How Does Your Analytical Horsepower Stack Up?

In our data-driven world, companies need to quickly harness, process and interpret the data they have on customers, their transactions and experiences. In order to inform business decisions with this data, two things are among the most important things to have:

1. A connected data strategy and management that is intuitive and available to those feeding insights to decision-makers.

2. Talent using that data needs multifaceted expertise to provide analytical horsepower.

You may ask, as a leader responsible for making sure decisions are data-driven, what is the right balance of investing in important technology and experienced talent to ensure you are a data-driven organization. There is no dispute that investment in both areas is very important and there is not a “one size fits all” answer, as the right mix depends on the industry and the company’s strategy. With that said, regardless of industry and strategy, your talent needs to have multifaceted expertise to have the analytical horsepower.

Analytical horsepower is more than technical data engineering and data science skills, though those skills are critical and always impressive. It is more than knowing how to write optimized code to handle volumes of data in real-time to get the algorithm to run without error for prediction (though it takes a really smart person to do that!). It is also more than running automatic imputations to fill in missing data, because by doing so, you may miss an important pattern in that data you should know about. Analytical horsepower includes having the expertise to make the right judgments on what data to use, how to use it, and how to draw insight from the results. These judgments are critical soft skills that ensure results are not misleading and appropriately guide decisions. 

While analytical horsepower includes expert technical prowess like knowing how to work with unstructured data and write the python code to theme, more importantly, it includes the expertise of knowing when to have the right balance of human lead and machine lead. Text theming can be highly machine lead, which is important, but an element of human curation could be needed to get the right theming for the right insights. That is why it is important to approach every situation by making sure you understand the business problem, decide what data is needed and choose the right method to deliver the outcome desired. For example, you can decide is it good enough to rely on machine lead coding to get general themes for a high-level understanding of recent call transcripts. While another time you may need a multiphase approach with a combination of machine lead or human curation to get more granular themes to understand specific customer pain points to take action on. Analytical horsepower includes the expertise to know what is best for the situation and be able to make the right judgments on whether to rely more or less on the machine. 

With the evolution of so much data, from the internet of things to survey data to first-party data, it’s critical to start with the problem, map out a solution and then bring meaningful insights to business questions at hand. Analytical horsepower includes having the business strategy knowledge along with the technical skills to know the rewards or consequences of making certain decisions based on the results of the analytics performed. It includes recognizing, to best answer your business problem, do you need more descriptive analytics or something purely predictive. Both types of analytics are powerful and many times a combination of both is needed. It takes analytical horsepower to know what combination of analytical methods to use when and how to bring results together.

Look for more from ENGINE on how to ensure your data strategy and analytics are fit for purpose when it comes to Customer Experience, Marketing Measurement, and Product Development throughout the year.

Written by Paula Sprowl, Senior Vice President of ENGINE Insights.