - beautiful, gorgeous, drives fantastic, best technology and green. I just couldn’t come to a conclusion, because the price was a serious consideration. The problem at hand was that the reasons I wanted to buy it were intangibles. 

That brings us to a key component in decision making - what do you do when you are facing a decision and there are numerous intangibles at hand? Paraphrasing Lord Kelvin, the famous British physicist, “The challenge of intangibles - when you can measure something and express something in numbers, you know what you are speaking about. If you cannot speak about it in numbers, then the knowledge is meager and you have scarcely advanced the state of science.” 

As I looked around in other areas of my life, I saw that there were a number of other decisions that I couldn’t make progress on. They were also stuck because of intangibles. Some good examples that come to mind in the life of a product manager are: the value of open source software, when your product is driven by OSS; the value of adoption of open source software; what does virality mean for success of a product - how much should we be investing in virality; how to choose between competing features to make the most impact for the organization. I could talk to these decisions very well, had anecdotal evidence and experts backing those decisions…but they fundamentally were intangibles.

I have always been fascinated by literature on decision making and find that most recommendations range from “Just do it” to mind-numbing, if-then-else analysis. Douglas Hubbard presents a scientific, probability-based framework to reduce uncertainty through your decision making progress.

Paraphrasing Hubbard, “Decisions fundamentally are about a choice in the path ahead where decision makers have imperfect information. This lack of information causes uncertainty. Measurement is a type of choice among others to reduce this uncertainty. For any decision, you can typically measure a large combination of things, but you can never achieve perfect certainty.” 

Thus, the right way to think about decisions is that a decision is a decision because one cannot articulate the right value of each of the choices ahead. The way to reduce the uncertainty, then, is to measure the things that reduce that uncertainty, knowing fully well that you cannot - and should not - measure everything that informs the decision, because too much information causes information overload. The key part of measurement is to assign an economic value to the measurement. Let’s say, you as a decision maker are looking to improve delivery of value from the organization. Measuring productivity of employees could be a type of measurement, and instead of measuring how much time an employee spends on each of the activities, quantify the business value of each of the areas she spends most of her hours on. Thus, if you get the employee to spend more time on activities that drive more business value, you can get more productivity from the system.

So how does one proceed, so that one isn’t stuck at the end of a spectrum that has “just do it” as the starting point and “analysis paralysis” as the ending point?

The ladder of better decision making starts by stepping back and questioning the decision. Check if you are asking the right question. Most hard problems in life and business can be solved by asking the right questions. If you are indeed asking the right question, then do the following (Hubbard calls it Applied Information Economics; with some color from my past readings on this topic).

  1. Define the decision. If the decision isn’t informing a significant bet, don’t measure anything. If the decision is easily reversible, just make a decision and move on.
  2. Determine what you know about the decision. Identify the key informational vectors that inform you more about the decision (hours working, vs hours spent on the most impactful activity).
  3. Compute the value of additional information (if none, go to step 5). Additional information is the information that you need to drill down into and measure.
  4. Measure where information value is high (return to steps 2 and 3). High information value implies tying some sort of economic benefit to it (improving productivity implied measuring and identifying most impactful hours). This really is the key to better decision making - tying some sort of economic value to your intangibles.
  5. Make a decision and act on it (return to step 1 and repeat as each action creates new decisions).

Typically, as you do steps 3, 4 and 5, you will find that each decision opens a cascade of micro-decisions that may have their own measurements to help build the case for the overarching decisions that you will rinse and repeat on. 

So how did this framework help me?

  • I made the call on the Tesla before I read the book and was pleasantly surprised that my mental model roughly followed the framework. I put an economic value on each of the subjective axis, compared with the cost of the vehicle and came up short. I passed over the Tesla for a Lexus RX 350, knowing full well that I will evaluate my decision in the year ahead and these values may change - c’est la vie. 
  • On the “business value” delivered by my team, I have instituted a “business value pointing” system offered by a tool called Aha. The vectors that constitute the business point are subjective. That is okay, because the scores on the vectors are informed by the expertise that my team has built up so far. We are drilling down on step 3 right now.
  • On understanding the “value of adoption of OSS” or “virality of a product” - I am on a journey and am hoping to find the right measurements to taking the decisions from “expert-led” to a state of science.

What decision frameworks work for you? Let me know in the comments section, below.

Harpreet Singh


Harpreet Singh is vice-president of products at CloudBees .
Follow him on Twitter and LinkedIn.