A Problem Well Stated is Half Solved

A Problem Well Stated is Half Solved

Post by Ray Poynter, October 2019

When I see people struggling to turn data into findings, and findings into stories, and stories into action, it is often because they do not really understand the question they are seeking to answer.  If you do not know what you are looking for, you are unlikely to find it.  And, if you find it, you are unlikely to recognise it.

When designing a research project, it is essential that the real problem, the underlying question, is accurately defined.  When I teach my workshop on finding and communicating the story in the data, I share an eight-step framework.  The first, and most important, step is defining the problem.

Key steps in finding the underlying problem

The key to finding the underlying question that needs to be solved is to discuss it with the person who has the problem that is being solved.  Some of this discussion should happen before the project is commissioned, but the discussion should continue through the design process and, indeed, be further refined during the data collection and analysis.

Key sources for finding the underlying problem are:
  1. Reviewing the request for the project
  2. Finding out what is already known (including what is ‘believed’)
  3. Have there been other attempts to solve this problem?  If so, what can we learn from those projects?
  4. Asking why the project is happening now (and why it did not happen before, why it does not happen later?)
  5. What actions does the business plan to do after it conducts the project?
  6. What would the business do if it did not conduct the project?
  7. What would success look like?
For example

If a hotel chain brand manager says to the insights team that they want to know the drivers of choice in the hotel business, the insights team needs to dig into the real issues.  Unless one is an academic, wanting to know the drivers of choice in the hotel business is not a real business question, at best it is a research question.

When the insights team probe the request, the brand manager might say something like “Because I want to increase the number of repeat business visitors.”  This is a good start, the team now knows the domain - business travellers - and within that domain, the interest is in repeat visitors.  They also now know that the business problem relates to increasing the number of visits (as opposed to, say, increasing prices or selling additional services).

The next question from the insights team might be “What do you plan to do when you have the results from this research?”  The reply might be “I will do a cost/benefit analysis and implement those ideas that show a positive return and which I have the resources to implement.”  Now the team knows that they need to produce options that can be costed and which are realistic in terms of the sorts of resources to which the brand manager has access.

Another useful question the team might ask is “What would success look like?”  The brand manager might say “A short prioritised list of options.”  Or, they might say “A long list of options, with the benefits and risks clearly outlined.”  Researchers need to be aware that different people have different definitions of success.  Some clients want a range of options, some want two options, and some want a single, clear recommendation. 

Problems in finding the underlying question

Whenever I teach this part of the framework, the most common questions I receive are about the difficulties insight professionals sometimes face in identifying the real business problem that is being solved.  This is particularly acute for research agencies who are not liaising with the owner of the business problem, but with the insight manager, who may not know the underlying problems.

The questions listed above in the key steps are still a good starting point.  In addition, you can try even harder to find out what the key topics are that are surrounding the business.  For example, have they been in the news, what is their share price doing, what does their website say their key aims and intentions are?

The final step, is to be bold and combine the following two elements:

i    Say, I think the underlying problem you are seeking to answer is …, is that right?
ii   If I deliver X, Y, and Z, will that be what you need?

Background to the quote

The quote “A Problem Well Stated is Half Solved” is usually attributed to Charles Kettering, who was head of research at General Motors from 1920 to 1947.

Join our mailing list