Skip to content
Big data

Creating data chemistry and detecting new patterns of innovation

Dr. Andrew Fletcher  Director, Thomson Reuters Labs™ – London

Dr. Andrew Fletcher  Director, Thomson Reuters Labs™ – London

The growth in data mining – and our ability to use the end product – is creating new opportunities for growth. As a result, we are also seeing new patterns in the way innovation occurs.

Last month, Professor David Gann and Professor Yike Guo from Imperial College London presented to an audience at Thomson Reuters London South Colonnade office as a part of our TechVision series. 

Watch the full talk (39:05)

How do we harness data to innovate?

These patterns, set out in Prof David Gann’s recent paper in the Harvard Business Review (registration required for free view of the article), cover five broad trends:

  • Pattern 1. Augmenting products to generate data: For example, collecting data about interaction with a product and using that data to create a highly personalized product, like Reuters TV
  • Pattern 2. Digitizing assets: Going beyond digitization of physical media to understand how metadata can be used to make data discoverable, linkable and searchable, for example using open identifier schemes
  • Pattern 3. Combining data within and across industries: For example, after 9/11, the engineering consultancy Arup created a new division to solve the problem of assessing the evacuation readiness of tall buildings by combining the data and modeling assets previously held in different parts of the business.
  • Pattern 4. Trading data: Creating the ability for micropayments and sharing of data. For example, sharing location data to dynamically adjust the insurance premium of a car.
  • Pattern 5. Codifying a distinctive service capability: Capabilities that a company already has internally can be turned into new managed services externally. For example, Accelus Org ID KYC is a managed service for due diligence.

Enabling ‘data chemistry’

For Yike, ‘data chemistry’ is one of the most exciting opportunities. As with chemistry in the lab, combining data elements together in different ways forms new products. Not all of these will be valuable but there is benefit from experimentation to discover what is both new and valuable. The experimental approach also gives an opportunity to show the value through running a pilot, or a proof-of-concept, which is particularly important when a new data product might compete with, or cannibalize, something that already exists. 

The economics of privacy

Generating new value from data surfaces an issue of data privacy. There are legal and technology challenges, but often the biggest challenge is economic. Are there clear incentives and rewards in exchange for your data? This is easier to map for consumer services, where an individual at least implicitly understands that they might get targeted advertising in exchange for using a free app or web service. In the enterprise environment the dynamics can be very different. If a company uses a service like IBM Watson to analyse their data, are they prepared for the system as a whole to learn from their data? Data innovation in this setting is still developing, and there will likely be some examples where a business is willing to share data in exchange for benefits, and some more sensitive areas where data will not be shared. Right now the default setting is for an enterprise not to share, which underlines the importance of experimentation to prove the value to a potential customer first before a new business opportunity can go live. The new patterns of innovation will reach everywhere, but they may not be evenly distributed.


Learn more

Visit Innovation @ ThomsonReuters.com to learn more about how we are pairing technology with human expertise and how you can get involved.


About the series

TechVision is a series of interactive seminars featuring industry thought leaders on emerging technology topics and trends.

  • Facebook
  • Twitter
  • Linkedin
  • Google+
  • Email

More answers