Your Data is Worthless*

[*Forgiving the use of the singular...]

Your data is worthless. Why? Because data has no intrinsic value. In fact, I'd go a step further, data is a cost: a cost to generate and a cost to propagate. Very often, a cost that your business should avoid.

It is a commonly accepted piece of wisdom that there is latent value in data; value that exists, but needs to be unlocked. This has resulted in tech companies scrambling to store as much data as possible. In the e-commerce world for example, this could be web referrer, basket, page history, location, device, prices, products, mouse movements, etc. Much like the belief that there is a beautiful statue in every block of stone, the act of unlocking this value requires a sculptor’s skill. What we have instead done with data is: stockpile stones.

The first problem with this course of action is that it overlooks the cost of collecting, managing and storing the data. The second is that it has led to a relatively tenuous view of what unlocking value means. Unlocking value necessitates the generation of value from a given activity. Unless you’re in the business of selling insights, then your insights in themselves are not a value-driving initiative: they are also a cost.

Collection → Insight → Implementation

Let’s simplify a standard BI / Data analysis process into three stages:

1. Collection - Collecting, storing, managing and making that data accessible (Cost, requiring an Engineering or Ops Team)
2. Insight - Doing the analysis and deriving the insights (Cost, requiring a BI or Data team)
3. Implementation - Taking action off the back of the insights (Potentially revenue-generating, requiring action from the rest of the company)

Without some idea of what insights you intend to generate, it can be difficult to know whether you’re managing your data effectively. Often this falls down in the collection stage. After all, how do you know that you will be collecting the right data points or whether it’s fit-for-purpose once collected if you do not know what analysis you want to conduct? I have seen this occur far too many times across a business, often in CRM systems and product usage tracking, where data is collected ‘just-in-case’. In CRM systems this can often be more costly, as it requires expensive sales reps to fill in additional data that is never used. This results in the following: 

1. The data is collected without proper direction. 
2. When it is eventually analysed, it is discovered that it is incomplete or not accurate enough to draw any conclusions from.
3. The data is scrapped and a new pipeline or collection methodology is employed based upon what is actually required.

A lot of the industry is focused upon insights. It is easy to forget: insights are only valuable once they are are used. Knowing that customers onboarded in a rush are most likely to churn is not only interesting, it’s useful. It’s also worthless if no one is going to action it. If it’s not actioned, then the analysis should never have been conducted. It was a waste of time and money (and in this case, a missed opportunity).

I have seen this confusion across business decision-making, where we misconstrue a cost as a revenue-driving activity. Implementation is where value is created, not analysis. If your data never reaches that stage, then you have an inefficient process. Do not collect nor analyse data unless you implement your insights.

A few simple suggestions for unlocking value

In order to unlock the value of your data, here are a few suggestions to follow:

1. Only collect the data that you need - You should not collect data without a clear use case. The use case helps identify and maintain the quality of the data. It helps to ensure that it is structured correctly and that there is a plan for monetising it. With GDPR in full-swing, this is now a legal necessity, as opposed to just an operational best-practice.
2. Give the data to people who can analyse it - Ensure that you have people capable of analysing your data. This doesn’t need to be a data scientist, it can be an analyst, consultant, financier or a data-savvy commercial rep. Make sure you put it in their hands as quickly as possible.
3. Have a plan to action your insights - Ensure that insights are created, shared and that there is a follow-up plan for its utilisation. Remember that the implementation is where the value is: make sure that there is a solid process to implement any idea you have.
4. Ensure that the people that create the insights are involved in the execution -Those that have derived the insight and completed the analysis should ensure that their work is properly understood. Involve them in the execution process.
5. Action your plans and monitor their success - Make sure you (1) follow-through on your plans and (2) have success metrics to evaluate how effective they were. Remind yourself that unless this is being done, the business has been losing money unnecessarily.

Key takeaways

• Do not store data for the sake of it: ensure that there is a clear use case.
• Don’t stop at insights: ensure that you have a game plan to action these insights once they are uncovered.