Enterprises have been relying on business intelligence tools like Cognos, Tableau and Power BI for decades now. These types of software revolutionized how businesses generated reports, analyzed them and made crucial decisions (both daily and long term) aimed at improving their market performance. They especially changed the speed at which businesses could generate information and reach these decisions.
But in 2020, these tools are beginning to show their age.
BI Tools Are Not Perfect
The primary function of most BI tools is to get business information in front of those who should act upon it. In practice, this means generating reports or providing interactive dashboards that display crucial business data. The problem is that these tools do little to perform analysis and pinpoint specific problems. This weakness has necessitated the creation of separate jobs that revolve around sifting through reports and managing dashboards — generalized BI analysts as well as analysts whose entire careers are specific to certain software platforms.
Many enterprises are then forced to choose between hiring additional employees to actually make decisions and perform necessary tasks or leaving the original employee with what amounts to two full-time jobs. Operators who need to be making daily decisions in the field are swamped with hundreds of reports that take time away from the operations they are trained to do to benefit the business. Some enterprises are spending unnecessary capital on analyst salaries.
The quality of BI dashboards and reports are also completely reliant on the quality of those who design them. In most organizations, the employees who configure BI dashboards are part of BI or analytics teams — this makes them experts at business analysis, but they rarely know the end users’ day-to-day struggles. This is one reason reports are often so prolific and dashboards come with endless configurations: The employees crafting them are simply providing as much information in as much detail as possible, with little thought of how that information will be used in the field.
A prime example of this problem is in retail, where merchants and buyers use reports to analyze KPI over time, compare sales year to year or month to month and identify anomalies like a sudden drop in the sale of bedroom slippers or children’s bicycles. They are often faced with massive sheaves of reports or endlessly complex dashboards that they have to spend hours poring over before they can come to the conclusions needed to make beneficial moves on the sales floor or with their suppliers.
Operators can, of course, provide feedback to their BI or analytics team to improve the usefulness of their reports and dashboards, but unfortunately, most businesses have no simple mechanism for this. Feedback is rarely taken on board quickly enough to enable valuable changes, and differences between operators mean that the same tweaks are unlikely to benefit every end user.
The bottom line is that most BI systems provide too many reports or too many dashboards for the end user to realistically incorporate into their business decisions. This leads to operators drawing contradictory conclusions, relying on incomplete assessments, and potentially making critical business errors that leave profits on the table.
How Can AI Help?
We’ve written before about how AI is best used not to replace humans but to assist them and make them more efficient.
Humans are best at:
• Making key business decisions.
• Innovating on designs and plans.
• Creating new and improved strategies.
AI is best at:
• Making day-to-day business adjustments.
• Structuring and organizing data.
• Prioritizing human actions.
By taking advantage of human and AI strengths, we can create BI tools that allow humans to do what they’re best at while AI does the rest. Business leaders can use the thinking above to improve their AI applications and imagine a future where their employees can work to the best of their abilities.
Some of the best applications of AI are in:
• Automating back-office operations that tend to be outsourced overseas or that take valuable time from more skilled employees.
• Automating processes like product data onboarding that humans simply cannot scale.
• Improving the quality of information available to operators, reducing their time spent dealing with BI tools.
This last bullet point is the most innovative. Rather than employees handling the repetitive, time-consuming task of sifting through reports and configuring dashboards, AI can churn through terabytes of data at incredible speeds: transactions, product information, procurement, vendor data, inventory, and store specific information can all be organized and analyzed by AI quickly and efficiently.
Best of all, AI can break this information down into bite-sized insights that are actually useful for operators making buying decisions and sales strategies. By strategically leveraging AI, operators can act quickly on anomalies that AI detects, and bring their organizations closer to target business metrics without wasting time and money.
If business leaders can reconfigure their thinking about BI and AI tools so that BI, AI and human employees can do what they are best at, enterprises can become far more efficient and innovate at much faster speeds than ever before.
This article first appeared on Forbes.