A man is searching for his keys under a streetlight. A passerby offers to help and asks:

“Did you lose your keys here?” 

The man replies: “No, but the light is much better here.”

Are you taking ‘the light is better over here’ approach to big data?

If you are looking at your data to improve decision making, you may be looking in the wrong place.

Data analytics can be shaped by a ‘streetlight’ bias – by looking at the data you already have. In most cases there’s no shortage of data – it’s readily available in structured and unstructured formats, stored in large volumes in data lakes or warehouses. 

Can you use the data to get an accurate answer to your business question?

No, because often these investigations don’t reflect the value in your data. You will get a result based only on the data that you know about. In other words, the streetlight only invites us to narrow our search.

It can be tempting to gather a large pool of data – everything you have. Then to get busy with the analysis.

In the case of one client, that was 18-years’ worth of oil and gas drilling data. At this point, we want to know what they are trying to achieve.

This is where we diverge from many data analytics agencies, who choose to follow this process:

The ‘streetlight’ way

  1. Data – What data is available?
  2. Insight – What insight can we get from the data?
  3. Actions – What can the business do as a result of the data?
  4. Value – What value can the business get from the data?

The Merkle Aquila way

We start the other way round, to find value in the data hidden in the dark:

  1. What are you trying to achieve?
  2. How are you going to measure it? 
  3. What insight do you need?
  4. What data do we need to gather?

So, if a games design company wants to increase sales, we would ask what player behaviour can we measure? What data do we need to measure that behaviour? We work from there.

We’ve helped global brands to discover what their business needs to achieve from their data, by first taking the time to listen and question. Only once we understand the business need, will we go and find the data.

And we find it’s better to get both the data scientist and decision makers together at the beginning. We will agree on the challenge before starting to get the data. A refreshing change, we hope you agree, from the C-suite being presented with data at the end of the project. By then the attention has shifted away from the real issues, and everyone huddles under the streetlight.

If you are looking for data in one place – because it’s readily-available – remember, only looking at the easy data does not guarantee better decision making.

At Merkle Aquila, we are experts at using big data for evidence-based decision making. If you need help with creating a data strategy get in touch.

Back to News