By adaptive - July 18th, 2012
Creating a social media channel strategy and campaign for a business is a relatively quick and simple task, however managing the data that results from the campaign can be a real problem. Soci...
Social media data is unstructured and can be spread across many different channels, which makes it difficult to calculate a return on investment and to get any meaningful feedback from the strategy. And it’s a problem that’s affecting many businesses. A recent survey by Aprimo found more than half of the businesses surveyed were not currently leveraging big data, citing simply not knowing where to start. The survey also found that of the businesses that did use data-driven insights into their marketing decision-making, 38% found it difficult to gather and integrate the data.
Andy Jones, Head of Solutions at Software AG, comments “Everybody is talking about big data these days. For many it has become a catch all phrase but it is how that data is accumulated and what then becomes of it, which is the real issue. Many companies don’t know what data they are accumulating, how they manage it and the processes involved in gleaning information from that data.”
Although there are now tools on the market like Datasift and Gnip and a host of tools and services that link in to the data provided by DataSift and Gnips making sense of the data that is flowing into a corporation, and then commercially acting upon the insights that data delivers is proving to be difficult.
Jones agrees “Importantly, cost-effective solutions are available which flexibly integrate with existing infrastructures causing minimum disruption to the business. Solutions like this will enable businesses to benefit from their data-gathering by managing and automating the organisation’s resulting responses and processes. These kinds of solutions ensure businesses are able to benefit from all that big data has to offer without the associated process headaches.”
Sorting the data wheat from the data chaff
One of the biggest problems for many businesses with social media is often the inability to spot a brand name amongst the flood of data – Twitter refers to the stream of tweets it sends to the likes of Datasift and Gnips as the Firehose for good reason – and for some businesses the only way to solve the problems is to build something in-house.
BT created their own solution to the social media data headache because they found that the tools that were available were unable to pinpoint their brand. Warren Buckley, Managing Director, Customer Services at BT explains: “We use a tool called Debatescape, which we developed internally. It tracks across social media channels with a very significant set of search strings, and it then filters out things that aren’t to do with BT. For example there is a musician in the states called BT. It takes those results and it delivers them through a workflow package to my dedicated advisors who will then respond.”
Another related problem is spelling. While you’re very particular about the spelling of your brand name, your customers aren’t. For example insurance business More Than, likes to spell the brand name as “MoreTh>n” but do a search for MoreTh>n on Twitter or Facebook and you won’t see many results as the customers are more flexible in their spelling. While a search for “more than” throws up many false positives as it’s a common figure of speech.
Additionally like many other businesses More Than also has multiple social media personas including morethanfrreeman and morethan. On Twitter morethan and morethanpetprofiles on Facebook, which means that not only do you end up searching for a brand name you also need to search for the multiple personas as well.
An additional problem is the number and diversity of social media platforms available, from big platforms such as Twitter, Facebook and LinkedIn through to vertical forums such as MoneySavingexpert in the personal finance world and AVForums in the consumer audio visual world, and on through to places that you would never expect. BT’s Buckley explains: “Debatescape searches across all of the public forum environments, from Twitter, and external forums like Moneysavingexpert.com , Thinkbroadband.com, YouTube, LinkedIn, through to Google Earth. Actually Google Earth is an interesting one as we have found that people were looking on Google Earth and for some strange reason leaving us complaints on the comment area for the area covered by BT centre and BT Tower.”
Answering in social media time
Another key problem with social media data is response time. Over the last twenty years the acceptable time for a response from a business has become more and more critical. With the postal age, the average expected response time to reply was weeks, with the internet age acceptable response times reduced to days, with social media response time is hours, possibly even minutes. It’s therefore important that you make sure that the rest of the business understands the urgency required in answering on social media platforms. And you also need to understand that the urgency may change depending on the type of query you receive.
While answering questions in social media timescales is necessary, the quick response also sets a precedent for the rest of the business eg if part of your business answers a customers’ query in 30 minutes then the rest of the business has to follow suit.
At More Than teams were created to respond in social media timeframe, which varied depending on the channel from hours to within 24 hours. The social media response teams consisted of a few people for each sector within the business – cars, pet, home, travel insurance - and were based on teams already created to deal with complaints under the financial services act, which requires businesses to respond within a certain time to customer complaints or the business can incur a fine.
Using big data constructively
The last element to social media is how you can use the data commercially within the business. Many corporations keep their social media feedback within the e-commerce or the marketing departments and little filters up the company hierarchy.
To make it easier to pass data up the hierarchy, many businesses use dashboards and net promoter scores (NPS) and by allying this with social media you can get real insight into the business. Richard Anderson, Vice President for Client Services International for BazaarVoice explains, “A client had a low-value product, that on paper looked a very successful product in terms of volume. However, it was only when the company used a NPS as a metric that they were able to see, that not only was the product getting negative feedback, it was also affecting their reputation as retailers of quality goods.“
Anderson added, This example illustrates one of the fundamental is“What they found from the standard metric, i.e. the number of units shipped, was it was doing really well, but when it came to a customer feedback metric - which they weren't getting from customer services team because the item was too low in value to trigger a customer services call - the reviews were so bad that it was putting people off the company.” sues that corporations are now wrestling with as their big data silos continue their expotential growth.
Although an NPS is useful, it won’t tell you everything. At More Than social media verbatims as well as numbers are used as a way of getting insight into the business. The verbatims are collected and categorised from the Twitter and Facebook social media and from additional feedback tools, and these are used at all levels in the business to show how the business is perceived by prospective and current customers.
One of the most valuable social media insight tools was a feedback tab that appears on practically every page of the website. The feedback tab allows users to respond in an unstructured way and works almost as a channel of last resort. It picks up when users are unable to get through via email, phone or forms and can highlight problems with the website or other general problems that fall through the standard channels.
The use of social media in business is still a comparatively young discipline, and there are no right or wrong ways to analyse social media data. However, that doesn’t mean you shouldn’t try, and even if you only manage to analyse part of your social media data, it’s still better than nothing. Managing big data is a major issue for corporations, but with new tools in development, businesses should soon have the analytical platforms they need to lever the data that is available to them right now.