On Monday of this week I introduced the topic of Data Value at EMC World for the first time. I framed the conversation in the context of data center evolution. Over the years we’ve seen a proliferation of storage services in the storage industry, from RAID to caching, snap copy to remote mirroring, encryption to deduplication, and so on down the line. I proposed that by 2020 we’re likely to see a new storage service emerge: data valuation.
This conclusion was reached as part of the joint research EMC has been conducting with Dr. Jim Short of the San Diego Supercomputer Center. Starting in January of 2015 we began the Architecting for Value initiative. The research, as highlighted by the graphic below, has two goals: 1) Discovering and documenting emerging valuation business processes, and 2) understanding and documenting the impact of these business processes on IT architectures.
The topic of Data Value was new to most in the audience, so I reviewed some of the valuation examples Dr. Short has uncovered in the industry, e.g.
- A Caesar’s palace database being valued at 1 billion dollars
- LinkedIn’s acquisition of Lynda.com’s video assets (valued somewhere over 1 billion dollars)
- Tesco’s sale of dunnhumby (900 million dollars)
In the same way that traditional storage services (e.g. volume management, replication) can be implemented in various ways and in various locations in the I/O ecosystem, I introduced five new approaches for implementing data valuation services in the context of a data lake architecture (the same approaches can be used in non-data lake architectures as well). These 5 methods are highlighted below.
Each approach can be briefly described as follows:
- Content processing: direct analysis of the content in a data lake
- Data protection ecosystem: indirect analysis leveraging data copies
- Content Ingest: valuation during initial ingest
- Application Agility: linking data value with speed of application development
- Content workflow: linking data value with business activities related to the data
One attendee asked me if there were other EMC World session related to Data Valuation. “Unfortunately”, I said, “not this year”!
I am certain, however, that future incarnations of EMC World will conduct sessions on this topic.
Over the next few posts I will dive down into these five approaches. Each approach inserts itself into the data ecosystem in a slightly different way for slightly different purposes.
If there are readers that are interested in participating in a workshop on this topic, please reach out to me via my Twitter handle (see below) or leave a comment for me on this blog post.
Steve
EMC Fellow
