What Exactly Do We Mean by Digital Disruption?

…And What Are the Data Management Implications?

Digital DisruptionEverybody is talking about “digital disruption,” The problem is that no two people mean the same thing by the term. But that being said, it would be a bad idea to just tune out the discussion because the term is being over-used or over-hyped.

It is a fact that businesses in every industry worldwide are being disrupted on a daily basis. Often by competitors that are from outside their industry, and as a result, off their competitive radar screens. What we mean by digital disruption is best described by a few examples:

  1. New all-digital banks and insurance companies are undercutting more established bricks & mortar financial services firms.
  2. Google, Tesla, and Uber are all investing in driver-less cars that may disrupt the incumbent automobile manufacturers.
  3. New sports equipment competitors are emerging who embed sensors in their products to provide immediate feedback to their users on their performance.
  4. Automobile manufacturers are selling transportation as a subscription service that includes the car, all maintenance, and car insurance.
  5. Healthcare companies are combining clinical data with genomic data to deliver highly personalized service and better patient healthcare outcomes.

What do all of these have in common?

The new competitors are implementing new business models that leverage data and analytics to deliver a better outcome or service for the end customer.

McKinsey recently came out with an update on analytics titles The Age of Analytics: Competing in a Data Driven World. This is a very lengthy report but well worth the read. For the purposes of this blog, I would like to go into the six disruptive models McKinsey identified and talk about examples of the models and the data management implications of each.

The Six Disruptive Models

McKinsey identified six models for leveraging digital disruption for competitive advantage. Let’s take each of these models, break them down, and describe the data management implications.

1) Business Models Enabled by Orthogonal Data

  • What: By orthogonal data they mean data that is new to the space, often sourced from outside the organization. By combining internal data with new, orthogonal data it is possible to create new business models that can change how an organization competes in the marketplace.
  • Example: Telematics that provide insight into driving behavior. Some insurance companies are offering discounts to customers who allow their actual driving behavior to be monitored.  We are talking about a “market size of one” here.
  • Data Management Implications: To do this, organizations will need to look beyond their application data silos to other sources of data both internal and external to the organization. They key to success will be on rapidly acquiring this data, ensuring that it is trustworthy, and then matching it up with the correct internal data (such as an insurance policy holder). If the data is coming from external sources, you will have less control over it, so careful monitoring and governance will be required – more or less depending on the impact to the business if the data turns out to be wrong, incomplete, or dated.

2) Hyperscale, Real Time Matching

  • What: This is about creating a digital marketplace that matches up the participants in real time so that they can conduct a transaction.
  • Example: Ride-share services who are matching people looking for a ride with drivers who are available to give them a ride. Or, matching any type of buyers and sellers.
  • Data Management Implications: The most effective marketplaces are usually those that are the biggest, with strong competition and many choices available. The data size for a given transaction may not be very big but the overall transaction volume is likely to be enormous. Some of the data will be just informational, so data quality may not be critical, while other data will be critical such as the identities of the parties and the matching criteria used. It will be important to support real time data management and to manage the data most where it is most critical to business success.

3) Radical Personalization

  • What: This is about delivering a highly personalized experience regardless of the type of product or service being used. This is an area where we see a large number of organizations investing to separate themselves from their competitors.
  • Example: Don’t you dream of going to a hotel reception desk and finding that they know and understand all of your preferences, don’t collect data they already have about you, and proactively make suggestions and take actions based on their previous experience with you? For a really funny but bad example of this see:  The Adobe Marketing Cloud video.
  • Data Management Implications: For this type of use case you are most likely using a great deal of data from widely disparate data sources. They key to success here is to be able to relate all of this data quickly to the right person, policy holder, patient, etc. so that it can be used in human real time to provide a better customer experience. This could involve knowing the customer’s needs, preferences, history, and past experiences with this organization, for example.

4) Massive Data Integration Capabilities

  • What: This is about capturing and combining very large data sets from a variety of sources. This most often involves matching that data with data from a known, internal data source so that you can derive a new insight about a customer, production facility, jet engine or other business entity.
  • Example: In the financial services industry this is being used to better understand customer’s needs and preferences so as to deliver to them more appropriate upsell and cross-sell offerings. This can result in higher revenues and higher customer satisfaction if they recommendations are perceived as genuinely helpful.
  • Data Management Implications: We are most likely talking about big data technologies to store and manage very large data sets. What is more challenging is the need to combine the external data with internal data that is stored in multiple silos to deliver an actionable and useful insight. Successful initiatives will require data management for traditional, structured data combined with data management for big data.

5) Data-Driven Discovery

  • What: The use case is about using analytics, in all forms, to gather new insights and to ask new questions that were never technically and/or economically feasible to ask previously. By combining many sources of data, data scientist will have the environment they need to experiment, fail fast, and iterate until they find a new and useful insight that can be used to deliver better outcomes and services. Increasingly, we are seeing customers using machine learning to deliver the results.
  • Example: Training systems to deliver algorithms that see patterns in user behavior that would suggest a buying preference in the future.
  • Data Management Implications: This typically involves many large data sets to create a useful algorithm. “Good enough” data may be sufficient in the early stages to test a hypothesis. But, when it is time to put this into production, trustworthy data for critical data items, will be a requirement.

6) Enhanced Decision Making

  • What: Managers routinely tell me that they have very pretty dashboards, but just do not trust the data enough for important decisions. To make “bet your career” management or operational decisions, the data has to be right.
  • Example: What if you could detect and eliminate 3% of your fraudulent activity or make your production environment 3% more efficient? In the case of an oil refiner, for example, something this small can be worth tens of millions of dollars.
  • Data Management Implications: This is about capturing and matching a wide variety of data but there is often also a real time element to it. First, important decisions must be made on data that management can completely trust. Second, they data may have to be delivered in real time or near real time to provide the business outcome desired. Getting to the really important data may involve applying real time filtering and rules to a data stream.

Wrap-up:  From a Data Management Perspective

The bottom line is that your organization is probably in the midst of digital transformation whether they use that specific term or not. What you will see is new business initiatives and business models being implemented that leverage one or more of the disruptive models described above.

As an IT professional, you are dealing with changing data types, changing systems, and now changing business models. What you need to be able to be adaptable in this type of world is a data management platform that works across all of these changing landscapes to give you a point of consistency to manage all of this change from.  You should not have to purchase the entire data management platform all at once, but it is critical that as your needs grow that you be able to add data management capability to fuel your digital transformation. A standard, flexible data management platform is essential to connecting a rapidly-changing organization and enabling IT to partner with the business side to drive a strategy that will truly differentiate the organization.

Having a solid plan and architecture for data management can make the difference between being a disruptor and a disruptee.

Comments

  • By orthogonal data they mean data that is new to the space, often sourced from outside the organization. By combining internal data with new, orthogonal data it is possible to create new business models that can change how an organization competes in the marketplace.