By Clive Gold, CTO Marketing, EMC Australia and New Zealand
The idea of Socialytic Apps is really interesting. Last week I mentioned the IDC briefing I attended and covered the first half of the discussion here.
Now the more thought provoking part and IDC’s other new term, the Socialytic app. To start they said that the socialytic app has the following attributes:
- Aggregate social data from public and private social data sources
- Real-time search and monitoring
- Social metrics dashboard
- Natural language processing (NLP) with linguistic analysis capabilities
- Visualization and simulation
- Social trending
- Human search
(There has been some reporting but CIO seems to have the best summary I found, here.)
The argument put forward by IDC is that companies are using social media externally, to increase brand awareness and manage and enhance relationships with customers and prospects. Internally they are using the tools to improve collaboration and productivity, with the intention of achieving the three essentials of revenue growth, cost reduction and risk mitigation. At the same time IDC’s research shows that CIO/CTOs rate business intelligence as being high on their list of priorities, with the intent of improving decision making, driving revenue growth and managing risk.
So it would seem these two, social media and BI, are inexorably related!
What’s the problem then? Well traditional BI is too structured and too slow for this work. Too structured as when data warehouses are set up they have a schema which pre-defines the questions that can be asked. Too slow, hence data has to be summarised to enable queries to be run and once again limiting the scope of the system. All that has now changed, with the new analytical databases that are now available. These systems are built to be highly parallel allowing the processing billions of structured rows in realistic-time as well as use tools, such as Hadoop, to process semi-structured datasets. But the real kicker is that these tools leverage standardised hardware, which drops one or two zero’s from the price tag, when compared to proprietary systems.
The key to me is the integration or mashup of this social media data with your internal systems. A cute example of this was shown during Pat Gelsinger’s keynote at EMC World. Where they looked for competitors in the audience, used public data to build a resume, used internal systems to get references, and then fired off the hiring system.
A new world is really on us, so do you use any social media integration into any of your core systems?