You know the refrain; ‘what does the data tell us on this’, ‘if only we had the data, we would be able to understand what’s going on here’ or my personal favourite ‘stop sending me spreadsheets, my inbox is full!’.
The use of data for decision making is not necessarily a new idea, it has really been a fundamental engine of growth since the industrial revolution. Computers, networks and connected devices have simply accelerated the trend. Data is everywhere, we are addicted to it: our reporting suites, spreadsheets, databases. There is even some discussion about an on coming data storage crisis (although there are some interesting avenues of research).
Wherever you sit in a business, everyone is now being asked to analyse data to improve performance. Be it product sales, website clicks, operational volumes or even your own performance metrics. Reviewing data has quickly become a fundamental skill to succeed in the modern economy.
However, there is an issue…potentially a big issue.
Whilst data volumes are increasing, the hours in the day are not (really)! Data overload is common, and when overwhelming, it becomes an obstacle to effective decision making.
Fortunately, the world of engineering and science can help for guidance. After all, they have a long history of dealing with complex data.
- Maximise use of existing data. All too often we are on a quest to get more. More datapoints, more segmented data and more detail. However before you embark on this expensive journey, have you maximised the conclusions you can draw from your existing data? Take for example the recent Pluto flyby. The space probe left 10 years ago, well before the invention of the iPad, there was no question of going back to update the instruments. However, using updated techniques (for example image processing), new meaningful analysis can be gathered, enhancing our understanding of the solar system. The same is true of a business process.
- The trend is your friend. ‘This is not the correct definition’, or ‘it doesn’t reconcile with the data over here’ are common frustrations and everyone has a view of what is ‘correct’. They are both consequences of trying to take absolute measurements. Focus on relative measures, systematic errors are removed and the conversation becomes much easier. ‘What can we do to improve this measure by x%’
- Don’t waste time obtaining data accuracy levels you do not need. Being able to calculate pi to 2,000+ decimal places may be extremely interesting, and impressive, but not necessary in day to day life, even for flying to the moon. Understand how accurately you can measure something vs how accurately you need to measure something to make a decision. It can really help and remember data is never perfect.
- In the absence of data, use hypotheses. Can existing data prove or disprove your hypothesis? Would this confirm your theory and is this to a level of confidence to take a decision? This is common in the world of science, a theory is developed and existing data is reviewed, often to find that the data to confirm the thinking has been there all the time!
- If you do need new data, really understand the cost vs benefit. What is the cost of gathering data with greater accuracy/reporting in a new manner, vs the benefit in terms of improved decision making and business benefit over the lifetime of the expense. In other words, do you really need to build the next CERN particle accelerator when that measurement could be completed with much simpler equipment.
Lastly a word of caution on infographics, commonly accepted thinking and easy conclusions. Whilst they can be extremely useful, they are also very persuasive in leading you to a prescribed conclusion (or set of conclusions). If you need speed, trust your provider and the decision is low impact, great. If not and you are looking for new insight, testing at a deeper level is really worthwhile.
Now, don’t get me wrong, undoubtably there is joy in finding that actionable piece of insight in a sea of numbers. For those who know me, yes I love it! However it is worth taking a step back once in a while on our use of data to drive the business forward.
We all want to be in a data driven business, but this does not mean the data is driving the business, after all… you are.
