Is your actionable insight being limited by your communication skills?

Data-driven leaders understand that analytics are crucial to providing actionable insight. But an equally important part of that process is how that insight is communicated to others. The most effective communication is when there is an effective interaction between the use of data and natural language. In other words, the ability to clearly explain what is meant in terms of insights obtained is a critical success factor.

This post is written by Tony Boobier, author of “AI and the future of the Public Sector” amongst other books. As a regular guest blogger, he has shared many past posts to help leaders improve their communication & thinking. Above & below, Tony explains his latest challenge for data & analytics leaders. Are you communicating effectively? Back to Tony… Taken independently, natural language when used alone can be misunderstood especially if not adequately defined. Vagueness can have major implications.

Take for instance the Bay of Pigs incident in 1961 when President Kennedy asked the military for advice on whether to invade Cuba. The President was told that an invasion plan had a “fair chance” of success, and perhaps not unreasonably, the President took it as a positive indication. The landing operation was undertaken by Cuban covert forces funded by the US Government. The operation was a failure. When asked afterwards, the author of the recommendation said that the expression “fair chance” represented only a 30% chance of success.

Does your use of probabilities confuse your audience?

Sometimes the use of numbers also can be misleading. The notion of a 1-in-a-100-year flood does not prevent the possibility of flooding occurring in consecutive years. This description is no more than a statistical device for explaining the likelihood of flooding occurring. Similarly, when we check our mobile device for weather conditions and are told that there is a 90% chance of rain, this only means that on days like this with similar metrological conditions, it rained on 90% of them. As with the flooding comparison, it is simply a mathematical device which expressed the likelihood of an incident occurring.

Gamblers are adept at playing the odds. In poker, rather than thinking that getting three-of-a-kind is “unlikely“, they know that the chance of that happening is about 5%. If two cards are dealt and are a “pair” the chance increases to 12%. Though both probabilities are relatively small, the difference between them is enough to influence a gambler’s decision at the table.

It might feel strange to apply such probabilistic decisions (in both life and business) where data might be incomplete. The reality is that we are seldom faced with decisions about events with a binary outcome which will definitely happen or definitely not happen. Rather, there is a whole range of probabilities which sit in between. To describe these are being “likely” or “unlikely” fails to capture the nuances of a situation. Both those expressions also mean different things to different people.

Is your language communicating more certainty than is warranted?

Precisely allocating probability to situations which are both complex and uncertain can be difficult. At best you might only be able to compare them to similar situations you have experienced before. Perhaps this is when the traits of intuition and experience can add greater colour. Avoiding what might otherwise be a black-and-white data canvas.

In a complex business, in an increasingly volatile environment, there is seldom any certainty but even less tolerance for ambiguity. Using words alone to describe a possible outcome leaves the door open to misinterpretation.

Perhaps there is no simple answer? Those who insist on data being the only truth need to understand that when making predictions it is seldom that any absolute certainty attaches. They need to understand the potential ambiguity of the statistics. On the other hand, those who are wordsmiths at heart need also to understand that the use of words alone, such as ‘fairly high’ for example, can also be open to different interpretations.

How are you using words and numbers to educate your business?

The true data-driven leader needs to be sensitive to the strengths and weaknesses of the use of both words and data. As they communicate potential outcomes they need to appreciate the uncertainty of the imprecision of both. But they also need to be able to communicate such imprecision effectively, using words and numbers to provide that clarity of understanding.

Many thanks to Tony for sharing his thoughts with us. I agree that such communication skills are an important focus for data leaders. With all the focus on robust analytics (numbers) & data visualisation (pictures) let us remember that words are needed too. Data leaders would do well to reflect & seek feedback on the effectiveness of their written communication.

On a similar theme, I was glad to see that Cole Nussbaummer Knaflic’s latest book is “Storytelling with You”. A book which focussed much more on communication & presentation skills. What has helped you develop in this area? As a data leader, how have you learned to use both words & numbers to educate your audience? How have you overcome some of the challenges that Tony highlights (like explaining uncertainty or probabilistic outcomes)?