The Future of Business Intelligence
Overloaded Dashboards and complicated Reports belong to the past. You will soon be able to ask your ‘Decision Support Digital Assistant’ all the challenging questions and get back answers, not just data.
Imagine if you could simply ask your digital assistant any question about your business; and get answers, not just data. In the not so distant future, you will be able to engage in progressive business conversations with your ‘Decision Support Digital Assistant’.
For example, if you simply ask: ‘How is my product performing?’ the system will:
1. Understand who is asking
The system identifies the user asking the question and retrieves the context (role, experience, perspective in the company, history of interactions, history of meetings and planning etc.). User identification happens seamlessly via multiple signals, including voice, location, input from the connected building etc. This is used for setting the context and personalizing the responses — for the same question, a sales person will get different answer from an engineering manager.
2. Identify the referenced ‘product’
The system analyzes the business question with NLP algorithms. By using the knowledge it has about the company (context, products, services offered, organizational structure, activity, performance, market, competition etc.), the system derives what the ‘product’ in the original question refers to — without even naming it. It then retrieves metadata, context, insights and knowledge about the product — to be used in determining what ‘performance’ means for the particular product, what metrics and KPIs are available and what content is available in the public domain.
3. Identify ‘performance’, retrieve KPIs & data index
The system will recognizes that the question refers to ‘performance’ and load all the metadata and indexes pointing to ‘product performance’ assessment — insights, KPIs and other analytical elements.
4. Derive the perspective of the business question
By using also the history of interactions with the specific user (and similar ones), it can derive the perspective of the question — for example ‘product performance’ means different things to different people in the same company: for a sales manager, it means sales volume, revenue, leads, conversion rates etc; for a quality manager, it means overall customer satisfaction score, quality metrics; from a CEO’s point of view its all about product profitability.
The ‘Decision Support Digital Assistant’ will combine all the above, to synthesize the right business answer and initiate an engaging, personalized business conversation.
The business being answers synthesized by the Decision Support Digital Assistant, may include not only internal statistics and insights, but also external — public domain content- enriching the actual response. For example, in the question ‘how is my product performing?’ the DSDA will attempt to locate relevant news about the product, social threads, references, complaints or other public-domain content about the specific product and similar ones — including competition.
By using this holistic approach, the ‘Decision Support Digital Assistant’ may reply to the sales person asking the ‘how is my product performing’ question, with more sophisticated, responses like:
“Your product ‘A’ is doing great, with a seasonally adjusted increase of sales 5% in your territory. Be aware though that there is an increasing online criticism due to quality issues of feature ‘B’. I have also e-mailed you a recently published patent application on a similar technology”
The user could follow-up and ask for more details or certain actions — all via voice; the DSDA may also present suitable insights on the nearest connected screen to the user asking the question — upon confirmation.
Business Intelligence systems of the future will provide answers, not just data. The complexity of analyzing data will be hidden under next generation NUI experiences. Insights and data stories will be incorporated in the right format, for the specific user and timing — to support or explain the business responses provided.
Referenced patent application: US 15/357574 20180144064