[ad_1]
Just over a decade ago, there was a lot of buzz around the concept of dashboards. It was the coolest thing to be able to slice and dice data in predefined drill paths. Businesses were starting to build dashboards for anything and everything, creating a huge surge in the demand for BI and dashboarding.
Organizations were developing dashboards with views across functions, geographies and even specific sets of audiences. Sometimes they even built two different versions of the same dashboard, as the business teams within a country or function didn’t like to look at their numbers the same way as their global or cross-functional counterparts.
A few years into this, some organizations have woken up to realize the hard truth: These dashboards that were painstakingly built are hardly being utilized by business users in organizations. They instead prefer bespoke analysis built by individuals on makeshift tools that suit their specific needs.
When we dig deeper into this, we realize that business users do not see the value in these dashboards for the following reasons: They are delivered too late, do not contain the relevant cuts of data required by the business teams, are sluggish in performance or simply are too complex.
The thing with dashboards is that they are purpose-built for something specific and can rarely handle scenarios beyond their scope without playing around with complex configurations. Also, they prove to be useful only when the users know what to ask and where to look in their dashboard for answers. This requires a lot of time to be spent in training users on how to navigate each dashboard.
In today’s world, a business user is simply left with one of the following choices to understand their business:
• Roll up their sleeves and perform an analysis on their own. This would typically involve working with IT teams to gather the required data for analysis to put something together in a spreadsheet.
• Raise a request with the in-house analytics organizations or a business analyst to perform ad-hoc analysis. This usually takes from days to weeks, depending on the complexity of the business question.
It would be ideal to pair a human analyst with every business user to help them derive insights from data. It is, however, not a scalable model. Organizations must strive to provide the next best alternative to business users — an AI analyst who can:
1. Answer their ad hoc questions in the most natural way possible
2. Understand what keeps them awake at night and proactively nudge them on the areas they need to be aware of in their business
3. Predict what is about to happen so that they can take preemptive action
4. Help them get to the whys of their KPIs easily
An AI analyst needs to go above and beyond and look at any KPI in a holistic manner and provide the following insights to the business user:
Descriptive
• Has the KPI grown or declined with respect to the base period?
• Is the rate of growth or decline faster or slower than the market?
Diagnostic
• Which areas of business are contributing to the growth or decline?
• Which business levers are driving the change? What is their impact on the KPI?
Predictive
• How is the KPI projected to trend in the next few periods?
• Would a decrease in price result in an increase in revenue?
Prescriptive
• Which areas of the business should the user focus on to improve their KPIs?
Providing answers to these common questions that business users grapple with on an everyday basis in an intelligent and automated way with an AI analyst will eliminate the time wasted in deriving insights from data, in turn leading to faster data-driven decisions.
[ad_2]
Source link
