The new faces of customer management

Integrating the never-ending stream of unstructured data into CRM systems is proving difficult

Remember when your data warehouse was the answer to a company’s customer information needs? That was the 1990s, and by the time the next decade had rolled around it was clear a data warehouse alone could not solve an organisation’s sales and marketing needs.

Usher in the era of customer relationship management (CRM) systems.

But in the new world of social media, a fire hose of customer data has been unleashed in the form of Tweets, Facebook updates and myriad other forms of unstructured data. Integrating that torrent into the sales and marketing systems of major corporations has to date proven beyond their capabilities.

According to the sales director at specialist data consultancy Servian, Nigel Peach, one of the issues companies face is a disconnect between many of the software systems they rely on.

CRM systems for instance mostly contain historical data, and while useful for customer retention, may be of little use in customer acquisition. That task has fallen to campaign management systems, but the two rarely talk to each other, and campaign management systems may not easily integrate unstructured data from Twitter and Facebook. Hence many organisations are looking to new advanced analytics tools.

Peach says the goal is to combine all data assets, both structured and unstructured, into a unified insights platform, or customer hub. “That is where the market is trying to get to, which is the whole Big Data paradigm,” he says. “Where people are struggling is tying those together.”

Exactly who is responsible for bringing all the pieces together remains undecided. Peach likens it to bringing together the “Madmen with the maths men”.

“We are trying to turn these marketing people into ‘maths men’, and make the marketing guys data insights people,” Peach says. “Traditionally they haven’t come from that. But the companies that will win are the ones that combine the external data assets with the internal data assets.”

Marketing momentum

Marketing is playing a much greater role in the direction of IT spending, so much so that research group Gartner claims the chief marketing officer (CMO) will spend more on IT than the CIO by 2017.

Right now most of the experiments with Big Data technology are taking place in the CIO’s office, but the business case is coming from marketing and sales, Gartner’s principal Business Applications research analyst, Praveen Sengar, claims.

“It will be the CMO driving it, because they will want to do personalisation, proactive engagements and targeted marketing,” he says. “The marketing department will be taking a more active lead, and IT will be very closely involved.”

Today however most organisations are still working on getting their internal data better aligned to their goal of having a “360 degree” view of customer activity. “This is what most organisations are focused on, and they have taken a considerable amount of time in getting to this,” Sengar says.

Because of this work, those organisations investigating the use of external data are doing so in isolation of their internal systems. Blending the two is expected to be a future stage of development.

“We have not reached that stage,” Sengar says. “That is a two-year journey further out, where we start seeing the internal data and external data coming together in a customer context.

“The reason for that is most of the implementations in Big Data today are generally experimentation or happening in silos. Many are still struggling with managing the internal data, and for this to happen business and IT have to come very closely together to define the use cases.”

Forging a new customer experience

Some industries, such as banking and insurance, are more advanced in plotting their strategies than others, and as Sengar indicates it is usually the marketing department driving initiatives. Director of marketing and digital at Budget Direct, Jonathan Kerr, says his group and the IT department work together on pretty much everything now.

“We create solutions together and we are all driven by a passion to make things better for our customers by leveraging the best technology out there,” he says. “Get that right and the rest takes care of itself.”

Budget Direct makes a strong distinction between the customer management applications used to administer the customer’s policy, and applications that collate data to help improve the prospect and customer experience. Technology from Adobe is used to ensure marketing campaigns remain efficient, with constant monitoring employed to give an indication of what prospects and customers are looking for.

“We get a huge amount of insight from the combination of our website feedback system and site analytics, which drives our usability and multivariate testing program,” Kerr says. “That effort helps us deliver very intuitive customer journeys. When you get that right, cross-sell and up-sell takes care of itself, because the other product lines and offers are presented at the right time for the customer.”

The insurer also combines its web and campaign analytics with feedback from customers via its call centre.

“For instance, we quickly learned [customers] wanted smarter, faster solutions via mobile, so we built Australia’s only 60-Second car insurance estimate for mobile,” Kerr says. “We got rid of six estimate questions by returning specific car make/model-type information when you type in your car rego.”

But while Budget Direct is highly advanced in the use of its own data, that is still a long way from being able to poll the vast amount of unstructured data floating around in social media. The organisation is not alone.

Executive manager of group digital strategy and innovation at Suncorp, Murray Howe, says integrating external data to better understand individual customers is something his company would like to try but is hesitant to leap into.

“Like most companies we see a lot of benefit in making our interaction with customers more relevant to their needs, and shifting towards more of a ‘sense and respond’ relationship with our audience,” he says. “The main barrier for that is consideration around privacy. This is a really great example of where technology, and maybe even audience sentiment, has outpaced the legislation.

“As a financial services company we need to be very sensitive about how we use our customer data, because it is at our core and customers have a lot of trust in terms of what we do with it.”

Unstructured social data is not the only type being used to drive outcomes for customers. The not-for-profit health insurer, HCF, uses health statistics data provided by the federal government to better understand its customers.

Page Break

According to HCF’s chief information officer, Patrick Shearman, this is important for HCF in terms of product development and determining what clients can claim for, and how much.

“You can start to tell how many HCF members in a particular postcode have had a disease, or what age they are,” he says. “That feeds into the overall claiming process and calculation of benefits.”

Much harder still though is the notion of integrating discrete social data with specific customer records. Even many of those organisations actively monitoring social media are doing so for sentiment analysis and ad hoc responses to questions and complaints, rather than proactive marketing and customer acquisition.

National Australia Bank for instance is an extensive user of social media, having established its Social Media Command Centre in late 2012. A team of up two dozen staff monitor different social media channels and use analytics tools to deliver insights into the conversations that are happening.

NAB’s general manager for technology strategy and innovation, Denis Curran, says most data analytics efforts right now are directed to creating business benefits, such as better risk weighting. “The holy grail is about creating new valuable product and service propositions,” he says.

“The new propositions come about when you look at things that can take richer sources of data, and that’s things like text analytics and voice analytics we’re experimenting with, and understanding more about customer s’ attitude and behaviour to things.

“You can see different aspects emerging. The challenge with many of those is the ability to turn them into valuable products for customers that they will appreciate and pay a premium for.”

Identification techniques

One of the greatest difficulties for marketers is in matching up external data from social feeds to internally-held customer data. How for instance can a company know that 10-year customer John Smith is also complaining about customer service via Twitter as @johnnysmith77?

Matching real people to online personae has become the specialty of Sydney-based consultancy, Digiviser. According to chief executive officer, Emma Lo Russo, her company can glean sufficient information from a person’s LinkedIn, Twitter and other publicly-available services to match up approximately 60 per cent of a company’s customer base to online personae.

“We use multiple markers, and each marker might be a combination of first name, surname, or any information that might be common,” Lo Russo says.

For example, a person’s public LinkedIn profile might contain their location and employer, and even their Twitter ID. Alternately, they might use their real name or employer in the brief description of themselves they post on Twitter.

“There might be enough in there that allows us to say alright, we know we’ve got someone, so let’s use that information across the other platforms,” Lo Russo says. “It is all probabilistically assigned, and then when we think we’ve got the candidate to a certain probability, we will say ‘here’s a match’.”

Even then however she says it is impossible to be 100 per cent confident, so companies must use such associations with caution. But even allowing for errors of assumption, the potential for targeted marketing is highly attractive.

“A person who indicates they have just booked a holiday, or are thinking of booking a holiday, would be a candidate for travel insurance,” Lo Russo says. “A wedding is an indicator that a person might be ready for a person loan.”

With the launch of the iPhone 5 last year, Digiviser was engaged by a telecommunications company to identify which subscribers might want to upgrade. Failing to reach those customers with the right offer might have led to significant customer churn.

Lo Russo says such techniques can also be used to better inform and target a broad range of marketing campaigns, by enabling better customer segmentation. Digiviser is able to tell a company where its customers ‘hang out’ online, what they talk about, and the language they use.

This can in turn lead to more targeted search and display advertising campaigns, with more tailored content placed in optimal locations.

While interest has been increasing significantly over the past 12 to 18 months, particularly for finding and converting sales leads, Lo Russo says technology is still holding many back. In the case of the telecommunications client, that was run as a separate campaign.

“Interest is definitely there, but the number of organisations that can have their CRM actually set up to enable that real-time integration of data are probably fewer,” she says.