Whitepapers

Three impossible questions about your data to get right

Three impossible questions about your data to get right

At the face of it the following three questions seem innocent enough, however, for most organizations they are contradictory:

  • Question 1: Do you capture enough data that you can accurately answer almost any key question about your company performance?
  • Question 2: Is the data clean and easily accessible to all required stakeholders?
  • Question 3: If you have the same question to two different people/teams would you get the same answer?

In the race for Big Data solutions, organizations are capturing more data than they have before and provide access to the live raw data to more people than they have before.

Increasing the amount of data you collect often effects the quality, as more nuances then come into consideration along with more levels of categorization. Double counting records then becomes more likely.

This means that as more data sources become available basic questions become increasingly hard to answer. Here are some examples of a few sectors of how different teams may approach answering a question even if data is in a clean format…

Retail: How many sales did we have last month?

The product or merchandise team are often concerned about stock levels. Therefore, they may look at the total number of sales minus and returns. Whereas marketing teams are concerned about the total demand, which would include the total units sold regardless whether they have been returned or not.

With total access to the data for all teams, it is easy for each team to focus on KPI’s relevant to them, but it can lead to confusion for more senior management when comparing reports from different departments.

Charity: How many supporters do we have?

With a multi-channel world at our fingertips, supports may take all forms. A Twitter supporter may also have been on a mailing list, an existing supporter may also donate offline or through a share site (like just giving).

The CRM manager would most likely use the number of people on the supporter database, whilst a marketing manager may want to add social supporters, but is this fair if you are double counting people?

Finance: What is our online conversion rate?

Banks, insurers, asset managers etc. All have different products, each with a different process of the journey. Take an investment company, they may allow you to; buy, sell, switch or transfer different products online.

If buy is the most common, should you use that? Perhaps they should be merged into one for an overall aggregated view?

As you go through the buying journey, you may find that an existing investor is different to a new investor and people drop off when they start having to enter details in. So where do you start the journey; once the user is signed up?

These examples all assumed that the data is in a clean enough format in the first place. In a lot of instances that is not the case. If it is not the case you may need an analyst to produce the reports.

An analyst can often be the saving grace as they will understand the difference between the different department needs and therefore provide consistent results. However, analysts can quickly become bottlenecks as demand outstrips their supply of time; at the end of the day, people need to be able to self-serve.

So, what is the answer?

Look out for the next article to find out…