Depending on the industry, businesses typically spend anywhere between 1% and 15% of revenues on information technology. Manufacturing tends to be at the lower end and financial services at the higher end. Information costs dwarf these costs however. Remember that information costs are primarily labor costs – people entering data, writing emails, searching for information – and so on. These costs range anywhere between 20% and 80% of total costs. Your CEO is an information worker, managers are information workers and of course so are people who work in call-centers and customer support.
Considering the very large amounts spent on information it would make sense to have some form of information strategy – a plan to maximize the utility of information and reduce its cost. And it is worth remembering that less information may well correlate with lower information cost – which calls into question the proliferation of information in an organization.
Before going any further it is worth discriminating between data and information. In many disciplines (economics, sociology, communications engineering) the concept of information is well defined. Information reduces uncertainty. Data on the other hand is just so many zeros and ones on a disk drive. It is worth remembering this beautifully simple definition because it is a reference point for everything else.
The aim of much technology investment is to extract information from data. Business intelligence, enterprise search and analytics all serve this function. But this is not a starting point, even though our love affair with technology causes us to often put the cart before the horse. The starting point is the art and science of asking questions. Where are your uncertainties, and which ones are the most important ones? The answer to this question depends on your role and what is important to you. The CEO may have uncertainties about an acquisition and is likely to use a consulting firm to address them. The CMO will almost certainly have uncertainties around the funding of various marketing initiatives. In this case it is highly likely that data mining projects will help reduce those uncertainties. So there are various levels of information that address specific uncertainties.
In most organizations there is no single individual charged with the task of maximizing the value of information and reducing its cost – which is strange considering its dominating position in the costs an organization incurs. Chief Information Officer is a title that implies such a role, but generally speaking it does not encapsulate information strategy and is more concerned with technology. The rise of analytics has introduced some new positions too – the Chief Data Scientist, or Chief Analytics Offer etc. But these positions are also more concerned with technique and technology.
Someone dealing with information strategy would be aware of where data costs are being incurred and the value these data brings to the organization – at a high level. And then the unthinkable might be thought. Reducing these enormous costs might mean reducing the amount of data that is collected and processed – particularly when it adds little value. Such a person would have a handle on the value being generated by data, and the cost associated with it. Since information is more ubiquitous than finance this person would have a remit that spans across functional boundaries – just as the CFO does. Is it going to happen? No – of course not. A move to place such a person in this position would be resisted by every functional head – including the CFO.
As a result most organizations will continue to pay more and more for their data (more people generating more data – social, email etc) and simply continue to collect increasing amounts of the stuff – regardless of its value. Big data encourages this notion, but the cost of storing the stuff is negligible compared with the cost of origination and processing.
In the so called information age most organizations still have no idea of what their information costs are, or how to use information for uncertainty reduction. The glimmer of hope is that data mining technologies implicitly incorporate notions of probability and uncertainty, and it is just possible that these ideas might get broader exposure. What can be said for certain, is the few organizations which happen to take information strategy seriously are very likely to see a significant return. When information represents 80% of your costs, you don’t really have to do much to show a significant return.