Home Artificial Intelligence 5 Questions Every Data Scientist Should Hardcode into Their Brain Hammer time Problems > Tech 5 Problem Discovery Questions A Key Lesson Conclusion Resources

5 Questions Every Data Scientist Should Hardcode into Their Brain Hammer time Problems > Tech 5 Problem Discovery Questions A Key Lesson Conclusion Resources

5 Questions Every Data Scientist Should Hardcode into Their Brain
Hammer time
Problems > Tech
5 Problem Discovery Questions
A Key Lesson

After I began my data science journey in grad school, I had a naive view of the discipline. Namely, I used to be hyper-focused on learning tools and technologies (e.g. LSTM, SHAP, VAE, SOM, SQL, etc.)

While a technical foundation is obligatory to be a successful data scientist, focusing an excessive amount of on tools creates the “Hammer Problem” (i.e. when you might have a very nice hammer, all the pieces looks like a nail).

This often results in projects that are intellectually stimulating yet practically useless.

My perspective didn’t fully mature until I graduated and joined the info science team at a big enterprise, where I used to be in a position to learn from those years (if not many years) ahead of me.

The important thing lesson was the importance of specializing in problems slightly than technologies. What this implies is gaining a (sufficiently) deep understanding of the business problem before writing a single line of code.

Since, as data scientists, we typically don’t solve our own problems, we gain this understanding through conversations with clients and stakeholders. Getting this right is essential because, for those who don’t, you may find yourself spending plenty of time (and money) solving the flawed problem. That is where “problem discovery” questions are available in.

6 months ago, I left my corporate data science job to develop into an independent AI consultant (to fund my entrepreneurial ventures). Since then, I’ve developed an obsession with cracking these early-stage “discovery” conversations.

My approach to recovering at this has been twofold. First, interview seasoned data freelancers about their best practices (I talked to 10). Second, do as many discovery calls as possible (I did about 25).

The questions listed listed here are the culmination of all my previously mentioned experiences. While it’s certainly not an entire list, these are questions I find myself asking over and all over again.

1) What problem are you trying to unravel?

While (in theory) this ought to be the one query on this list, that (unfortunately) just isn’t how things work out in practice. Repeatedly, clients aren’t clear on the issue they need to unravel (in the event that they were, they probably wouldn’t be talking to a consultant). And even in the event that they are, I often have to catch up to grasp the business context higher.

This query helps in each cases because (ideally) the client’s answer prompts follow-up questions, allowing me to dig deeper into their world. As an example, they could say, “We tried making a custom chatbot with OpenAI, but it surely didn’t provide good results.”

To which I’d ask, “What was the chatbot used for?” or “What makes you say the outcomes weren’t good?”.

2) Why…?

A natural follow-up query to “what” is “why.” That is one of the crucial powerful questions you may ask a client. Nonetheless, it may well even be one of the crucial difficult to ask.

Why” questions tend to make people defensive, which is why having multiple ways of phrasing this query may be helpful. Listed below are just a few examples:

  • Why is that this necessary to your small business (your team)?
  • Why do you must solve this now?
  • What does solving this mean for your small business?
  • How does this fit into the larger goals of the business?
  • Why do you must use AI to unravel this problem?

This query (or any of its variants) is an especially effective strategy to get context from the client, which should (again) spark follow-up questions.

To proceed the instance from before, the client might say, “We’ve several support tickets that we wish to categorize into 3 levels of prioritization robotically, and we thought an AI chatbot was a very good strategy to solve that problem.” This provides far more context to the “We tried making a custom chatbot” response from before.

What are we doing?” and “Why are we doing it? are the 2 most fundamental questions in business. So, getting good at asking “what” and “why” can take you (very) far.

3) What’s your dream end result?

I like this query since it (effectively) combines the “what” and “why” questions. This enables clients to talk to their vision for the project in a way that will not come through when asked directly.

To learn something recent, I often have to take just a few passes before it finally clicks. Similarly, I find that to really get to the basis of a client’s problem, I want to ask “what” and “why just a few times in other ways throughout the conversation.

That is harking back to Toyota’s “5 Why’s” approach to attending to the basis explanation for an issue. While this was developed in a producing context, that is something that readily applies to problem-solving in data science.

Two related questions listed here are: What does success appear to be? and How would we measure it? These are a bit more pragmatic than a “dream end result” but are helpful for transitioning from asking “what and why” to “how?

4) What have you ever tried up to now?

This query starts on the trail toward an answer. It does this by bringing out more of the project’s technical details.

As an example, this (typically) gives me a very good idea of who can be writing the code. In the event that they’ve already built some basic POCs, then the client (and their team) will probably be doing a lot of the heavy lifting. In the event that they are ranging from scratch, it is likely to be me or sub-contractors from my network.

On this 2nd scenario, where the client has built nothing up to now, one can ask just a few other questions.

  • What’s the present solution?
  • How do you solve this problem now?
  • What have others done to unravel an analogous problem?

5) Why me?

I got this query from master negotiator Chris Voss. Who frames it as an efficient strategy to reveal people’s motivations.

Often, this sparks additional context about what led them to you and how they see you fitting into the project, which is useful in defining the subsequent steps.

Sometimes, nonetheless, people don’t have good answers to this query, which can indicate that they don’t actually need to work with you and are holding back their true intentions (e.g. they need free consulting or a competing bid).


Please enter your comment!
Please enter your name here