The best place to work … and other mythological places
There is an obsession with measurements, especially now, with the Digital Transformation fad. The most common practice in the business world is to measure everything and of course, It must be a number, NPS, Best Place to Work ranking… the magic number. Did you think about the consequences of taking this too far? Black Mirror, an award-winning Netflix series, invites us to reflect on the consequences of the abuse of technologies taken to the extreme, it is the modern version of Isaac Asimov’s stories. Let me tell you a real story.
Coming back to reality
A few years ago, doing a consultancy for an international holding, I ran into a large wall in the entrance hall that read “Best Place to Work 2016”, I was happy, it was the first day of the consultancy. The offices had been remodeled recently, it was a gigantic open space, shared tables everywhere, espresso machines, a relaxed atmosphere, thousands of sticky notes decorating the walls…I thought: “This is gonna be great”. I didn’t know what was expecting me.
When alms are high, even the saint suspect — Anonymous
After just 2 weeks of work there I realized that this company was not the “Best Place to Work” as I had read on the wall. The first thing I did was talk to the teams, the teams are the ones who suffer the most from organizational inefficiencies, they know all the hacks to survive. Following the path of the crumbs, I discovered, for example, 2 Business Units with objectives that directly competed with each other. It is like telling a person to store as much water and another to consume as much water as possible, and to “improve” the situation you create an economic incentive. A true example of organizational schizophrenia.
You get what you measure
During my investigation on incentive’s structure, I realize, for my surprise (or not so much) that part of the bonus was associated with the Best Place to Work’s survey results, in other words, if I completed a survey with good results and the company obtains better ranking position that the year before, it automatically guaranteed monetary income. This is an example of an easy and effective way to generate perverse behaviors within an organization.
Other business units use the NPS or Net Promoter Score for their products and services, an indicator that became popular in 2003 with the Harvard Business Review article “The One Number You Need to Grow” and is currently widely used by the industry, although this one also has its technical problems. Regardless of studies that challenge evaluation of the indicator, the problem is not the indicator itself, but people’s incentives to improve it, as mentioned in Goodhart’s law
All metrics of scientific evaluation are bound to be abused. Goodhart’s law […] states that when a feature of the economy is picked as an indicator of the economy, then it inexorably ceases to function as that indicator because people start to game it.. — Wikipedia
This behavior could be related to what Daniel Pink explains in his best-selling book “Drive: The Surprising Truth About What Motivates Us”, a study concludes that when tasks are mechanical, monetary incentives show outstanding results on the performed task, but when tasks imply cognitive efforts or creativity, extrinsic incentives greatly worsen the results.
Does your work require creative or mechanical thinking? If you respond creative, are your main indicators of success and incentives based on standard numerical indicators? Have you figured out how to hack those indicators yet? Share on the comment section below
AI to the rescue of metrics
When we talk about things like the best place to work or customer satisfaction, we are talking about intrinsically complex things, so there must be more than one variable to measure it. In 2016, the University of Cambridge published a document “The Fallacy of the Net Promoter Score” in which they argue problems with the NPS measurement’s method and propose an alternative model that incorporates, among other things, the voice of the customer. They resolve the analysis of the information using machine learning models and algorithms, and they extract the emotional tone from customer feedback comments after that information is quantified to weight it with other variables such as income.
But there is a problem with artificial intelligence, and that is that, although it is very useful for processing historical data and generating more accurate projections, the human component is much more complex than certain economic variables or comments. The fact I behave in some way in the past does not mean I will behave the same in the future. Again we are trying to simplify something that is intrinsically complex. Maybe there is a better way to understand human qualitative information
Deal complexity with complexity
Another problem with the Best Place to Work and NPS surveys is that they make very explicit what they are trying to measure…in fact, they have it in the name! By declaring the objective of the survey in an explicit way, you are increasing the chances of been hacked. That said, is there a way to measure something without making it so explicit? After thinking about it for a while and in what seemed to be a dead-end, I visited Designers and UX colleagues, they think differently and have experience in taking into account the customer perspective. That’s how I found The love/breakup letter.
The love/breakup letter is a qualitative elicitation technique in which you ask consumers to write a love/breakup letter to a brand or company. This method allows us to understand emotions and feelings based on concrete experiences. We tried it on some clients and it’s extremely fun and enlightening at the same time.
There is an issue with this technique and is that once the experiences are captured then we need to analyze the data and get insights, and by doing that we are trapped by the infamous cognitive biases. When we analyze the letter under the “eyes of experts” the most likely to happen is we lose perspectives of highly abstract and creative information. This technique is fine for elicitation but incomplete for the signification. But … Is there any qualitative or/and quantitative method that takes customer experiences without been bias?
Dave Snowden, a world reference in knowledge management and complexity, proposes through complexity methods and tools, to delegate the interpretation of their own stories to the people, in other words, allow them to capture and interpret at the same time, in this way the risks of cognitive biases are reduced. After that, you can present the information to them, in such a way that you allow them to seek emergent patterns without being influenced by expert bias
As an example, to the Employee NPS question “On a scale from 0–10 how likely are you to recommend our organization to your family or friends?” ( btw recommend organizations is not a common topic when I chat with my family or friends) Snowden proposes an alternative… “What story would you tell your best friend if they offered him a job at your organization?” Then you provide tools to people so they can interpret their own experience. This sensemaking technique is an alternative to the already known hackable surveys.
Story of universal metrics
In 1955 a supercomputer was able to predict the outcome of the United States presidential elections. It did this by selecting a single voter and asking her a series of simple questions. That supercomputer was the MULTIVAC and it was the imaginary of Isaac Asimov who gave it life. 65 years have passed and still, that technology does not exist and most likely never will. We love to dream of the possibilities of predicting the future, but we forget that the future does not exist. The best thing we can do is understand the present, describe the present, and at the most identify the adjacents possible.
Computers are better than people in some tasks, but a person with a computer always beats the computer. People have emotions, feelings, intelligence, we imagine, we dream, we are wrong, we are perfectly imperfect. Let’s learn how to deal with the imperfections and complexities of human nature, perhaps in that way, we will be better understanding organizations. There are no best places to work, there are some places better for some people or others. Try to accept the idea that maybe there are things that can’t be measure by numbers.
Not everything that can be counted counts and not everything that counts can be counted — William Bruce Cameron
