At first glance, a CDO coming out against dashboards might surprise you. Isn’t that like a chef coming out against plates?
But what if dashboards have become an outdated way of centralizing and displaying key information? That’s the opinion of several prominent CDOs, including Cindi Howson, chief data strategy officer at business intelligence firm ThoughtSpot.
She recently talked to TechTarget about the limitations of dashboards:
Why is it time for the death of the analytics dashboard?
Cindi Howson: The biggest thing is that dashboards have not been able to keep up with the needed time to insight. If the COVID-19 pandemic showed us anything, a backwards-looking dashboard that takes three months to build is not commensurate with the pace of business. Data workers report that 92 percent of their time is being spent on low-value operational tasks, and 84 percent of workers who need these granular insights reported poor experiences with their analytics solutions.
People talk about the dashboard backlog that they have. If you look at dashboards, it takes weeks, and in some organizations months, to build. They’ll have a three-month backlog of requests to add a new filter, to add a new visualization. [During the pandemic] there were so many new questions — which employees don’t have personal protective equipment, what can we do online if we have to shut down the restaurant? Analysts created reports, but perhaps the reports didn’t quite answer the question, and then there was a back-and-forth. In [traditional dashboard] world, you’re creating data extracts, so it just adds to the dashboard backlog.
The other thing is small data sets. If you think about cloud data warehouses, whether it’s Snowflake or Databricks Delta Lake, you want to be able to get all your data and not have to pre-aggregate things. Think about when we were having shortages of toilet paper. You have to get to the granular level of detail to be able say, ‘Here’s the SKU that we need to restock.’ Look at people who were laid off and were late on their credit card bill. You don’t need the aggregate data. You want to know that this is an individual customer who is never late on their bill and be able to figure out what’s going on with them. Let the business user ask the question of the data, or let the AI ask and answer it.
Have you begun to see organizations move away from analytics dashboards, or is the demise of the dashboard still more theoretical?
Howson: It’s absolutely happening. I think about [ThoughtSpot] customers who are speaking publicly about this, like Scott Peck of PricewaterhouseCoopers who says, ‘No more dashboards,’ and Juergen Kallinger from HP who said they’re out of the dashboard business during [cloud data warehouse vendor] Snowflake’s conference. The key thing that all of them say is that they’re getting to the higher level analytics — rather than just descriptive about what’s going on, they’re getting to the diagnostic level about why — and the harder business questions.