top of page

Warren Berger’s Beautiful Questions or why every good analysis starts with a good question



written by human, not by AI




Picture this: you get your hands on a bunch of data. Let’s say it is Sales data. In and of itself, this dataset doesn’t tell you anything. It is information in a raw form, often organised in columns and rows. But it doesn’t mean anything. Then, something sparks in your mind. A process starts, likely driven by a powerful force called Curiosity: you wonder about something, and that wonder takes the shape of a question. “What is this dataset about?”, “What can it tell me?”, “What are its limitations?” “What can’t it tell me?”, Why am I looking at this dataset?”, “Where do I begin?”


No good analysis ever has started without questions. The act of analysing is, in itself, the same as asking questions. One of the reasons we analyse is to find an insight. But what is an insight? According to the Cambridge Dictionary, an insight is:


“(the ability to have) a clear, deep, and sometimes sudden understanding of a complicated problem or situation.”

But there is very little that is clear or deep, let alone sudden, in data analysis - at least when you start. It is often a painfully crafted method of searching and researching until a nugget of wisdom reveals itself. And it takes a whole lot of questions.


I like to imagine each question geared towards analysis as the light of a small flashlight while my data is in a dark room. If I have a narrow question, I’ll have a small ray of light, pointing just one spot dimly - and I’ll very likely miss a whole lot of what is happening in that dark room. If my question is more open-ended, I will have a broader light beam, and I’ll be able to cover a larger surface area of the dark room - but even then, I’ll only be able to see what’s right in front of my light beam and nothing else. The more questions I ask, the more bits of the dark room I’ll uncover, which will enable me to connect the pieces to help form a fuller picture of what the room would look like when in full daylight.


Questions allow me to go through this room without tripping too much on the objects scattered on the floor as I dig deeper into what may be hidden in its corners. Questions can also help illuminate our inherent biases: if I entered this dark room with a pre-conceived notion of what I was going to find, I would try to use my flashlight to quickly look for the object that would confirm what I expect - but if my questions are good, I’ll end up shedding light on a bunch of other unexpected things.


As someone analysing data every day, questions are my ultimate tool - even if I have no other tool nearby, I can uncover information about any subject, situation, or event just by asking a lot of questions. But questioning in this more exploratory way is almost a lost art - one that we mastered as kids but were un-taught as we grew up, and as we become analysts, we have to learn again.



The lost art of asking beautiful questions


Warren Berger calls himself a questionologist. He wrote two books on the subject of asking questions: “A More Beautiful Question” in 2014 and its successor “The Book of Beautiful Questions” in 2018. I read both together, and I believe they are actually better enjoyed if read back to back, so this is a double review.


In “A More Beautiful Question”, the author talks about the art of using questions as a discovery tool. He mentions several times across that text that as children from the ages of two to four, we ask a whooping average of 40.000 questions. But as we grow up, this figure tends to fall sharply, especially after we enter school. He argues how, as we grow, we are expected and rewarded (by the school first and then by work) to always have answers but are punished for asking too many questions. Questioners carry a stigma of being disruptive or inconvenient. Often, people fear that by asking questions, they’ll display vulnerability. And the more we grow into the expectations of our societal roles, the fewer questions we ask.


But as we enter into our roles as analysts in the workforce, we are required to be able to inquire again. Curiosity, for example, is considered one of the 3 C’s of Data Literacy - alongside Critical-thinking and Creativity. The one thing all these three skills have in common is questions. We ask because we’re curious about something. Upon closer inspection, and while trying to gain deeper knowledge about the object of our curiosity, we keep asking and analysing it in different ways, sparking our critical thinking. And when we connect this newly acquired information with other pieces of knowledge and start inquiring “what if”, we get Creative. So, if this is the case and these skills are now required of us, why does it seem so hard to ask questions as naturally as a four-year-old?


Berger explains in his book that questions are powerful challengers of the status quo: by asking, “why do we do things this way?” or “what if we tried doing things differently?” we challenge pre-defined norms. And while there may be an understanding in companies today that an amount of disruption is required to drive evolution - nobody likes having their ways of doing things challenged. As the author explains:


“(…) questions challenge authority and disrupt established structures, processes and systems, forcing people to have to at least think about doing something differently. To encourage or even allow questioning is to cede power - not something that is done lightly in the hierarchical companies or in government organisations, or even in classrooms, where a teacher must be willing to give up control to allow for more questioning.”

A More Beautiful Question” is a wonderful manifesto for reviving the lost art of asking. It goes deep into why we don’t ask questions as frequently and offers some frameworks and exercises we can adopt to re-learn our lost ability to look at the world with inquiring eyes. It is a fantastic book because it is theoretical and practical at the same time. It is written in a very approachable way and is primarily concerned with asking questions instead of providing answers. In a world full of books offering quick, easy, bite-sized answers as one-size-fits-all solutions, Berger wrote two dedicated to asking the right questions first - because you can only get beautiful answers if you ask a more beautiful question.



How to move from asking to doing?


Berger delves deep into what questions are, how they work, how we perceive them and how they can be instrumental when our work is to turn data into knowledge. He also offers a few frameworks we can follow to help us frame questions to enable action. Here are two that I really like and often use:


The Why-What if-How framework:

  • Start by framing the situation as a Why question: instead of asking analytical questions more traditionally, such as: “What are our Sales Numbers this month versus last month and against target?” try asking them in a more open-ended form: “Why is it that our Sales seem not to have a clear trend upwards or downwards?”.

  • Surface ideas and hypothesis using What If questions: the second step involves theorising situations from our Why questions. “What if Sales are actually seasonal and comparing months is not enough?”; “What if Sales were down?”; “What if Sales were up?”; What if Sales were more constant across the months?”;

  • Evolve your What if question into a How question: this will help propel the questions closer to an actionable form. ”How can we guarantee Sales Trend up, despite seasonality?”; “How can we prevent Sales from falling?”; “How can we boost our Sales growth?”; “How can we make sure Sales volumes are spread more evenly across the year?”


This can help us see problems differently and gear them towards a more actionable form. It is particularly useful when trying to uncover the real requirements of a group of clients or stakeholders asking for an analytics product (like a dashboard, for example).


The second framework consists of a multi-step exercise that helps generate questions around a theme, produce questions, curate those questions, select the best ones and then reflect on lessons learned from the activity. It is helpful to help define and prioritise how a problem can be framed and which questions are relevant and desirable enough to be pursued.

  • Design a provocative Question Focus: for example, “Sales Are Down”. or “We won’t reach our Sales Targets this Financial Year.” This statement must not be in the form of a question. It must be an affirmative or negative short sentence that is thought-provoking.

  • Produce Questions: give everyone a few minutes to come up with as many questions as possible related to the Question Focus. No discussion or comment around the questions is allowed. Ask away. Everything goes: Why are Sales Down? Are Sales Down? When Are Sales Down? What if Sales are not down? Does it matter if Sales are down? Are the targets achievable? Are the targets correct? Where did the targets come from?

  • Improve the questions: everyone must then try to change their questions a bit, either opening or closing them. An open question will have a broad answer - why, how, what if - a closed question is often answered with a quick yes or no. Example: open: “What if Sales are not down?” (many possible answers) closed: “Do we want Sales to go up?” (yes or no?)

  • The group picks their three to five favourite questions from the whole bunch: This helps prioritise the best way to inquire about a subject.

  • The group discusses and decides on the next steps for action on the asked questions: how will they be investigated and potentially answered?

  • The group reflects on what they learned about the discussed subject: there will be heaps of new ways to look at an old subject!


While this may seem like a lot of time and effort, it can help narrow down big problems into more fundamental steps of inquiry, making an analysis better directed and defined from the get-go.



Why can’t you just give me the answers?


If you have been doing data analysis for a while, you probably had at least one boss or client that told you this: Why can’t we just have the answers? All this asking is making everyone wary that you’re wasting precious time. Business means speed, and everyone is concerned that this is deviating from the objectives - which are all defined as do, do, do - with no time for a little ask.


Berger also talks about this phenomenon, which makes him wonder: “why are you evading inquiry?”. He lists four main groups of justifications people give as to why they don’t ask questions and explores them in a bit more detail:

  • Questioning is seen as counterproductive: it’s the answers that most people are focused on finding because the answers, it is believed, will provide ways to solve problems.

  • There is never time: the right time to ask questions never seems to present itself; either it’s too soon or too late.

  • Knowing which questions to ask is difficult: so it’s better not to ask any questions.

  • What if we find we have no good answers to the important questions we raise? It’s better to avoid such uncertainty.


We are obsessed with answers. Since we’re young, answers are rewarded in the form of tests, interviews, and yearly performance reviews. Nobody gets promoted because they asked good questions. That is until they yield very good answers. The quality of your outputs will always be tied directly to how long you spend sharpening the tools you’ll use to craft them. Your main tools as an analyst are questions. Make time for them - they’ll save you time in the long run.



The Book of Beautiful Questions


Warren Berger asks lots of questions in “A More Beautiful Question”, but most of them are formed to help us understand how questions can be powerful to help us. Perhaps in the spirit of giving some answers to his most avid readers, who probably asked for more examples of such beautiful questions, he followed with publishing “The Book of Beautiful Questions”.

The sequel to his first book is less philosophical and more practical. It offers us a range of questions we can borrow and reframe to ask our own more beautiful questions. He starts by defining who are the five biggest enemies of good questions:

  • Fear: of being judged by others or being perceived as ignorant;

  • Knowledge: or the “trap of expertise” because we don’t know as much as we think we know;

  • Bias: which filters our perceptions of the world, always looking to be confirmed;

  • Hubris: which can lead us to believe that our biases are correct, even when confronted with conflicting evidence; “if I don’t know it already, it can’t be that important”;

  • Time: or the supposed lack of it.


The author then presents us with lists of groups of gorgeous questions to help us challenge our biases, detect bullshit, asses conflicting information, open up possibilities and much more. It explores questions we can apply to make decisions, spark creativity, connect with others, develop stronger leadership skills and get deeper into the meaning of life and many other things. It is a book I often go back to for inspiration when stuck on what I should ask next. It is definitely a worthy successor to Berger’s impactful “A More Beautiful Question”.



Should you read it?


I recommend both of them to anyone who has to look for answers but is always left wondering what the actual questions should have been - which is the definition of anyone in an analyst role.


These two books will help you get more clarity into the process of asking questions, why they matter and how to get better at doing it. It will help you frame your questions to learn more efficiently, honing your Curiosity, Critical-Thinking and Creativity muscles.


Asking good questions is an often overlooked but important soft skill for any data analyst. If you have to choose, I would recommend going first with “A More Beautiful Question”, and once you’re ready to dive deeper into the art of inquiring, grab “The Book of Beautiful Questions”.



Always check your local library first to see if any of the books I recommend are available. If they’re not, consider donating a copy!

 

Get a copy at your local library | Amazon


 

If you subscribe to my monthly Newsletter, you’ll get a summary of all recommendations, plus more of my data viz musings.


You can also follow Data Rocks on LinkedIn or read these and other articles on Medium

 

written by human, not by AI

47 views0 comments
bottom of page