by Charles Roe
At the 2011 Enterprise Data World Conference, Mr. Krish Krishnan – the President of Sixth Sense Advisors Inc. and an expert on data warehousing strategy, architecture and implementation – gave a presentation titled “Social Analytics: Measuring the Digital Age.” In this presentation Mr. Krishnan discussed the current revolution and considerable changes coming to the world of Business Intelligence (BI) due to the rise of social media over the past three to five years. A major paradigm shift is underway in the business world as social media coupled with existing BI techniques has created Social Analytics (SA), whereby “the wisdom of the crowds” is fast becoming a central focus necessary for businesses to compete in the ever-expanding virtual markets of the World Wide Web. No longer can the business world focus on traditional mass marketing techniques and ignore personalized approaches to single customers, no longer is simply having a website enough to increase brand identity, no longer are the conventional enterprise structured data assets enough for companies to create and implement successful Customer Relationship Management (CRM) and Supply Chain Management (SCM) strategies. Companies who want success in the world of BI 3.0 must have their fingers on “the pulse of the audience” through the application of a detailed SA strategy or be left behind.
What is Social Analytics?
In his June 2006 article “The Rise of Crowdsourcing,” Jeff Howe coined the portmanteau “crowdsourcing” by combining “crowd” and “outsourcing.” He used the term to explain how companies are outsourcing problems and decision-making procedures traditionally given to individual employees or committees, to the general public on the Internet. Instead of the public only being consumers of products and services, they’ve now become potential partners for the business world, participating meaningfully in the business process.
Crowdsourcing became important as central theme in the creation of Social Analytics, due to advent of three important factors that happened around the same time:
- Emergence of user-based communities: With the rise of online communities formed around common interests, users could become part of a community and give their sentiments on certain products and services. Conversations then center on specific topics, which in turn generate enormous amounts of content, which further drive the discussions within the community. The cycles repeats back and forth and from that comes the spread of similar communities, offshoots and entire networks of interrelated opinion-based communities that could essentially destroy or create a company’s reputation overnight. CRM went digital and then viral within just a few years.
- New data collection technologies: Many different technologies came about that could process enormous volumes of text, recognize patterns and align those patterns into a company’s data warehouse or BI applications. This allowed for a rapid growth in data mining, predictive analysis/forecasting and statistical optimization of the vast quantities of data being created within those user-based communities.
- “Long Tail” Retailing Strategy: In conventional business statistics, there is an approximate 80/20 percentage split, whereby 80% of sales are accounted for by 20% of the products – so large ticket items were traditionally regarded as more important. In the past, an appliance store would rather sell a washing machine than ten mixers. The Long Tail retailing strategy puts the focus into the other 80% of the probability distribution, where retailers focus on selling large quantities of small ticket items, rather than small quantities of large ticket items. Amazon, Netflix, eBay, iTunes and Audible are a few examples of online businesses who have successfully implemented the Long Tail strategy.
Those three factors combined together to make crowdsourcing and the ability for businesses to tap into “the wisdom of the crowds” essential. Then came the Social Media (SM) revolution and seemingly overnight everything changed. With the advent of Facebook, LinkedIn, Twitter, blogs and microblogs, video and photo sharing sites, social media aggregators, eOpinion sites, bookmarking and a host of others, SM has become the driving force in the shift of the traditional business paradigm. SM can make or break the reputation of a company in “a heartbeat,” so a business must know what is being said, where it is being said and how to take those comments and turn them into applicable metrics. In 2009, Domino’s Pizza had a social media nightmare when an employee-created video of them badly preparing food went viral with over 2 million hits in six hours. It has taken Domino’s Pizza more than two years and a pointed SA process to fix their reputation.
Social Analytics is the collection, processing and reporting of data of a company’s entire SM online experience, which are then put into BI applications, so that a given business can see “the pulse of the audience” concerning their business. Through the creation of a SA platform, a business can turn traditional customers into virtual partners, leverage the SM online environment to build their business, enhance reputation, drive sales and allow the business ask the questions, create the products and personalize the shopping experience for a vast quantity of customers outside their traditional realm. By using SA processes, market share becomes more than stock market quotes and revenue charts – market share becomes a personalized system of “micro targeted” brand ambassadors who feel like they are a part of the company since the company’s marketing strategy comes to them through individual, actionable idea focused only on them.
How Social Analytics Changed BI: The Enterprise Social Data Evolution
Traditional enterprise data assets that are sitting in data warehouses such transaction based systems, books of records, operational systems and others are still important – they are not going to go away with the advent of SA. That data will still continue to be turned into necessary business information through data integration processes, decision support systems, demand planning, supply chains and a host of others. But, that data must now be categorized with the unstructured data collected from SM integration.
Social Analytics has changed the entire business paradigm from one of mass marketing to one of personalization at the individual customer level. This in turn has changed BI from one that analyzes internal, structured data like sales revenues and combines it with volumes of data from a business’ external, social media experience. If done correctly this entire data transformation takes place over five general steps known as an Enterprise Social Data Evolution (ESDE), where you take data that is traditionally enterprise driven and move it up to where it becomes personalized:
- Data: Enterprise Data (ED) is collected in standard ways along with new, unstructured data (UD) from SM channels.
- Information: The ED becomes useable information through BI applications and other Data Integration processes. UD is then integrated/transformed into content and becomes useable information as well.
- Content: ED/UD are combined together, linked and categorized as content and integrated into an effective metadata system.
- Knowledge: That integration then gives a business a Knowledge Base which can then have taxonomies, ontologies and others placed on top of it to create a fully functional Enterprise Social Data system.
- Personalization: All of that information can then be harvested so that each individual customer gets the information they want, just in time, on demand and a business goes from focusing on mass marketing to personalized marketing.
Those five steps are part of the entire ESDE process where a company begins the adoption at the tactical/listen and learn level and moves up to a strategic/engage and respond level with their customers (see chart below).
A large software company (client confidential) is an apt example for this entire process. With the launch of their newer releases of software, they began to lose market share – many customers (especially those in small and medium businesses) didn’t like the entirely new interface presented with this new release and began migration to other applications based on Open Source and other competitors. The company here initiated the process of collecting and integrating SM data with their already in-place ED. They did sentiment analyses, collected enormous amounts of data through their Analytics team, did geo-spatial studies that showed various regional sentiments and preferences, then put all that data into their Enterprise BI applications and did full SA integration. They created OLAP cubes so they could have a multi-dimensional perspective of what was actually going on in the software spectrum, which in turn allowed them to pinpoint the problem: they needed retail stores in specific areas so they could solve small and mid-market customers’ problems more effectively, one customer at a time, rather than providing a single mass market on-the-phone technical experience. They now have a number of stores and after eight months of operation showed a significant drop in attrition and in-fact had higher customer satisfaction ratios, all through a successful implementation of ESDE process.
Continue to Part 2…