Don’t Let Bad Data Spoil Your Analytics

In order to draw valid conclusions from your analytics, you need quality data

Analytics can be a powerful tool for change at your business—provided, of course, that your analytics is based on quality data. Unfortunately, in today’s data-saturated world this isn’t always the case. Here are some tips that will help ensure you’re using good data to draw sound conclusions about your business.

Get Enough Data

Relying on a data pool that is too small can lead you to draw incorrect conclusions and make unnecessary or even detrimental changes. For example, you wouldn’t want to change a franchise policy based on the results of a site audit of just two out of your ten franchise locations. This might lead you to implement changes that would work for those two locations, but harm the others.

Merge with Care

Sometimes, you may find yourself needing to merge data from two different systems in order to run analytics on it. When this is the case, you need to be very careful that the two data sets truly are compatible. One obvious problem would be if one set measured time in hours and the other measured it in quarter hours or minutes. You’d need to set a standard and then convert one of the data sets before you could run your analytics.

Consider the Source

When you’re using data to drive important business decisions, you want the data source to be as unimpeachable as possible. Numbers that come directly from your own accounting or warehousing systems should be trustworthy. However, if you’re tracking a more subjective metric like employee friendliness or retail store cleanliness, you will need to consider the source. These metrics typically get evaluated by managers during site visits or customers who fill out surveys or reviews. You need to consider the possibility that the human making the judgment may have some bias.

Consider the Collection Method

You need to be confident that the data you’ve collected really is a representative sample. For example, if you’re getting self-reported data from employees or from customers, it’s possible you’re not getting an accurate picture because not everyone is responding.

Beware of Confirmation Bias

Confirmation bias happens when data analysts unintentionally seek out information that supports what they already believe to be true. Naturally, this leads to erroneous conclusions. If you have a theory as to why your retail stores are losing money, don’t seek to confirm it by only reporting on the factors you think you’ve identified. Report on all possible contributing factors to avoid overlooking something.

Use a Quality Analytics Tool

If you’re interested in tapping the power of analytics for your business, consider MyFieldAudits. MyFieldAudits is a cloud-based tool that you can use to conduct digital site audits, run reports, and create digital dashboards to communicate real-time insights to stakeholders. For a free demo, contact us at info@MyFieldAudits.com.