Misreading the Tea Leaves

Misreading the Tea Leaves: Editor’s Note – This article follows a simple format.  The first section is about a fictional company Smart Kitchen and Cindy their CHRO.  This story helps us present real-life situations where biased decision making occurs.  After we introduce a concept through Cindy and her colleagues, we will present the science behind the bias and what you can do as a business leader to mitigate it.  Finally, we will close out the story, and see how Cindy and her colleagues resolved their situation.

Misreading the Tea Leaves

Misreading the Tea Leaves – Part 1

Cindy had to be careful to not let her mind wander as she drove up the mountain road.  It was going to be an interesting few days for her and the Smart Kitchen team.  She knew this past year had been a rough one for the company, and for her colleagues on the leadership team.  As the Head of HR, she was focused on turning the leadership team into a high performing team, but so far, her attempts to improve their performance were meeting resistance.

So, when Jesper (the CEO) announced they would all do a strategic retreat, Cindy knew it would be the ideal time to get the team working more effectively, especially when it came to making strategic decisions.  In preparation for the meetings she had worked with her HR team to identify some of the biggest issues facing the leadership team and the company.

Overall, the company was seen as too conservative.  While they were incredibly innovative in the technology area, Smart Kitchen wasn’t known for taking risks.  The perception of the leaders was that they were too confident in their past successes, and therefore too set in their old ways.  If they were going to bounce back from a tough year, Smart Kitchen would have to be willing to try something new.

When she finally got to the mountain lodge and got inside, she noticed that everyone else had already arrived.  She quickly walked to the meeting, and sat down next to the Head of Manufacturing, Ted.  She had always liked Ted, and he seemed to really appreciate HR’s perspective on the business.  Sometimes he could be a little dominating, but overall, he was a good guy.

On the other side of Ted sat Sofie, the Head of Marketing.  She was relatively new to the company, having just come from a small Bio-Tech company.  Even though she had no kitchen appliance experience, Cindy had championed for Sofie to get the job, and so far, she had brought a great new energy to the team.

Rupert, the Head of Sales, came in chatting with Jim, the CFO.  Both seemed to be taking their time talking about their weekend activities.  No one was really paying attention to the CEO, Jesper, at the front of the room.  Cindy thought he looked a little stressed, as he walked back and forth in front of the presentation.

“Now that everyone is here, we can begin” grumbled Jesper.  “While our Finance team is still finalizing our year end reports, I can give you a heads up.  The Christmas season did not give us the boost we needed, and we will end the year with a 4% loss.  We can’t let this happen again.  Over these next two days, I want to come up with a plan on how we can turn around our business.  But first I want to know why this happened.”  Everyone’s eyes darted from one person to another, seeing who would be the first to speak.

“For me it is pretty straightforward” chimed in Rupert (the Chief Commercial Officer).  “We got hit by the poor housing market. I mean, obviously there were some things we could do better, but the poor housing market meant that fewer people were refurbishing their kitchens.  In places like San Francisco, Seattle, & Houston, the housing market is exploding and our distributors are barely able to maintain their inventory.  Other big cities like Los Angeles, Philadelphia, and St Louis aren’t growing near as much, and my gut tells me that is why demand in those areas have stagnated.”

While the team reflected on the implications of this conclusion, Cindy reflected on how Rupert came to the conclusion.  It seemed a little too convenient of an answer and based upon his intuition.  If she was going to help the team make better strategic decisions, she would have to get the team to be a little more analytical about data.

“Maybe the housing market played a critical role, but I know Sofie (the new Head of Marketing) has just finished an in-depth customer analysis” stated Cindy, as she turned to Sofie.  “What does your analysis say?”

“Actually, I don’t know if it is related to the housing market” replied Sofie.  “I think it might be…”

(End part 1 of Misreading the Tea Leaves)

Misreading the Tea Leaves

Why you shouldn’t trust your gut

Your gut / intuition is not as great of a decision maker as you think.  Gut decisions are made because we see pattern or situation in the data.  From experience you have seen that X situation precedes Y result.  So the next time you see X situation arising, then you automatically start thinking about Y.  Our brains are built to see those patterns and make decisions from them.

And for many years our ability to see patterns has been at the heart of our evolutionary process.  Pattern recognition helped us know which food was poisonous and which was edible, and what are the signals for a dangerous predator nearby.  This was how we built tools and became the dominant species. Without pattern recognition, we would never learn from our mistakes or build off of our success.

But our ability to see patterns, can lead us to see patterns where none exist.  Here is a picture with 3,200 randomly placed dots.  As you will see in the picture, the dots are not perfectly distributed.  In some places there are clusters, while in other places there are large gaps.  Our brains have a hard time understanding randomness, and so the clusters and the gaps are perceived as part of a pattern (even though they are not).

Misreading the Tea Leaves - image Clustering-illusion-random-dots on http://cavemaninasuit.com

But this tendency to see false patterns is not limited to optical illusions.  We see patterns in all types of data.  For example:

  • During WW2, Londoners were certain that German V-2 rockets were targeting specific neighborhoods and following a preset pattern. The truth was they weren’t.  The bombs fell in random locations.
  • Basketball announcers love talking about a player having a hot hand. While the player may be making more in a current streak, statisticians continuously prove that hitting your previous basket has no bearing on you making future baskets.  There is no hot hand.

This tendency to see false patterns in randomness is called the Clustering Illusion.  In the business world, this can seriously derail your company.  Data on your market, your customers, and your competitors can be vague and full of missing information.  And that makes you even more susceptible to the Clustering Illusion.

We want to believe that when we have all the puzzle pieces laid out before us, we can easily see where the pieces are missing and decipher the bigger picture.  This isn’t the case.  What are random bits of data, could suddenly be used to justify a strategic decision.

That is why scientists and statisticians focus on the difference between Correlation and Causation.  Correlation is the scientific term to mean that two things have some relation.  For example, height and foot size are correlated with each other.  Tall people tend to have bigger feet, but that does not mean that one causes the other.  Short people can have big feet, and vice versa.

While correlations can be very valuable in making decisions, we should never assume that there is a Causation between the two factors.  Especially since random data can also form a correlation.  One website that I really love is called Spurious Correlations (http://www.tylervigen.com/).  Basically they show how some of the most random things in the world can be scientifically correlated (to an extremely high level).

For example, this analysis shows that the amount of Japanese passenger cars sold in the US is positively correlated with suicides by crashing motor vehicles.  But even though there is a 93% correlation (crazy high in scientific standards), no one actually believes that suicidal people are buying Japanese cars for the purposes of driving them off a bridge.  (Data sources – U.S. Bureau of Transportation Statistics and Centers for Disease Control & Prevention)

Misreading the Tea Leaves - image on http://cavemaninasuit.com

Another example shows that cheese consumption in the US is correlated with the number of people who die because they have become tangled in their bedsheets.  Aside from the ludicrousness of the correlation, I wonder why the Center for Disease Control & Prevention tracks the number of people dying from their bedsheets.  (Data sources – U.S. Department of Agriculture and Centers for Disease Control & Prevention)

Misreading the Tea Leaves - image on http://cavemaninasuit.com

Finally, my favorite analysis shows the ‘danger’ of Nicolas Cage.  You might not have known it, but the number of people who drown by falling into pools is significantly correlated with the number of films by Nicolas Cage.  Keep that in mind the next time Leaving Las Vegas or Conair comes on the TV.  (Data sources – Internet Movie Database and Centers for Disease Control & Prevention)

Misreading the Tea Leaves - image spurious-correlations-share on http://cavemaninasuit.com

This biased mindset towards patterns and the tendency to treat a correlation like a causation is something I saw a lot in leaders.  While in the middle of making strategic decisions, the leaders would suddenly start acting like Nostradamus.  They would comb through the P&L statements, competitor reports, and customer satisfaction results looking for indications that their opinions were right.  “When our US market does X, and our competitor’s price shifts to Y, then we will know the market is rebounding.”

In the story, Rupert may be right in saying that the housing markets in San Francisco, Seattle, and Houston are exploding along with the Smart Kitchen sales.  But that does not mean that those two factors are related.  Coincidence occurs a lot more than we are willing admit.  Without fully understanding the data, Rupert could be drawing the wrong conclusions about what the data is telling him.  That is why we need to have a better understanding of statistics.

Misreading the Tea Leaves

How we can start thinking with our heads and not our gut

So, what can we do if our brains are hardwired to see imaginary patterns and draw faulty conclusions?  In order to overcome these biases, we need to leverage statistics and analytics.  Now most of you when reading that already started to frown.  Statistics is one of those few classes in high school and college that somehow always managed to be taught by the most boring teacher.

I remember my early classes on statistics focused primarily on teaching the underlying math of statistical analysis.  We spent less time understanding what to use the different analytical tools for, and more time completing the calculations.  This gave an unfriendly taint to learning statistics, because few people really enjoy learning complication math equations.

On the other hand, many people like finding answers to difficult questions.  They want to know if one thing is related to another.  They want to be able to predict a result if they choose a specific action.  They want to understand what is important versus what is just background noise.  And that is what statistics can do.  When you can understand the purpose of each statistical tool, you can unlock almost any complex problem.

Now 15 years ago you would have had to obtain an advanced college degree to build up your expertise in statistics.  The good thing is that modern technology has rapidly improved our the easiness of running advanced analytics.  What used to be done by hand, or specialized statistical software, can now be done through Excel.  All you have to do is know which statistical tool to use, and how to interpret the result.

That is why we recommend business leaders receive basic statistics training.  The training should ignore all of the math and calculations (since that is automated through software).  Instead, leaders need to learn how to analyze problems they are facing, determine which source of data would help provide an answer, choose which statistical tool is most appropriate for the problem, and finally interpret the statistical results.

By knowing how to pragmatically use statistics, leaders will be able to find the true patterns amongst all the randomness.  But in reality, they cannot do it alone.  Your company is most likely sitting on a mountain of data, but most of it isn’t formatted in a way that is easily accessible.  The gold is in the ground, you just need help in mining it.

That’s why we recommend having an analytics team in every company, and connecting that team to the top leadership.  Your company’s leaders have difficult questions to answer, and instead of trusting their gut, they should have access to thoughtful analysis of hard data.

But that leads me to a whole new problem I have been seeing among leaders.  They don’t believe in statistics, even when it is presented to them.  They put far greater emphasis on their experiences and their own interpretations of the data.  It is the business equivalent of alchemy.  Science disproved it long ago, but for some reasons they just want to keep believing they can turn lead into gold.

While statistics can seem daunting, in reality these recommendations are not that hard to implement.  Training your leaders about statistics doesn’t require an outside training vendor.  A basic introduction to statistics can be taught by a knowledgeable colleague.  Similarly, an analytics team doesn’t need to be a gigantic investment.  Start with one person, and build it from there.

While these recommendations will require some investment from your side, they will help you in making decisions based on real data, and not just your feelings or your misrepresentations of the data.  Investing in making unbiased decisions is one of the most valuable things a company can do.  Think about the consequences if you don’t.  Misdiagnosing some strategic data will be far costlier then hiring someone who can analyze the data for you.

Recommended Actions:

  1. Identify your colleague which has the most knowledge about statistics. Have him/her help create a basic introduction to some basic statistical tools and how they can be used in your company.
  2. Put every leader who is working on strategic priorities for the company through the introduction to statistics. Follow up with them on how they are applying their learnings on the job.
  3. Build an analytics team that can transform your company’s data into something accessible that your leaders can use to make their decisions.

Misreading the Tea Leaves

Misreading the Tea Leaves – Part 2

“Maybe the housing market played a critical role, but I know Sofie (the new Head of Marketing) has just finished an in-depth customer analysis” stated Cindy, as she turned to Sofie.  “What does your analysis say?”

“Actually, I don’t know if it is related to the housing market” replied Sofie.  “I think it might be related to where our target customers tend to live.  San Francisco, Seattle, and Houston are all tech hubs, and those areas might just have more early adopters for our products.  I think the other areas are lagging behind because we haven’t been able to simplify our products to reach a wider customer base.  While tech enthusiasts may love our features, I just don’t know if we have nailed the right value proposition for the average family.”

“While I hate hearing that our products are positioned right” muttered Jesper.  “I would hate even more if we didn’t address something that’s failing.  What do you think Jim?”

Jim (the CFO) leaned forward in his chair, and said “I think both are partially right.  We’ve known for a long time that the housing market impacts our turnover, but I think Sofie is onto something with this customer analysis.  While I wish I could provide the answer this minute, I think this requires a little more analysis.  I recommend that my team combines Rupert’s sales numbers with Sofie’s customer analysis.  I got a couple of ‘stat-nerds’ that would love to take a crack at this.”

“Fine.  Tell them they got a week to pull the numbers together” replied Jesper.  “I know my reputation is one of trusting my gut rather than the numbers, but let’s be honest here.  Trusting our guts didn’t work this year, and we don’t fully know why.   So, another action I want us all to take is to partner with Jim’s team a lot more this year.  It’s fine to have opinions, but I want us to use a lot more hard data.”

Cindy couldn’t believe her ears.  She had always known Jim was a numbers guys, but she was surprised by Jesper’s turnaround.  Maybe this disaster of a year really woke him up for the need to change.  She knew it was going to be a tough offsite, but this was a great way to start it.

Jesper continued “But this meeting isn’t just about explaining the past.  We need to turn this company, and unfortunately, I’ve got more bad news.  This past weekend, I was playing golf with Matt from Fasttrack Distributors…”

Editor’s Note – The story of Cindy and Smart Kitchen will continue in the next blog.

Misreading the Tea Leaves

Highlight and Tweet it – Humans are built to recognize patterns, but that means we just start thinking there are patterns in the randomness.  Any leader who says they can read / interpret the market is probably falling under the Clustering Illusion.  #biasbreaker  #Cavemanthoughts https://wp.me/p9toKb-29

Leave a Reply

Proudly powered by WordPress | Theme: Baskerville 2 by Anders Noren.

Up ↑

%d bloggers like this: