sand trickling through hands
PHOTO: Ben White

As president of artificial intelligence (AI) contract analytics company Seal Software, Jim Wagner knows his client base. First, he knows that senior executives are very interested in the data analytics process. And the second, he knows that if Seal Software had to count on those senior executives to undertake the necessary data manipulation for those processes, “we would have no business,” he said. Seal Software is addressing this paradigm in two ways: It is working to minimize the amount of configuration or manipulation required by delivering prebuilt solutions. It is also bringing the AI results directly into the applications. “For example, we have a significant investment in bringing our platform and the AI outcome directly into Microsoft Word, a solution that senior business leaders use comfortably every day,” Wagner said. 

Wagner is not alone with his observation about senior executives and their unwillingness or inability to get their hands dirty with data manipulation.

A Deloitte study shows that C-suite executives tend not to have hands on knowledge of analytics and related data manipulation. Sixty-seven percent of executives surveyed are not comfortable accessing or using data from their existing tools and resources. And while 76% of survey respondents report that their analytical maturity has increased over the past year, most are still using traditional tools such as spreadsheets (62%) and business intelligence programs (58% combined). Sixty-four percent rely solely on structured data from internal systems or resources, which means they miss the insights from unstructured sources such as social media comments, product images and customer audio files. This is unfortunate as the survey also showed that executives who incorporate unstructured data into their approach are 24% more likely to have exceeded their business goals.

Related Article: The Data-Driven Organization Is an Endangered Species

Look for the Shortcuts

These habits are likely to be firmly in place and ordering senior executives to training is not likely to result in a changed attitude around data manipulation. Instead, a big part of convincing executives to manipulate data is demystifying the process — and like Seal Software, offering as many shortcuts as possible. 

“Data analytics are mistakenly perceived as mystical, labor intensive and costly expenses,” said Gennadiy Borisov, CEO of Internet of Things (IoT) development firm Klika Tech. “While some complex problems can require significant investments, tools are readily available for aggregating and analyzing vast amounts of information fast and inexpensively using prebuilt, collaboratively developed data models.” A hesitant business leader may be more apt to consider tools starting with these prebuilt models that don't require a data scientist wizard to use, he added, noting that hundreds if not thousands of such data models are available on data science repositories such as GitHub, “where you can find collaboratively developed models that power data platforms by startups such as Domino Data Lab and Sense.” 

Related Article: Why Your Company Needs a Chief Data Science Officer

A Step-by-Step Guide

Few senior executives have the time or the acumen to dive into such a project with assistance though. Indeed, hand-holding may even be necessary. Dave Mariani, founder and chief strategy officer of AtScale, suggested that the analytics teams should work with the executives to determine which KPIs the executives use on a daily basis. “The KPIs selected should be KPIs that are used for operational benchmarking — board level KPIs are the best,” he said.

The analytics team then needs to calculate and store the chosen KPIs and they need to be updated at least daily — and closer to real time or multiple updates per day would be a stretch goal, Mariani said. 

Then the analytics team needs to choose a data visualization platform that does not require installed software and can be accessed easily with single sign-on via a phone or tablet. “The visualization needs to be simple dashboards with minimal parameterization and customization — it needs to just work “as is” with minimal manipulation,” Mariani said. Finally, he concluded, the analytics team needs to measure C-suite usage and get feedback on what needs to be improved to make their usage a “habit.”