We all deal with data on a day-to-day basis. We handle data as knowledge workers, data crunchers, data creators or consumers in our daily routines. And, I am not even talking about your work yet. You pick up your newspaper in the morning (yes, some of us still do!) and scan through the stocks. You are the consumer of the data, which was prepared by an army of data crunchers and knowledge workers. You react to a piece of information on that page by picking up your smartphone and tweeting about a stock. Now, you have turned into a data creator. How? You have given the world of data consumers a piece of information (your tweet) from which they can derive more intelligence.
Business Intelligence is a thing of the past. Companies used to look at their past performance and understand their mistakes or achievements. In today’s world of tweets and posts, even news agencies can’t keep up with the speed at which information is traveling across the world. Enterprises need to be extremely proactive rather than being reactive. So, a few years ago, the concept of Social Business Intelligence was introduced. There is abundant information available in all the petabytes and zetabytes of social feeds across all social networks. The ability to single out the right channel, to kill unnecessary noise and to make sense of the remaining social information into actionable metrics is called Social Business Intelligence. For example, a retail company monitoring specific social feeds of its consumers can understand that they actually prefer its newly introduced product in red more than blue. Using this data, it can instantaneously prepare its supply chain to ship its next batch of products in red rather than blue.
However, there are inherent challenges with gathering such social data. To start with, social data is not structured. Consider a tweet such as this – “Love the fit of #Gap’s new #slimfit #jeans. Zipper not smooth”. How do you associate a metric with such a tweet? Is it a positive tweet or a negative tweet? How do we classify this tweet to make predictive intelligence from this? To address this, we have many tools out there that perform “Sentiment Analysis” on such social data. This type of analysis assigns weights and scores to social data and makes the data more meaningful for Social Business Intelligence dashboards.
There has also been another trend in the last couple of years – an increase in adoption of social media platforms within enterprises. This can span just within employees or even with customers and partners. These tools help in keeping collaborative teams closer and well-informed. And, this is where I will introduce Social Business Intelligence 2.0. The social feeds have grown to gargantuan proportions that they have created silos of Big Data by themselves. Next, the information in such social data is not useful freestanding on its own. They need to be cross-referenced with other data stores in the enterprise such as Product Master, Customer Master, Pricing database, Support Center application etc. This is the way that coherent information is created from such chaos.
Let me give you an example of such a Social Business Intelligence 2.0 scenario. Customer X logs in a support ticket about a system failure in one of their servers. The support system instantly posts a note about that ticket into the corporate social feed under the support channel. Since a few stakeholders such as the Account Manager, Systems Engineer and on-site consultant have subscribed to this specific customer’s posts, they get notifications about the post and are aware that the customer has an issue. They can immediately call up the customer expressing their interest in addressing the issue ASAP. This increases the level of customer satisfaction as well. But, what has also happened behind the scenes is that there were three other similar issues (with the same component) raised by three different customers in the past 1 week. But, those were not trouble tickets but just social posts in support forums. The correlation is made between the various social posts and a social notification is sent out to the relevant R&D team automatically. Members of that team see that post, understand the problem and release a fix immediately addressing the issue. Next, an automatic notification goes to each of those affected customers’ technical contacts indicating the availability of this fix, which addresses their problem.
Yes, I know that sounds like a fairy tale and hardly ever happens that way in any company. But, that is not wholly true. There are quite a lot of companies that are leading the charge on such initiatives and making great strides in bringing such solutions to reality. What I want you to focus on is the underlying technology to make this happen. First, yes, you need a social platform that can primarily connect everyone together. Also, the APIs of this social platform should be open enough that it can be called from other applications or from the integration platform. Second, you will need a correlation engine or a Complex Event Processing (CEP) engine that can associate disjoint data and identify patterns in them. The third most critical thing you need in such an environment is an integration backbone or an Enterprise Service Bus (ESB). This will be the glue that connects all of the moving parts in this environment together.
An ESB within a collaborative enterprise helps in enabling Social Business Intelligence 2.0 by –
- Understand and parse structured / unstructured data from various social sources
- Integrate social data from various sources into a correlation engine or a sentiment analysis engine
- Connect up with various Master data sources and lookup databases within the enterprise. This will help in classifying social data into relevant taxonomies.
- Integrate support systems, R&D systems etc with social data streams so that automatic posts can be sent into appropriate channels based on incoming / inferred data
- Pump all metrics and social analytics data into dashboards to create real-time actionable views for relevant stakeholders
An ESB can also leverage complementary technology such as in-memory management solutions to process Big Data effectively. Social Business Intelligence 2.0 is already here. I urge you to start considering your integration backbone / ESB as an effective platform to launch it within your enterprise today.
Until next time, Ciao!
RSS Feed
Email Subscription
Does InfoStream/Pulse and CEP of webMethods work together Out-of-Box in version 9? I am asking so because CEP is on IS and InfoStream is on cloud. I hope Niravana plays an important role to achieve the same.
Rankesh – Yes, Infostreams and CEP (webMethods Business Events) will work together. Btw, Infostreams is an on-premise solution.
I have always done processing in my head to figure out sentiment analysis related to stock tweets, but have longed for a sentiment analysis engine to do this for me automatically.
I am curious how an internal R&D team might make use of the same Social Data. The above case you presented works for Support organization/CRM, but what about internal Dev/QA. Obviously they want to make the product better.
Joey, very interesting question and something that has to be discussed over a whiteboard. I don’t think I can justify that by answering in this response. However, to keep it concise, I will say that metrics drive more metrics. Some organizations implement a balanced scorecard or a similar notion to associate metrics and align top-level corporate goals to department-level objectives and all the way down to individual performance metrics. So, to your question, the Dev/QA team’s goals can be tied to metrics like number of fixes released, number of issues raised, number of calls from Support teams etc. And, those metrics are aggregated from the social feeds. So, if a developer’s performance metrics are tied directly to such metrics, it will boost his/her performance to make the product better.