Analytics and Context - Big Data Comes to Communications

5 Mar 2014

At the Lync Conference 2014 last week, I participated in an interview on the topic of "In Context Analytics" with Kieran McCorry of HP. Kieran is a Strategist in New Service Development for the Enterprise Services CTO Office at Hewlett Packard.

Kieran described a new product and service that HP is offering that brings "Big Data" techniques directly to the user's desktop or device. The interview as well as an HP video clip and a PDF highlight that the HP software integrates with the user's instant messaging (IM) system, such as Microsoft Lync, to track the IM conversations.

As the user types along, the In Context Analytics software is tracking for key words and phrases that indicate a potential need for or value from additional information. That information is then immediately popped up in an adjacent window for quick reference during the chat. As more detail is provided during the chat, the window can evolve in parallel, as shown in the video clip.

This information search can reference public information, such as web pages, as well as information internal to the enterprise and even the user's own information such as e-mail messages and documents that the user has on file. The In Context Analytics continue to learn by monitoring all that the user types, and updates the reference data by tracking new information.

There may be many benefits from this type of tool and the value will vary depending on the user's business role. The benefits will likely fall into two broad categories:

  • Time savings when the information can be provided without the user spending seconds or minutes searching through folders, logs, calendars and e-mail records. If the user's role has a high transaction volume, these small savings can really add up when multiplied by a large number of events. For example, a text chat help desk engineer who has been processing 40 text sessions a day at around 10 minutes per call could lower costs by 10% or increase capacity by 11% if In Context Analytics could shave just one minute off of each call, perhaps by finding and popping up relevant sections of product manuals and knowledge base (KB) articles.
     
  • Improved business outcomes when the information helps inform a conversation, a collaboration, or a negation. For example, if In Context Analytics assisted an online sales person to raise their close rate, say, from 20% to 25% of customer interactions, this small change in close rate (+5%) could result in a large revenue multiplier (+25%).

Of course, there's been a lot of buzz about big data applications in highly complex or analytic applications, such as health care diagnostics. Yet those applications still seem distant from the information worker's desktop. In Context Analytics seems to be one very practical example of how big data can be made 'small' for user productivity and convenience.

Of course, this is not the first example we have seen of near-real-time context-based tools. Many call centers have used various tools to monitor customer interactions in real-time to aid call center agents with first-call resolution of a customer problem or with relevant up-sell recommendations based on analysis of the customer's history. In some cases, software can even listen to the customer's words and tone to aid the agent in delivering customer satisfaction. We have also seen, in the Avaya Flare product, some analytics invoked at the outset of a call to pull up contextually relevant documents or messages, although that did not include the near-real-time information retrieval provided by the HP In Context Analytics engine.

It will be interesting to watch the adoption of this HP capability to see how the value is realized, and to watch the advancement to see how it optimizes business processes.

  • Will it be possible to measure the benefits, as suggested above?
     
  • Might it lead to partial automation of some transactions or business interactions so as to minimize the user's time and labor in response to IM requests? Note that we saw a version of that in the Lync Conference keynote that illustrated the start of a medical interaction as an IM session with an automated "bot" after which the physician came into the session to convert the symptom identification into a diagnostic decision and care recommendation.
     
  • Might it be possible that the analytic software could heuristically and automatically convert repetitive IM dialogs into a "bot" type of session based on a cumulative knowledge base so that the value of experts could be magnified and shared much more effectively?
     
  • Is this extension of communications to include software assistance just another example of what Gurdeep Singh Pall meant by "Universal Communications" as the new definition for what we have known as UC?

As with any new technology there are many intriguing questions. Thanks to HP for putting this tool at our disposal so we can continue the UC journey of delivering "communications integrated to optimize business processes."


This paper is sponsored by HP.

Unified Communications Strategies Logo Sm

Also on UCStrategies.com on this topic:

 

Comments

There are currently no comments on this article.

You must be a registered user to make comments

Related Vendors