<<< Part 4 - This Series
[ Part Five of a Five-Part Series ] - Final Showdown – ‘Real’ Industry Vertical Solutions
Over this series of posts, we have been looking at various aspects of data, managing those aspects and utilizing data in our favor. We have also discussed about value of data - which data adds value and which does not and why it is important to consider expenditure on data as compared to the returns in terms of business value add.
It is time all of this came full circle. Ladies and Gentlemen, let’s talk 'real' solutions now.
It is imperative that solutions quantifiably address a business problem, Else, they are not real. Which is the reason we brought data strengths together to create Industry Vertical Solutions, things like:
- Dear Ms. Customer, let’s increase your campaign effectiveness by XX%
- Dear Mr. Customer, let’s decrease your inventory cost by YY%
- And yes, let’s do that with the most resilient platforms and frameworks; and with low TTM
Everything towards solving a real business problem and truly quantifiable.
Since a lot of the programs we execute are critical to our customers’ business and operations and are under NDA, we created some PoCs and case studies to help illustrate generalization of data management concepts leading to solutions.
Case Study #1: Energy Sector
Consider a power distribution company – one of the biggest challenges is maintaining a balanced sync between energy distribution and consumption. Traditionally, the data sources in use have been power generation stations, distribution grids and charging stations. In recent times, companies have started using smart grid setups and smart meters at consumption points. But there is still a gap left to strike the sync between distribution and consumption. Unless that gap is minimized, extra distribution will always lead to unnecessary loss of power.
So, we brought in some other ‘useful’ data pieces to complete the equation – not just any data, but truly ‘useful’ non-traditional data – weather, socio-economic info and end-customer sentiment and emotion. How does this help? Good question – consider a festive season, lights everywhere and its beautiful. People light up their homes and spend time with family and friends. But how do you arrive at a near-accurate consumption estimate for a given region? Here goes:
- Weather Data – is it sunny, raining, snowing? If it is sunny, would people need extra cooling? If it is raining, given historical trends, what should the advance estimate for blown-out circuit breakers be? If it is snowing, what would the additional estimate for heating be?
- Socio-Economic Data – how many people in a given part of the geography have the spending power to be able to light up their entire house? Or heat or cool a room rather than use centralized heating/cooling?
- Sentiment/Emotion Data – how many people are posting on Facebook about their upcoming vacation to Hawaii to avoid the snow? Or sharing other social updates? What it translates to, is a lot of business intelligence hidden in data in the form of sentiments and emotions
We put together a PoC to sample and utilize data across all these non-traditional sources alongside the traditional sources and the prediction accuracy was obviously way better, no surprises there. The right data can provide an entirely different and yet-unexplored level of insight into business outcomes and the required steps to get there.
Case Study #2 – Retail Sector
Similar to the Finance PoC, we brought together both traditional and non-traditional data to be able to derive insights which were previously hidden:
- Traditional sources: Store/Point-of-Sale data, Logistics info, Inventory and Billing Records
- Non-traditional sources:
- Sensor Data – How many people stood in front of a display, for how much time and how many of those turned into actual sales (basically the conversion and attach rates). Where did customers go from there? What did they finally buy?
- Churn Info – What are the trigger points and contributing factors? How can those be avoided by translation into tangible operational steps?
- Footfall – How many people entered the store? Is there a correlation between dollar value and time/day/seasonality? How many of those people went to the coffee shop next door? Should the retail store incorporate a coffee corner inside?
- Socio-economic and Weather data
- Customer Sentiment/Emotion – what are people saying about the retail business on social media? Are they providing good reviews? How can they be motivated to tell their friends and family about their purchases? How then, can that be turned into a friends and family discount and pricing packages for even more sales? One of the most underestimated influencers on business is customer emotion. Customers’ word of mouth about products has greater influence than product advertising. If a friend recommends a product to us, we will generally prefer that product over what we see on billboards – that is the power of human emotion
Data is our middle name. Explore ‘real’ industry vertical solutions powered by data-driven intelligence with Hitachi Data Systems - ‘real’ solutions that provide value to your business by leveraging the right quantum and type of data, providing a quantifiable business outcome that by far exceeds investment.
<<< Part 4 - This Series