Add insight to information for best results

The ability to exploit data in a myriad of different ways is key for successful organisations in the information age. According to a report by consultancy McKinsey, "One of the trends that will help shape businesses and the economy over the coming years is leveraging information in new ways. Organisations can do this by using information to make smarter decisions and to develop insights that create competitive advantage and new business models...

The ability to exploit data in a myriad of different ways is key for successful organisations in the information age. According to a report by consultancy McKinsey, "One of the trends that will help shape businesses and the economy over the coming years is leveraging information in new ways. Organisations can do this by using information to make smarter decisions and to develop insights that create competitive advantage and new business models."

But while few would disagree with this as a vision statement, it fails to reflect reality.

Data morass

Neil Macehiter, partner at analyst company Macehiter Ward-Dutton, believes that, ultimately, all businesses are concerned with exploiting information and using it more effectively. "The challenge is that they have vast amounts of it and so often don't know what they've got," he says. "They may not always be collecting the most relevant stuff or be able to access it, and the quality and consistency is variable."

So while most enterprises find themselves swamped with data, many fail to make the most of it, despite having introduced rafts of business intelligence (BI) tools from different vendors.

One of the problems here, says Andreas Bitterer, a vice-president of research at Gartner, is that organisations are often simply unaware of the possibilities offered by such technology and so tend to focus on historical reporting and analysis activities.

"The trick with BI is not finding out what happened yesterday, but predicting the best next action," says Bitterer. "The value comes from forecasting and looking into the future, but many companies are missing a beat here. They're so focused on the reporting and analysis bit that they're not using what they've learned and applying it effectively."

Organisational culture

Another issue is that enterprises tend merely to implement business intelligence. But they also need to introduce a sound enterprise-wide approach to information management and governance and revisit organisational culture to promote collaboration and information-sharing.

"It's no accident that the I in information technology stands for information because it's about exploiting data in different ways," Macehiter says. "It's partly about making smarter decisions, but it's also about combining information with other assets such as people to harness innovation."

While companies have tried to grapple with this issue for years under the banner of knowledge management, traditionally it has been a top-down approach with a tendency to be prescriptive. But the arrival of new management techniques could change this.

As the McKinsey report points out: "Google fosters innovation through an internal market: employees submit ideas and other employees decide if an idea is worth pursuing or if they would be willing to work on it full time."

Prediction markets

Another idea starting to gain some credence in the US is that of the prediction market, a strategy already adopted by Intel in conjunction with more standard short-term forecasting techniques in an attempt to estimate product demand more accurately.

Vuk Trifkovic, an analyst at Datamonitor, says prediction markets are "a sort of freakonomics" that work on the same principle as fantasy football leagues. "People set up exchanges of ideas, put pretend money on a given notion and track it as it gains or declines in value," he explains.

Although prediction markets are no less subjective or vunerable to hype than real ones, Trifkovic believes the concept has value, particularly if combined with more scientific ways of predicting performance. This is because "they can be a really fun way to engage people more and foster a sense of community".

The problem lies in how to balance the use of automated, standardised processes with a more creative approach that relies on human intelligence and flair.

"An organisation's ultimate differentiator is its people," Trifkovic says. "So you have to ensure that everyone's working in a standard way so that things happen and there isn't redundancy in business processes. But you also have to provide people with the right tools to promote creativity and I don't think anyone's put their finger on how to do that yet."

Another consideration is organisational culture, which in many instances is typified by the "information is power" mindset. At a personal level, let alone a departmental one, people are often reluctant to share information with colleagues and need to be persuaded with carrot and stick methods.

Tim Jennings, a research director at Butler Group, says, "It's about setting down guidelines in an information charter and rewarding people for sharing or contributing to corporate knowledge bases. Quite often you see organisations setting out information policies so that staff know what is expected, but it's generally just an advisory thing and isn't built into appraisals or contracts."

One technology trend that will help in breaking down information silos is the increasing integration of business intelligence tools into enterprise application suites such as those from SAP and Oracle. This is because analysing and exploiting information will become progressively embedded in the way staff work rather than simply being a separate activity undertaken by subject specialists.

Feedback loops

A further important step forward in optimising internal data usage will be the incorporation of performance data of all types into business processes to create a feedback loop. Although the biggest barrier to this at the moment is poor data quality and a lack of faith that information is accurate, it is a logical outcome of the raft of performance management initiatives taking place at the moment.

"If you start taking all this information and distilling it down into key performance indicators such as customer enquiries need to be resolved in two days, it's perfectly possible to feed that information back into your CRM or ERP system," Jennings explains. "So if a process fails, staff or systems receive an alert and are instructed to take action to deal with the situation."

While such a scenario is currently the exception rather than the rule, it is more widespread in customer information than elsewhere even though there is still much work to be done in the field of master data management, which is crucial in creating a single view of the customer.

Collaborative filtering

Another area likely to grow in importance is the move to near-real-time modes of working and pull- rather than push-based modes of selling. McKinsey cites online retailer Amazon as at "the forefront of advanced customer segmentation" in that its recommendation engine correlates the purchase histories of individual customers with others who have bought similar goods in order to suggest items of potential interest.

Beyond the online world, the idea is to provide call centre agents with insight into customer behaviour so they can up-sell products and services at the most appropriate time, for example, or to alert check-out staff that a given client might respond to a special offer.

While a move to real-time can require major upgrades to IT infrastructure, some organisations are already embracing the concept in a limited way. Differential pricing models are starting to emerge although relatively basic, they are likely to become more sophisticated over time.

In future, customers can expect to be charged according to their spending power, purchasing history and the time they visit a site, or a combination of these and other factors.

McKinsey points out that some US toll road operators "are beginning to segment drivers and charge them differential prices based on static conditions (such as time of day) and dynamic ones (traffic)."

Another manifestation of the pull concept is enabling individuals and communities of customers to design their own products. BMW, for example, allows customers to customise their own production vehicles, while encourages people to design their own T-shirts and vote for a winner, before selling them to participants.

"We're starting to move to more of a pull model, which is where technologies such as Web 2.0 come in by building communities of interest," says Macehiter. "These communities enable you to tap into useful information, aggregate it and target an active customer base for mutual benefit. So it's about the customer telling you what's appropriate and adding value rather than just being a passive purchaser."

Mash-ups offer further interesting possibilities. "If you know where your customers are and you use this data in combination with mapping technology such as Google Maps, you can determine the ideal location for a store based on what products to sell where, for example," says Bitterer.

New business models

The exploitation of information in this way may also lead to the emergence of new business models. The ability to aggregate more data as a result of digitising more processes and activities may create "by-products or exhaust data that companies can exploit for profit", says McKinsey.

For example, retailers with digital cameras installed to prevent shoplifting "could also analyse the shopping patterns and traffic flows of customers through its stores and use these insights to improve its layout or the placement of promotional displays". They could likewise sell such consumer behaviour data on to their suppliers to help them improve their merchandising activities.

A second scenario is the growing importance of information-as-a-service. Hosted information services already exist in the shape of credit bureaux that check customers' financial risk ratings on behalf of third parties such as banks. Bitterer believes "we ain't seen nothing yet" and that the market is set to explode in the next few years.

Walk before you run

But before they can take giant leaps, most companies have much work to do just to get the basics right. There are coherent enterprise-wide information strategies to devise and rafts of data to cleanse.

Before they can even begin, says Macehiter, organisations need to change their attitude from one of "what can we do with our data?" to "what are we trying to do with the business? what information do we need to do it? and how can we capture that?"

"Look at the information assets that you have and require, and keep very focused," Macehiter recommends. "There's too much of a temptation to boil the ocean and come up with answers to everything, but that's not going to happen. And in some cases, it may simply be a matter of harnessing more effectively what you have already."

Read more on Business intelligence software