Big data: Embracing the elephant in the room
The sheer mass of information available about consumers on the web is enough to scare marketers into ignoring it. But for those brave enough to confront and harness big data, there are large rewards.
Big data can be a big problem for brands. Information from large, complex data sets - weblogs, social media, smartphone analytics and even medical records - is difficult for brands to capture, manage and process within traditional database systems.
But big data is also the next frontier for innovation, competitive advantage and productivity, according to a research report from the McKinsey Global Institute. It warns that too many companies are missing the opportunities because of a lack of data management expertise.
GlaxoSmithKline’s (GSK) customer relationship management consultant James Parker agrees that firms need to invest in operations to process big data immediately. “There are even larger data sets today [than 10 years ago] and although consumer behaviour is best monitored by the marketing department, the benefits are to the rest of the business,” he says.
Retail is one sector that has great potential for big data. Its customer transactions, both on- and offline, conversations and intentions can all be brought together so that brands can better understand how to reach shoppers.
Wickes head of CRM Natalie Henty says brands must not be scared of the huge amount of data being generated and what it can help to achieve. “You must be focused on what you want to get out of big data or you can spend too much time analysing it,” she says. “Rather than trying to make use of every single piece of information it is possible to gather, you must get actionable insight.”
Some online businesses, such as Amazon, are already leading the field in this area. The site uses collaborative filtering technology, which allows it to develop automatic recommendations for customers based on their purchase history data.
Kurt Kendall, McKinsey’s business lead for its consumer and marketing analytics centre, acknowledges that companies have had information on their customers for years but claims things have changed. “With the right capabilities, you can take a whole array of new data sources - web browsing data, social data and geotracking data - and develop a much more complete profile and with this information, segment better,” he says.
“When a customer is standing in the store, a retailer can mine their databases and determine whether this is someone they want to keep and what it will take to keep them,” says Kendall.
“The retailer can combine information it has about past purchases with details about the customer collected through the company’s Facebook page and elsewhere to determine exactly what price to sell the item and if it is right to negotiate,” he adds.
Wickes works with a number of data companies, including data analytics firm Fuel, and collates information from various sources to steer its business and marketing decisions across different product categories. Henty agrees that combining big data sources allows Wickes to understand much more about the people who buy its goods.
“We analyse large segments of data and break them down into small pockets of consumers so our marketing spend is more focused,” she says.
This type of segmentation is made more powerful and innovative when big data is involved. It allows more micro-segments to be created and more personalisation using information such as clickstream web data, which logs how people browse a particular site.
The next stage for many brands that have already achieved detailed segmentation is to add in new areas. Social media sites provide brands with data for sentiment analysis, which can be integrated with retail EPOS data from tills to better understand how the way people feel about the brand translates into actual sales.
When this is achieved, people who already like a brand can be targeted individually in-store or with relevant DM and email content, effectively linking the on- and offline environments where consumers interact with a brand.
But while big data allows retailers to more effectively segment and market to their customers, the same information is also being used by price-comparison businesses to push down prices in the market. With more consumers using their own data to search and compare products and services, there is a risk that companies may see their margins squeezed as they try to compete.
Wickes’ Henty accepts that big data can put a downward pressure on prices but thinks retailers can still benefit. “The greater visibility for people to shop around means more opportunities for personalisation and to target individuals with the right level of discounts,” she says.
Direct Marketing Association (DMA) chairman Scott Logie says any downward pressure on prices will be fair for everyone if big data enables brands to target consumers more effectively. “If people allow their data to be used in marketing, it seems right that they should get something in return, in the shape of special offers or loyalty bonuses,” he says. “This has been happening for years with loyalty cards but with big data it is just not so explicit.”
While the DMA is excited by the potential of this emerging area, it warns marketers not to ignore the potential pitfalls of using large data sets. DMA chief of operations Mike Lordan says brands must be aware of best practice in handling data when dealing with so many possible sources. He claims that only one in four people say they are willing to share their full details with brands online, according to a DMA survey of consumer attitudes towards privacy.
“We’ve seen high-profile cases of companies losing data or experiencing data breaches,” says Lordan. “If brands are to profit from using big data, they must build trust. If they don’t, the size of the data sets will be eroded.”
Online retailer Notonthehighstreet brings together more than 2,000 creative local businesses and works with online data and technology company QuBit. The e-retailer’s interim chief marketing officer Maya Moufarek is aware of the DMA’s concerns but says analysing unstructured data is crucial to optimise pricing and to gain a competitive advantage online.
“The e-commerce industry has become distracted by endless analysis of traffic numbers with limited segmentations; a sort of analysis paralysis,” she says. “In-depth big data analysis brings the customer back into focus, whether in terms of how they react to your marketing or what they think of your products. This understanding is central to effective ecommerce.”
It is not just the sales process where big data is growing in importance. Manufacturing brands that do not sell directly to consumers can also use internal and external data to improve the design of existing products and create new ones based on any emerging consumer needs that the analysis highlights.
Procter & Gamble asks consumers for their ideas during product development and works with academics and entrepreneurs to reduce its own research and development costs through its Connect and Develop innovation programme. Meanwhile, IBM has an R&D facility near Winchester where brands turn big data into credible insight to improve products.
IBM’s emerging technology programme leader, Peter Waggett, says brands must work faster and dig deeper to get answers to business problems. Traditional data warehousing and business-intelligence solutions are not always enough.
One of IBM’s brand partners is examining whether every conversation recorded by its call centre could be turned into text, which could then be data mined. “This would give the marketing team additional insight into how customers feel about the brands and how they are likely to react to new products,” says Waggett.
Also in the business-to-business sector, Microsoft is using customer sentiment data to gain extra insights. Microsoft UK’s insight manager for the developer and platform evangelism team, Maxine Cook, says her brand can discover via the many online IT forums how the trade is reacting to a piece of technology and its competitors.
Microsoft and real-time mapping company Onalytica segment developers based on what they are talking about and the positive or negative buzz around the brand or a particular product.
“We are gathering data that tells us where our messaging is being understood and where it is causing confusion in the industry,” says Cook. “From the results, we can act locally in the UK or feed back to the parent company. PR is no longer the sole channel for monitoring and responding to reputation issues.”
During the summer, the company monitored developers’ reactions to previews of Windows 8, which has been designed to run on tablet computers. “The feedback we get on anything is constant and in real time and we combine the data with information gathered from our customer satisfaction surveys,” says Cook.
“We still use traditional market research but this is about taking the temperature of the market on a daily basis and measuring it against our business objectives so we can react quicker,” she adds.
It is this ability to react fast and use big data intelligently that will ensure both business-to-business and business-to-consumer brands improve their operating margins.
Media companies are also seeing the potential. The Press Association asked analytics company MarkLogic to create a content store, metadata repository and syndication engine for all its unstructured data, which includes news and sports video, tweets and articles.
The news agency holds approximately 15 million images, 60 million news stories and billions of individual data items, but was finding it hard to combine and measure data to add value to stories. “Sports, news and weather data were in different silos and we needed to join up the content to get better results,” says The PA’s director of technical architecture and development, John O’Donovan. “A Premiership footballer can appear on both the front and back pages.”
The PA now stores its data in a flexible XML format and it is indexed for easier searching. The company is considering cloud services for big data, including the next-generation cloud system called Hadoop, which would allow it to load and process vast amounts of data and explore some of its under-utilised data sets. For example, there is a huge amount of data that would provide an intriguing insight into how politicians spend money, reveals Donovan.
With so many firms not yet up to speed on how best to use data, the opportunities out there must surely outweigh fears on handling such vast quantities of information. Since global data quantities are predicted to rise by 40% each year, but IT spending is expected to increase by just 5%, according to McKinsey, the firms that spend more in this area are likely to reap the benefits.
With trillions of bytes of information available for companies to collect about their customers, suppliers and operations, the brands that invest in big data now are likely to see big rewards.
GlaxoSmithKline’s (GSK) CRM consultant James Parker believes the benefits of big data must be sold into every department within an organisation, rather than being seen purely as a marketing function.
“We have great analytical resources with data experts concentrating on specific projects, but the real value of big data comes when the experts are working towards common goals across the business,” he says.
Parker sees parallels with the early CRM debates of the late 1990s when the challenge for brands was to piece together information on particular consumers across all departments. Although he believes it is wise for marketing divisions to oversee the big data process, it too must be something shared across the whole organisation for it to have an impact.
GSK, which owns brands from Sensodyne to Lucozade, is already tracking consumers online and repurposing the data to benefit particular brands. “Consumers are providing signals of what they are doing, seeing, thinking and sharing at a point in time and this is valuable to help drive relationships,” he says.
The company is working with St Ives Group’s data management division Occam, which says GSK is aiming to build direct relationships with 1 million consumers in a year using social media as a base for research and multichannel marketing. Targeted offers and promotions will drive people to particular brand websites where external data is integrated with information already held by the marketing teams.
“We identify people who are mentioning a particular brand of ours but also track everything else they are talking about in the public domain to build a customer profile,” says Parker. “This is still an immature market and brands need to learn more about how to leverage this data.
“Ultimately, you must have the data- processing power internally to pull things together,” he adds. “There must be enough analysts to make sense of it. At the moment, many marketing departments are not willing to pay consultants or agencies to do it.”