Big Data is a term these days one can hear or read about practically anywhere. Hundreds of companies offer big data solutions, hundreds are trying to convince the business users to “leverage big data”.
You only have to enter “big data” in the Google search bar to get a mindboggling list of articles and entries describing all kinds of complicated technology solutions, platforms and offers.
What most of these companies do not do well relates to explaining in simple, non-technical terms what big data means to the business user. “Big data” in itself is a technical term – one which doesn’t actually describe a business opportunity – so no surprise that searching for “big data” will lead to primarily technical materials.
As one of those business users, I wanted to write a blog post to attempt to describe a range of potential business applications leveraging big data – without even using the term “big data”.
This is about asking “In what concrete way can I (or my business, or my community) benefit from new insights, and where are those insights coming from?” and not about “Big Data” or “Volume, Velocity, Variety,” or any other similar terms, which in themselves do not really mean anything.
So let’s ask that question.
In what concrete way can I benefit from new insights – and where are those insights coming from?
1. Customer Interaction Analysis
Different customers deal with companies differently. Some interact once per month, others several times and using different channels (in person; by phone; over the website; even by email).
Understanding details like:
- How often customers interact with you
- What channels they use
- What the interaction is about
- What the outcome of the interaction is (e.g.: has customer service solved a challenge?)
can lead to business insights such as:
- The cost of serving my customer (total cost of interactions)
- How those costs impact the value of the customer (total profitability of customers )
- Which channels are used by what customer – and the commonalities among customers using certain channels (leading to: how can customers be motivated to use lower cost channels)
- How outcomes impact customer behaviour (the value of a happy customer vs. an unhappy customer)
These kind of insights can allow businesses to increase sales (by better serving customers), reduce costs (by focusing customers on less costly channels) and overall better manage their business (attract more high total value customers and balance priorities to reduce time invested in pursuing and maintaining relationships with low value customers).
2. Potential Churn Analysis
In industries like financial services or telecommunications, acquiring a long-term relationship with your customers is crucial. Acquiring new customers is costly – and the value of the customer is often dependent on how long the customer stays with the company. At the same time, research shows that customer churn is very significant in both of these industries. Ernst and Young suggest, for example, that 50% of banking customers changed or considered changing banks in 2012.
Before customers change their bank or telecommunications provider, they very often comment on their level of satisfaction in social media.
Paying attention to social media can help to identify:
- Which customers are unhappy (and therefore likely to churn)
- Which unhappy customers are both vocal and socially influential (and therefore likely to influence the decisions of other customers or potential customers)
- Which potential customers (customers of a competitor) are not satisfied with the service they receive (and therefore they are open to hear from other offers)
A company which listens can therefore:
- Reduce churn by reacting to unhappy customers and
- Gain new customers by targeting unhappy customers of the competition
Both activities can lead to very significant sales and profit impact.
3. Customer Microsegmentation
The leading companies (from Amazon to Beyond the Rack to Google) are collecting a vast range of information about their customers. They use the insights from their data to:
- Offer products that are highly relevant to their customer
- Provide information that reflects the interests of the user
- Anticipate and react to upcoming life events
- Send out email campaigns with highly customized (even individualized) messages reflecting only the subject matter the addressee is interested in
As the technology for collecting structured and unstructured data, removing the irrelevant noise, and combining various data sets to create a holistic view is more widely used, customers will increasingly favor companies who understand them, offer them relevant products, and don’t waste their time with sales pitches for items and services they have no interest in.
We are in a period of transition: customers do not yet penalize companies who treat them as an “average” by taking their business elsewhere. But very soon they will, and companies who understand this will benefit from it.
4. Personalized Public Services
Public services provided by government organizations are in most jurisdictions very democratic: everybody gets the exact same service under the exact same conditions. While on one hand this is very fair, on the other hand it can be a major waste of resources. Users do have different needs – some need more of service A but less of service B, while others need only service A and not service B. Creating a model which tracks historical user needs and predicts future ones can allow government organizations to tailor their public services – to better serve each user and to overall significantly reduce costs. But without solid, broad insights to needs, the impact of the services, the way the services are delivered such customized service offerings are difficult to implement.
5. Improved Insurance Pricing
One of the most challenging aspects of providing insurance services is the actuarial activity, which models risk probabilities in order to determine the cost of providing insurance. Today those models are based on averages – in the car insurance industry, this is based on age, distance driven, type of car, gender, location, past driving record and other factors. They are not considering actual true driving behaviour – despite the fact that it is the true driving behaviour which will determine the probability of an incident. Most cars have “black boxes” installed in them which track (recent) driving behaviour.
What if we gave insurance companies access to the black box, allowing them to determine exactly the driving style of the car’s owner and, on that basis, develop a highly accurate price to insure each individual car. This higher degree of individualization rewards good drivers and better anticipates potentially risky drivers.
Device-driven data to aide insurance isn’t limited to vehicles, either. Some healthcare companies have already started offering customers discounts in exchange for using toothbrushes with sensors that tell the company how often and how vigorously the customer is taking care of his or her oral hygiene, letting them reward their low-risk customers with reduced rates.
Those five examples are just a small sampling about how data can be used to deliver significant results. Business applications like these are what make “big data” relevant to modern organizations and their decision-makers. But, as mentioned before, “big data” in itself does not mean anything – it is the concrete business case to which it is attached that makes it meaningful.
The relevant discussion to organizations isn’t “big data” – it’s “big business impact”.