Friday Commentary: Data Insights Or Cornflakes – What do you Base your Decisions on?
Estimated reading time: 4 minutes, 14 seconds
An insight is actually a source of information and is not data points. It is obtained by the analysis of data that directly impacts the business. Analysts thrive on data; executives need insights.
Understanding the importance of an insight is critical to actually listen to it. The more cornflakes (data points) that get spread around, the less each one means to us. Delivering data is a very difficult task in today’s data noisy society. Headlines in the newspaper or magazines report statistics and forecasts to create powerful headlines and sell more. We buy papers with the headlines; “Lose 20% of you weight in only three weeks” or “40% of your tax money is wasted in bureaucracy”.
Whatever the data points are, they are there as a reference. They are facts to build a story on. The average person in the street actually trust data completely without checking the validity of the source on so many occasions that we forget that we are losing the control over the impact is has on us. A classic example is that we rely in Google to check facts and figures and it seems that the increase of info graphics have made us more interested in data as it looks cool and interesting, but we are still not checking the source of data points.
Often the general press serve us data points that match with what we wish to hear. Some companies even seem to buy research with the intention and hope to see a certain outcome and ask the research company to help us find the answers. Buying cornflakes of different data points and then stringing them together into a story we wish to read, is a very false way of doing business research and it misguides the end interpreter.
In my everyday work reports and data points surround me. Most of my clients want the story before they know if the facts are true or built on solid thinking foundations. It still amazes me how many large cooperation’s and global organisations want to run before they can walk. In some ways, they seem to think that it is better to report something than nothing at all. I have seen media campaign tracking studies and web analytics reports that miss-guide a client into spending millions of Euros in the wrong direction. It makes me wonder if this result is because the agency felt obligated to show some “facts” and findings just to keep the client happy, or if it was because the agency simply didn’t have enough knowledge and resource to look beyond the obvious headlines themselves. Interpretation of data takes deep thinking time, good analytical expertise, and a deductive reasoning and some agencies are struggling to perform on the level needed for accuracy and unfortunately many agencies or consultants are not brave enough to stand up and say that they might not have the answer.
A recent survey by InfoChimps of 300 IT professionals revealed “80% of respondents said the top two reasons analytics projects fail are that managers lack the right expertise in house to “connect the dots” around data to form appropriate insights, and that projects lack business context around data.” Reference: http://sloanreview.mit.edu/article/predicted-to-perform-how-to-hire-analytic-talent/.
My top 5 advice on what is critical for success for data insights:
1. Understand that insights need to be built on a data collection that is solid and have a longer perspective than just spending time looking at a week or a month in comparison. Data collection sometimes needs months to become comparable and statistically accurate before any insights and conclusions can be pulled from the source.
2. Managers and teams need to be able to understand the insight and reasoning behind it. Successful business leaders and marketers today are successful because they understand the reasoning and caveats that come with it. They listen to the story and don’t try to build their own wishful biases that are not there.
3. Always know what goal you have and define your KPI:s that can define the progress and/or success. For example, calculate the value of your customers (new versus existing) in order to understand the cost you are willing to pay per incoming lead.
4. Surround yourself with good data insights people (internal and out sourced) who understand the data points; and critically analyse them before they start any interpretation. That is, they are neutral to the outcome. Ensure these people have a leading role in the organisation and can influence decisions taken because of their data credibility.
5. Never trust data before you know its source, accuracy limitations (for example, the survey respondent rate) is and the purpose for it being sent to you – did you request it, or is there a sales pitch behind it? Critical management doesn’t mean that you are negative, it just means that you want to know the background before you start spending your time analysing the numbers.