Noise, Overload and Trust
Noise, Overload and Trust: The data challenges
The list of challenges to companies seeking to become more data-centric is long, but leading it is the ability to determine which data are truly useful.
The shift to a data-centric world is opening up new opportunities all the time. But this does not mean it is an easy task for companies to take advantage of them. Uncertainty over the future and difficulty in filtering the signal from the noise; immature tools; limited skills and expertise in this domain; the challenges of dealing with IT—all such concerns come into play, just as they did at the dawn of the internet era.
Across all functions and industries, a fairly consistent set of barriers emerged. The most pressing worry related to the difficulty in assessing which data are truly useful, selected by four in ten respondents (see Figure 17). "The ability to separate noise from the important things is a challenge, you don't learn it overnight," says a Fortune 100 chemical company CFO. "The access to quality data becomes more and more important with each passing year. And it also creates a thirst for more data, as you begin to understand what can be done with it."
"Concerns over data quality followed closely behind, along with a related worry about data overload. About one-third (34%) of respondents agree that they are increasingly worried about the quality of strategic decisions being impaired by information overload. In part, as BP's Mr Gray argues, this can be a function of maturity within the business—and a willingness for leaders to push back on reporting purely for reporting's sake.
"We went through this process a few years back, and ran into a break point when an executive questioned one day who was actually reading yet another thick report on some issue," he says. "It was a watershed moment, as other leaders starting asking the same question, so a lot of this got cut back very quickly. We've been through that, but it is absolutely a risk for people to be aware of."
Beyond these data-specific concerns, a lack of skills (35%), issues relating to organisational silos (25%), disconnects between IT and the rest of the business (16%), and a fear of the unknown among the management team (15%) round out the key barriers. Each of these can be major issues in their own right. For example, Mr Puryear notes that many companies are grappling with the accumulation of 20 years of legacy systems and applications. "This notion of unnecessary complexity in the IT environment is something that crops up a lot, and the reality is that so much of the data that companies need are locked up within these systems," he says.
The fine line on trust
Elsewhere, other concerns come into play for either specific industries or countries and regions, from worries over tightening privacy regulations that are restricting the access or availability of consumer data, through to considerations of trust. Coutts provides a useful illustration. As Mr Brannan explains it, while an online retailer like Amazon might reasonably be expected to track a customers' purchases and insights in order to sell them more, a private bank has to tread far more carefully. "Our relationships are based on discretion and trust. If we sit there and talk about the fact that we know a lot about you, we can easily step over that very fine line," says Mr Brannan. "So there's a question of how we present our insights back to clients and whether that's acceptable to them. Anything that erodes trust is a huge risk to us." Professor Neely argues that reputational risk is a significant issue for firms. "As you pull different sources of data together, you get to know more about people, whether customers or employees, and you have to be very careful about what you want to do with some of that information," he says. At Aimia, Mr Johnston explains that such concerns are paramount within the business, which has in turn led it to develop a core set of values relating to data, dubbed TACT (for Transparency, Added value, Control, and Trust).
New needs, new skills
These challenges apply just as much to using data for optimisation purposes as they do for more strategic ones. But as leaders seek to make the evolution from one to the other, several issues become more pressing. One that crops up repeatedly is the move beyond the skills of the traditional business analyst role, to bring in an advanced new skillset, matched with deep industry expertise.
At the core of many challenges, argues Mike Balay, vice-president for strategy at Cargill, the global food manufacturer, is the need to better align the management team's assumptions about the world with reality. This gets to the core of management theorist Peter Drucker's "Theory of the business", which argues that when these assumptions are true, whether about customer needs or organisational strengths, it is far easier for the business to be successful. "What's brilliant about Drucker's theory is that when these assumptions are in alignment, your business successes will seem effortless," says Mr Balay. "The companies that don't do well are those that don't share the same assumptions internally or don't sync them up to the world outside."
What's important here is that data essentially supercharges this theory, by providing far more information and options, such as the possibility to segment customers in multiple new dimensions. But this comes with tough new challenges: "How is your organisation going to process the insight? How are you going to get connect insight to an actionable set of strategic choices? And how are you going to get the organisation aligned behind you?" says Mr Balay, who makes a strong case for the need to invest in advanced skillsets to support this (see the Case Study below for more).
This is crucial. As Aon's Mr Clement advises, one of the challenges is that data capture the world as it exists today, not what might be possible. "The biggest opportunities come from fundamental change. If you take the old horse and buggy example, the data would have pointed you to adding more horses, not to inventing the automobile," he says. To help Aon improve on that front, the company has been continuously investing in its skills and ability to interpret data. "Ten years ago we had a lot of analytical capability, but the organisation just wasn't ready to take advantage of it," says Mr Clement. "You need the talent and we are catching up on that."
All this points to a marketplace in transition. As highlighted earlier, only a small minority (12%) of executives think their organisation has been "highly effective" at translating data into useful and insightful information so far. "Access to data is no longer a challenge these days. We have a tremendous amount of access and we can measure just about anything," says Xerox's Ms Carone. "To me, where we continue to have some challenges is how to translate this data into real, actionable results for the company," she says.
Over the past few years, Samsung has risen to become one of the biggest consumer electronic brands in business. In 2012, its popular smartphones accounted for the highest proportion of global sales. That success has come on the back of its innovation, design and other factors. But behind the scenes, the company's achievements have also been driven by its efforts to become a far more data-centric business.
Asim Warsi, vice-president for mobiles at Samsung India, explains that the company now tracks the sale of every single one of the tens of thousands of handsets it sells on a daily basis, which in turn powers the rest of the business. "This data is the basis on which our demand planning is done, which is the basis for production planning, and in turn the basis for raw material sourcing, so it's really the lifeblood of the organisation," he says.
From a sales perspective, the company knows every sale that happens, and where and when it took place, which helps supply the analytics that determine which sales channels and markets are performing, how much stock to hold, where credit could be considered, and more. "If there is a particular item not selling, or slowing down, we immediately know of it and can take proactive action," says Mr Warsi. "By being able to track our business down to the lowest metric and building our analytics back from that, it gives us an enormous visibility on our markets, channels, supply chain, customer behaviours and more. If this is not powerful, then I don't know what would be," he says.
At the same time, the company is strict about cutting away the superfluous data that might obscure this picture. This is just as important in running a data-centric business. "We purge all data that is non-decisionable, so we only really take the data that matter," says Mr Warsi. This is a challenge that many other firms are still grappling with.
All companies thrive on insights at some level. But as the world becomes more data-centric, these insights become increasingly difficult for companies to find. Mike Balay, vice-president for strategy at Cargill, gives a simple business scenario that his strategy team might grapple with: resetting pricing strategies for a particular product line. As this division gets more data, it can start to do more complex modelling of its customer segments. "You suddenly find out that it's not just a customer, but it's a customer in a particular place, with a certain product mix, a specific history with us, and a particular risk management strategy, all of which raises incredible complexities," says Mr Balay.
By overlaying this against price strategies and asset utilisation, the strategy team might uncover that the unit is actually throwing away 30% of its profitability with the last 10% of sales it is doing, as one example. Such insights can transform the fortunes of a division, but the volume and complexity of the data make such findings harder to uncover than they were historically. How to solve this? The simple answer: hire the smartest possible people. "The stars of the future are the people able to work with these large data sets and turn them into something actionable," he argues. This new breed of "super quants", or what Mr Balay dubs "elite symbolic analysts", are what will be needed to successfully grapple with the firm's complex network optimisation challenges.
But getting such rare skills requires a major corporate commitment. "Are you willing to hire and pay those people? Give them freedom and recognition? Make them partners in your business?" asks Mr Balay. None of this is easy, not least as this talent is rare, expensive and often unconventional. "But the conventional wisdom is not interesting. It is the weirdoes who are interesting," he notes.
