By adaptive - August 7th, 2012
Social media tracking is evolving quickly with some platforms offering sophisticated analysis. How do enterprises in the B2B space gain insight into how effective their social media activity actual...
Social media tracking is evolving quickly with some platforms offering sophisticated analysis. How do enterprises in the B2B space gain insight into how effective their social media activity actually is? Dave Howell reports
The big data that corporations are now having to manage means in essence gaining meaningful insight into their customer bases and commercial partners. In the B2B space social media has become a new way to make what can be highly lucrative connections that simply would not have been possible a few years ago. However, with the torrent of data that is now flowing into your business, how do you analyze this information and ultimately lever this data into potential sales leads?
The McKinsey Global Institute report commented: “Digital data is now everywhere – in every sector, in every economy, in every organisation. While this topic may have once only concerned a few data geeks, big data is now relevant for leaders across every sector, and the consumers of products and services stand to benefit from its application.
“The ability to store, aggregate and combine data and then use the results to perform deep analyses has become ever more accessible as Moore’s Law in computing. Leading companies are using data collection and analysis to conduct controlled experiments to make better management decisions.”
One of the fundamental aspects of analysing the big data in your company is this information delivers potential leads that can form the basis of marketing campaigns. For the B2B sector that is characterised by highly focused relationships with customers, the big data in your business can give your marketing the edge it needs.
A recent report from the Interactive Advertising Bureau in association with the Winterberry Group said: “The plethora of first-party data now being amassed and analyzed by both publishers and advertisers is being used to build rich audience profiles that, marketers say, can enhance advertising effectiveness by enabling improved targeting and message relevancy. Today’s dominant approach calls for the development of unique customer/prospect profiles, which are then segmented and modelled as the basis for identifying what are commonly called “lookalike audiences” for follow-up marketing across channels.”
Advanced data analysis can come in many forms. Vendors in the traditional business intelligence sector are quickly updating their tools to take advantage of the new channels of information that social media is delivering. IBM sum up these new tools as: “Data mining, predictive analytics, simulation and optimization. Both predictive analytics and simulation are methods for creating a model, which is a logical representation of a problem or a process. When data is available, predictive analytics techniques are used to create the model. When there is little-to-no-data or the data is unreliable, simulation can be used to create the model using business rules.”
Vendors that are currently offering analytical solutions for big data analysis include SAP, Kognitio, ParAccel, SAS, Cloudera (that distributes open source software based on Apache Hadoop), Google’s BigQuery and of course IBM. “To get the most out of big data, you need tools and platforms that can analyze diverse data types, and you may need tools that can handle the velocity of streaming data in real time,” advises TDWI Research.
One corporation that is levering the information flowing into its business is Ford. Speaking to ZDnet their Big data analytics leader John Ginder said:
“We recognize that the volumes of data we generate internally -- from our business operations and also from our vehicle research activities as well as the universe of data that our customers live in and that exists on the Internet -- all of those things are huge opportunities for us that will likely require some new specialized techniques or platforms to manage,” said Ginder. “Our research organization is experimenting with Hadoop and we're trying to combine all of these various data sources that we have access to. We think the sky is the limit. We recognize that we're just kind of scraping the tip of the iceberg here."
Often corporations can become paralysed with the sheer volume of information they have available to them. Erik Brynjolfsson, professor of management science at Massachusetts Institute of Technology’s Sloan School of Management, says: “It’s not just big data in the sense that we have lots of data. You can also think of it as ‘nano’ data, in the sense that we have very, very fine-grained data – an ability to measure things much more precisely than in the past. You can learn about the preferences of an individual customer and personalize your offerings for that particular customer.”
In their report into big data analysis, Forsyth Communications state: “Studies show that a majority of organizations aren’t sure how to maximize the demonstrable value. Some business leaders see only the mounting costs of storing so much data, not the value in it. The big data imperative is not to discard it but to use it to begin an analytical journey leading to better insight all round.” Says Michael Olson, founder and CEO of Cloudera: “Businesses that are getting all of these status updates (on Facebook) and user-generated messages today need to understand all this and how to digest it.”
For the B2B enterprise focusing the analytics in use is the key to delivering convertible leads. As Mark Dunleavy, Managing Director at Informatica describes: “In a recent Ovum Research study, entitled Optimising Enterprise Applications: The Data Connection, providing a single version of the truth was ranked as the foremost challenge facing IT executives today. Gartner, meanwhile, ranks a single view of the customer as second only to improving reporting and analysis as a tactical master data management goal.
“Achieving a single source of truth across applications is high on the agenda for businesses - there is an obvious need for comprehensive master data management. And this single view must encompass data from multiple business functions and multiple sources – including, more than ever before, social media data.”
The B2B sector because of its highly focused customer base is perfect to exploit the insights that big data can offer. The McKinsey Global Institute report concisely summed up the opportunities that big data can offer to all B2B corporations: “The use of big data will become a key basis of competition and growth for individual firms. From the standpoint of competitiveness and the potential capture of value, all companies need to take big data seriously. In most industries, established competitors and new entrants alike will leverage data-driven strategies to innovate, compete, and capture value from deep and up-to-real-time information.”
Dai Clegg, Big Data Evangelist, IBM UK.
Q Clearly the analysis and practical application of the information contained in Big data is the driver for all corporates. How are enterprises approaching the analysis of their data silos at the moment?
A Big data is irrelevant until you analyse it. And the first place is to start is by deriving value from existing data sources. So we do see companies trying to build a unified corporate enterprise data warehouse, but that’s really an idea whose time has past. And that is why I see the characterisation of Hadoop technologies as just a way to ingest unstructured data into a relational database. It’s a very powerful concept when that is a requirement – as it is for many use cases, but the idea that it is the only use case is flawed. There’s just too much data and the cost of cleansing, loading, managing it all into one place doesn’t make sense.
Some silos are legitimate data marts – so for example, telco companies will often offload call data records for billing, churn analysis, network performance analysis and other applications. And there are analogous use cases in every industry, which means the right architecture is a federation of stores. We have this concept emerging of what Gartner calls the Logical Data Warehouse and what IBM is delivering to customers as Smart Consolidation. But you have to acknowledge the need to apply some analytical techniques across multiple platforms – searching for patterns, exploring different sources, correlating different sources. IBM recognised this challenge – which is why it acquired Vivisimo – who lead in this space.
Q Does the big data from social media need to be approached differently to big data gathered from other areas of a corporation’s activities?
A The key difference with social data is that it is unstructured text, whereas most corporate data is structured. People often don’t think of instrumented data or sensor data as structured. But it is possible to know exactly what each bit or byte it represents and it is easily transformed into a relational structure, so it can be analysed in the warehouse. And many organisations use this approach.
Point-of-sale data from retailers is market basket analysed (to reveal combinations of products that tend to be bought together) and smart grid data and other network sensor data is analysed to optimise network behaviour and predict failures. Sometimes the economics of such an approach might favour Hadoop analysis on commodity hardware, whereas time to value might favour an existing warehouse with high productivity tooling and existing skills.
Genuinely unstructured social data is typically very low value per byte. Most tweets tell you nothing and maybe, at best, add a minute weight in the pan of your analytic decision-making. So the Hadoop approach– low cost, more amenable to textual analytics – would be strongly recommended for social data.
Q Is the main problem with big data especially from social networks the fact we don’t yet have any established metrics that corporations can use to measure and analyze their social media data against?
A Not necessarily. Where social data is most immediate and has most value is in crowd sourcing applications. So, for example, a movie studio might re-edit a trailer in the launch of a blockbuster mid-campaign, responding to the tweet-stream about the characters in the trailer; this happened with Planet of The Apes reboot in 2011. Caesar, the chimpanzee character, was given more prominence in the re-edit because he was creating the most positive sentiment.
Social media gives you access to much larger samples for what would have traditionally done with more limited focus groups or other small sample survey techniques. So even though a single tweet may be unreliable, or un-interpretable, if you have hundreds of thousands of them, you can have high confidence that they are representative.
Q Who should be analysing the big data that a corporation has pouring in? Is this an IT issue, or a marketers responsibility?
A IBM’s experience is that there is most success when the line of business embraces big data. IT is the enablers, and in many cases are the pathfinders trying to push an ROI to the business. But real progress seems to depend on the business recognising the value, either because IT enabled their vision or because they just ‘get it’. Luckily marketers of all lines of business have the most to gain and are pre-disposed to want more data from more sources. Is it Marketing’s responsibility? Not specifically, but you’re right that if they don't step up it’s going to be slow going.
Q Can Big data offer the level of personal connections that brands want to make with their customers?
A I think so. I always liken it to high quality service in a restaurant – where the staff are practically invisible (I’m talking very high quality!) but where the appear from nowhere to top up the wine or whisk away the starter plates before you even realise that it needed doing. When a brand can give that level of service to you online, there will inevitably be lots of people who will find it invasive or just creepy. I know lots of folk who don’t actually like that level of restaurant service, though I do, when it’s done right. But over time we’ll come to expect it and then rely on it.