FEATURE1 March 2011

Order from chaos

Companies used to have too little data. now, if anything, they have too much. Can market researchers be the ones to corral all that information, or must they sit and watch while their influence declines? Tim Phillips investigates.

“Data is streaming in from multiple sources simultaneously… 2011 will be the year when this data starts to converge on market research departments.” That was the call from Forrester research director Reineke Reitsma in her report on predicted developments in the research industry this year.

Forrester’s opinion is that the clientside market research department is both the natural clearing house for customer data and the one place where the problem of gaining insight from all these data sources can be solved.

But to do that, clientside research doesn’t just need the tools and expertise to make the process work. It needs political power – the cooperation of the rest of the business and the assistance of external agencies and technology suppliers.

It also needs to strike a balance between policing (ensuring that data is collected in the right way and reported to the clearing house) and consulting, to make the collection of data and the discovery of insight a joint project between the research function and the rest of the business. Forrester expects this process to take years rather than weeks. But is it an inevitable transformation for organisations that use research? Or an impossible one?

The data deluge
With 15 years on the clientside running research and insight projects for organisations from the Department for Education to McDonald’s, Lisa Edie has doubts. “The whole concept looks like an agencyside idea of nirvana compared to what’s going on in client organisations,” she says.

“The whole concept [of one department corralling an organisation’s data] looks like an agencyside idea of nirvana compared to what’s going on in client organisations”

Lisa Edie

Edie lists the types of information that a department would need to corral to truly act as a clearing house: market research and customer intelligence, of course, but also transactional data, which she warns is “a very different animal” to research. Most difficult of all, she says is the ‘soft’ anecdotal data that comes from customers, and competitor intelligence in the form of ads and marketing materials observed by a company’s staff as they go about their lives. “The other thing Forrester doesn’t cross into,” Edie continues, “is external information, things coming up on the periphery – looking at different service operations that aren’t competitors.” Clearly, anyone seeking to bring all these strands together to create something useful faces a daunting task.

Reitsma’s colleague Roxana Strohmenger explains why Forrester’s market insights team believes it can be done. “All these companies are experiencing a data and information explosion. We are deluged with this information, and we have to take a step back and control it,” she says.

Strohmenger believes it will rapidly become apparent to forward-looking companies that they need a function to separate information from noise, and to protect standards. For Forrester, the recent tsunami of data originates from two sources. First, the DIY research taking place in other departments or functions (much of which the clientside research department may not even know exists). Second, the data that flows in from social media monitoring which, again, is often done by other departments. In isolation, they offer as many problems as insights, and can potentially give misleading and contradictory signals.

Nice cop or nasty cop?
If clientside research is only vaguely aware of what the organisation is collecting and how it collects it, some form of policing, Strohmenger argues, is overdue.

“We are deluged with this information and we have to take a step back and control it”

Roxana Strohmenger

“We’re seeing a lack of internal checks and balances on internal data collection. We’re seeing other departments that may not have professionals on their team who have the right methods to collect data. Poorly designed surveys, bad execution… the questions are double-barrelled, or the analysis is not stating what the insights are.”

Ian Lewis is director of research impact consulting at Cambiar Consulting, helping clientside research departments become more effective. Until 2010 he was in charge of the international consumer insight and strategy function for Time Inc, and has previously done similar jobs at Sterling Winthrop, Pfizer and Unilever. He spells out three options for research departments who want to increase influence. The most drastic, used by a minority of the departments he works with, is a rule that all customer research goes through the research department or must be approved by it: research as an internal insight police force. This is the ‘nasty cop’ approach, which Lewis warns is very difficult to impose successfully.

Instead he favours two light-touch approaches. One is to create an environment where business units can conduct their own data gathering, but with “guard rails” in place. He cites Microsoft (not a client) as an example. The company has 5,000 employees who want to acquire data about customers, together with a high degree of technical ability and an appreciation of the value of targeted marketing.

“The guard rails could be a single software platform, they could mean requiring training for someone who wants to conduct a study, they might mean that the research department has to vet proposals to do research,” says Lewis. This approach also defines the situations in which it is acceptable to do research, and identifies repetitive research tasks in order to standardise them.

The other light-touch approach is where the internal research department does not mandate that research is done through it, but creates its own low-cost research products for the rest of the business to use. The ability to call on this shared expertise becomes an incentive for its internal customers.

These ‘nice cop’ approaches can serve to build understanding between research and the business. Another reason not be nasty cops, Edie points out, is that it’s time-consuming, tedious and might not work because to capture information systematically requires the willingness of the employees from whom you are capturing it. Insisting that the business reports soft information can also create bureaucracy, meaning that such attempts to boost the power of research can end up having the opposite effect.

“It’s very difficult to cultivate, because you’re saying that your staff have to be responsible for their own filing. Why doesn’t that work? It’s dull, it’s boring, there’s no incentive to do it. It’s like taking medicine. You’re lowering the power of the customer insight function to being an administrator telling people that they’re not doing their filing properly.”

Actions and words
Harris Interactive has developed a service called Research Lifestreaming, which combines social media monitoring with survey research to provide clients with two streams of data as one. The online researcher has got some of its panel members to allow their social media activity to be tracked, so that their survey responses can be linked to behavioural information. It has already tracked more than 20 million Facebook posts.

“Clientside ownership of social media is very important,” says Harris’ marketing director Ed Chatham. Often it gets pushed off into marketing, or even PR as a reputation thing.” It’s important, he believes, to take social media research beyond stating that ‘Some people have said some things about your brand’, and facing up to the ‘so what?’ test.

“When people talk about phone networks, or the banks that rip them off, we know if they are customers [of those companies]. That gives us a deeper layer of information.” For Harris, Lifestreaming is a way of making sure research doesn’t miss out on data from new sources. “With increasingly fragmented audiences, that needs to be controlled by a central department,” Chatham says.

Tools for the task

?Forrester’s Roxana Strohmenger identifies enterprise feedback management tools as essential to a clearing-house model: making existing research available to a wider audience, and saving time and money from being wasted on unnecessary new research.

“We’re taxing customers with frequent surveys, or they are receiving a survey for their recent shopping experience from the e-commerce department, while at the same time the customer experience department is sending a receipt survey about the same shopping experience. The market insight professionals should take control and use tools like enterprise feedback management as a way to bring the data together in one location,” she says.

This makes for better use of money, as well as better use of time. Insight Marketing Systems produces the Research Reporter tool, which delivers secondary research from sources such as Factiva and Mintel to one location in an organisation, where it is indexed, tagged and made available to other departments. Customers include Coca-Cola, Nestlé and Reckitt Benckiser.

“The research teams can have a single subscription and get the content integrated into their portal. The internal clients might subscribe to specific topic areas, interests or brands and the research department tags the content as it comes through,” says co-founder Chris Forbes. “They have gone from having a subscription for which they have no idea whether it is of use to the business to having a subscription where they can see exactly what is being used.”

Any clearing house will need some form of self-service mechanism to deliver insight to the business, and that will require skills in making data searchable. “One of the things that client researchers haven’t started doing yet is moulding their content to this channel,” says Forbes. “They are still doing big PowerPoint decks, which are fine as a presentation aid – but not when you’re sitting on your own, looking at your internal portal, and you’ve found 5,000 hits, one of which is a 50-slide PowerPoint deck, and the information you want is on slide 50.”

Can it be done?
Strohmenger, while recommending that clientside research departments start on the journey to providing a clearing house function, doesn’t underestimate the time or investment it will take.

“To move from being a service bureau to a strategic partner, a third eye or gut check for each department, is not something that one day the department says, ‘We’re doing it’, and it’s done. The market research department has to develop a marketing plan about how this will be beneficial to the company as a whole,” she says.

That means doing a research audit, and quantifying the value to the business of the research both inside and outside the research function. Only then can the research department compare the value of a distributed, laissez-faire approach with a centralised, consolidated source of insight.

It may also mean, she adds, re-running bad research to show how more powerful insight can be generated when professionals are involved, or acting as a form of relationship counsellor to involve more than one department in a common research project. Showing rather than telling.

Some organisations already have a central clearing house structure, or are attempting to integrate channels for research. The first qualification for researchers who want to transform their departments, says Steve Reiman, who recently moved from the clientside to become head of qualitative research at Harris Interactive, is the ambition to take on a tough job. Working for mobile operators, most recently O2, has made him aware of the scope that a clearing house role can have, even encompassing the job of training contact centre staff in how to interact on customer forums.

“It takes a brave, forward-looking or ambitious insight person to spread their wings. Some researchers are purists, and think all of this is highly dangerous,” he says. “Inevitably this involves the insight department with the people responsible for channel management. It’s a massive headache to companies that they need to integrate those channels to integrate the customer experience.”

“You need a mechanism for control of the information that is running around the company. Philips is a large organisation, I don’t know all the surveys that get done”

Federico Trovato

At Philips, vice president for strategy and consumer market intelligence Federico Trovato believes the company has successfully created a clearing-house structure based on this sort of continuous contact, but he warns that openness can create problems as well as solve them. “Policing [alone] doesn’t work long-term. It may work in an emergency, but long-term you need to show that you are adding value, giving a more complete view of your consumers,” he says. “I have some challenges: some data, like market trends, can be politically difficult if it supports an idea or a project, or starts a new work stream. But if there is good cooperation in the company then it can be worked out.”

He overcomes conflict by positioning the customer and market intelligence function as an internal consultant, willing to take on tasks for other departments that would usually be out of the remit of a research department. One example is a media effectiveness audit that compared approaches across a broad range of markets, rather than leaving each business unit to analyse its own data. “We sourced the data from our media department so we could understand the value of the investments made in each market. We made a huge analytical effort to try to understand the value of our media investments to the business. That led to decisions to pull out of advertising or invest more – the media departments tend to keep this data for themselves, but we acted as a clearing house.”

For Trovato, this is not rooted in software tools, but in a systematic approach, supported by the business and validated by results.

“First, take stock of everything that is happening. You need a mechanism for control of the information that is running around the company. Philips is a large organisation, I don’t know all the surveys that get done. But I’m trying to bring a way of working which means you have a brief that you fill in, and a request for data, and if it’s stored in a centralised system you can have a mechanism in place so that if you are requesting the same type of data three or four times there’s a warning signal.”

Ultimately, though, even a successfully beefed-up research function might wind up becoming an agent of control. Lisa Edie warns that the attention of clientside research could be diverted by taking on a policing role – collecting data, telling people what not to do – while the responsibility for creating insight falls to a broader knowledge management function.

Research is just one of multiple data sources needed for insight, says Edie. An econometric model would typically only attribute about 30% of decision-making to research, she reckons.

Perhaps the only way for clientside research to get above that 30% is to begin to integrate fresh data sources, such as CRM and business intelligence. Forrester’s analysis predicts this will happen – but gradually. “Once we have capabilities, let’s advertise successes, build examples of how this works so people will see how this is a benefit. A bit of politics, a bit of communications strategy, then slowly bring in the other departments and see how it works for them,” says Strohmenger.

“But it’s going to be a long arduous process. It will not be solved within a year.”

1 Comment

13 years ago

Although I absolutely agree with this analysis, I think Steve Reiman hits the nail on the head in terms of the huge problem this represents for market researchers - which is down to their training and narrow world view. To be able to deliver this type of useful business information from various sources the 'provider' has to be comfortable with merging, mixing and exploring various data/information to help answer a 'business problem', be a jackdaw and be comfortable and truly understand the value of secondary research (consistently and slighlty 'snobbishly' dismissed as 'oh desk research' by the MR community)...this is just miles away from most researchers ability and comfort zone. The first reaction of most market researchers to a business question is "I 'll design you a survey" not "wWat have we got?" ; the value of research data is consistently overvalued against other sources and researchers generally would rather cut their arms off than mix their data with that from another source! Busy executtives have to make quick decisions by weighing up a whole series of imperfect information...market research is just one of these. sources of, yes, imperfect information Researchers have to grasp this or be further marginalised. Market analyist skills are required to do this and Market Researchers who really want to consult and add value need these skills as much as sampling, questionairre design, etc etc

Like Report