Fools’ gold?

First you needed a website. Then you needed a social media strategy. But how many of us have actually started integrating data from social media into our overall customer database management processes? And how many of these are actually integrating ‘hard’ data – such as data that tracks customers as they undergo situational changes – say by moving house, getting married, or changing job?

Not many, it seems, according to Natalie Cowen, head of brand and communications at First Direct. “Most brands are only just beginning to get to grips with social media,” she says. “Until they’ve got through this test-and-learn phase, it’s hard to know what kind of data you’re going to get via social media, and therefore what to do with it.”

Many marketers openly admit that this is the case. Natasha Hill, supporter marketing director at Cancer Research UK, explains that there are obstacles to the effective integration of social media data, such as the inherent difficulty in matching a user name on Twitter or Facebook with a customer on the database. The charity has focused on creating a single, central database, with its social media strategy forming a part of the communications mix, but not one that provides much in the way of hard data.

What’s more, marketers need to tread carefully when it comes to data collection on social sites. “These sites aren’t about sharing data with third parties,” says Emily Subden, digital producer within the digital communications team at Oxfam. “People sign up because it’s a protected environment. There have been mass outcries on Facebook when it has tried to change privacy settings; it could end up alienating people.”

Most brands are only just beginning to get to grips with social media

Another concern for marketers is the fact that individuals on social media sites frequently project personas that do not reflect their true feelings. “Often people are effectively operating behind an avatar. I’m not sure whether you can use that for marketing,” warns Ian Cracknell, marketing director at mutual insurance company UIA.

This is of particular concern for financial institutions that need to glean data from regulated sources in order to protect customers against issues such as fraud. According to Keith Moor, director of brand and communications at Santander, social media is of little value to the bank’s data teams at present. “Social media does not represent a secure channel for hard data collection,” he says. “It’s not validated data. It’s much softer; it tends to be attitude or opinion. It can’t be incorporated into models or profiles.”

For these and other reasons, many brands haven’t yet attempted to track activity on a granular or individual level across the various social platforms such as Facebook, Twitter and LinkedIn. The vast majority are still focused on ways to boost brand awareness or engagement through such channels. Daniel Nelson, social media marketing manager at MyDeco.com, says the key benefit of a Facebook presence for the online retailer is enhanced brand engagement.

The broad information – such as gender, age, country of residence – that can be gleaned from activities on the brand’s page, he says, enables his team to tailor content in line with what is generating the most ‘likes’ or comments. “On Facebook, we can see that 25- to 34-year-old females make up most of our fans, which means we can produce content with that in mind,” he says. “You can get an idea from the interactions on your page what your fanbase is responding to, but we’re not at the point where we’re doing data mining on social channels.”

In technology terms, there are tools out there to aid data collection and management, he says, but they tend to be quite expensive and constantly involved in a game of catch-up with the social media platforms – and among themselves. “It’s something of a goldrush. There’s always a new one coming out that’s apparently the definitive one. We get a lot of cold calls, but when we ask, ‘Can it do this?’ they tend to say, ‘Not yet,’” he adds.

So, for many marketers, the extent of any data mining on social channels is currently limited to often fairly rudimentary sentiment analysis or attempts to identify online influencers. Those that claim to have managed to convert ‘soft’ data from such sites into ‘hard’ data say this is only possible by creating some kind of value exchange with the customer or prospect.

This might be by asking them for information or ‘hard’ data about themselves, such as an email address, in return for serving them better in some way – or by directing them to the company website, where processes are in place for more formal web analytics and data capture.

Facebook as a lead generator

Perhaps unsurprisingly, given its focus on the student population – a demographic notoriously active on social media, insurer Endsleigh’s social media data strategy appears to be well advanced. Richard Norton, customer marketing manager, explains that although the focus at present is to keep the brand front-of-mind, Facebook – the brand’s key social platform – does now represent a valid pipeline into the overall funnel of leads, and one which can ‘absolutely’ be matched to leads in the overarching database.

Norton says that the marketing team is currently working on a data warehousing project that aims to centralise feeds from its policy holder database, as well as feeds from Facebook and other channels, all into one pot, to be able to perform a merge and dedupe in one location. This will help to offer a degree of integration, he says, although he admits there are limitations.

“There are two stages on Facebook,” he explains. “First somebody becomes a fan, but this is of no use in data terms. But we are pulling in students via a games mechanism tied to prizes. There is a data capture element when people submit their answer. You need this to capture hard data.”

Often people are effectively operating behind an avatar. I’m not sure whether you can use that for marketing

The target audience for insurance company MoreThan represents a slightly older demographic, yet marketing director Peter Markey explains that the brand is currently working with interactive marketing services provider Acxiom in order to identify customers or prospects with a propensity to use social channels. “Acxiom has software which we can run through our prospect database,” he says.

“We are trying to track this on an individual level. It’s early days, but all things are possible.” He says that looking ahead the ideal stage would be one where it is possible to join up all the data sources for rich insight into the customer, and one where the dynamic nature of social media is played through into the business.

It’s worth noting, too, that although social media currently represents to most brands a source of ‘soft’ data, such as attitude or customer opinion, that does not mean this data is of no value to marketers. It may not be readily available in a format ‘hard’ enough to be reliably incorporated into spreadsheets, matrices and so on, but it’s still useful according to Chris Davies, commercial director at BBC.com. “Soft data is not inferior, it’s just different. Combining the two, hard and soft data, is really powerful,” he says.

Davies says he suspects that the major web analytics providers are in the process of examining services that will make the process of social media data analysis easier to integrate into customer data processes. “I suspect that the power of the two together [web analytics and social media analysis; hard and soft data] could be months rather than years away,” he says.

It will take time for brands to find out how to best integrate social media data into their current infrastructures, but certainly the data professionals at Sky have recognised that it can significantly supplement data from their existing sources, adds Danny Russell, marketing strategy director at BSkyB. “It is something we are looking at closely and with interest,” he says.

And there is certainly data there. Tom Alcott, social media analyst and managing director at The Social Network Company, points out that the universities of Manchester and Greenwich are leading the way with social network analysis, and that they have entire departments set up for it. “We are at early adopter stage,” he says. “The statistics need to be simplified and the jargon agreed. The data available for social network analysis is growing daily, and the more it grows, the more you can see patterns.”

After all, Facebook alone now has more than 600 million users globally, with 30 million in the UK. It’s a veritable data goldmine – and one that hasn’t been at valued at $65bn for nothing.