Aggregates are misleading and averages tend to lie

Aggregates are misleading and averages tend to lie Business illustrations by Storyset

Summary

Data driven businesses collect massive amounts of data today, which continues to increase exponentially. The question that often surrounds most decision makers is whether their team is actually consuming their own data in a meaningful way – to get rich insights or timely warnings.

Consider an example: a data driven company with an online presence measuring typical KPIs like revenue, views, CPC and order conversions across multiple dimensions:

  • Thousands of products/offerings
  • Hundreds of geographic locations
  • Dozens of channels
  • Dozens of ad-campaigns
  • Dozens of payment processors
  • And the list goes- on!

Here is the good and the bad about this data.

The GOOD:

There are millions and millions of combinations of dimensional values that can affect and influence your KPIs

The BAD:

There are millions and millions of combinations of dimensional values that can affect and influence your KPIs

The BAD becomes GOOD when we can understand, interpret, and leverage the insights hidden in the data. At this scale, the current process of KPI monitoring is highly time consuming, mundane, and an inefficient use of analyst’s time. As a result companies often revert to looking at aggregates and believing the averages. They build fancy visualization dashboards on existing analytics tools and establish some static protocol for retrieving information.

Any data driven team will quickly realize that visualizing the KPIs at the aggregate level and believing the averages can often leave you blind sighted to anomalies and opportunities in your data, which are time sensitive and highly dynamic. For example, it is FAR MORE powerful to know that 45% of the measured increase in conversion rate is being driven by the house product category, women iOS users, across California, through Facebook marketing campaign, and in the age group of 40-50 years THAN just knowing that there was a 10% increase in conversion rate across US. At the same time, the knowledge of a 10% increase in conversion rate across the US also hides that the Facebook campaigns amongst users from Texas on the same product line is not driving any conversions.

We at BoostKPI firmly believe that making data driven decisions based on aggregates and averages alone can often take companies on a trajectory that slows their growth. Contact us at BoostKPI to see how we automate this process of true business insights- so your data driven decisions are beyond the limits of aggregates and averages.