Information Search Economics
We generally only do three things with information: we create or acquire it, we manage it, and we search it. While we’ve become fairly adept at doing the first two, information search is a surprisingly new concern in many organisations. Compliance issues have largely driven the uptake of search technologies, and since it is often senior managements’ head on the block, funding enterprise search projects has not been a problem.
While it may be prudent to acquire information search capability, there is also a real need. The volume and complexity of information continues to escalate and getting hold of information in an operational context has become increasingly difficult. Inefficiencies abound, and while the ‘don’t fix it unless it’s broke’ approach may still cause some inertia; there is a growing operational pressure to address the search problem.
Before looking at the economics of search we should put aside the categories of search technology offered by the IT industry. Enterprise search, business intelligence, digital asset management, content management and any other technologies that are highly focused on search will inevitably converge and become a single technology. So we’ll use the generic term ‘search technology’.
The economics of search are fairly easy to understand. The cost of search should be less than the value it delivers. As always the ‘cost’ part of the equation is fairly easy to establish – the value part isn’t. To get started we’ll consider some of the inefficiencies that search can address:
- Duplicated effort – either because information could not be found or because there was no knowledge of its existence.
- Excessive search costs – as people spend more time hunting down specific information in a rapidly growing information inventory.
- Lack of knowledge sharing – simply because there is no infrastructure to share knowledge.
- Undiscovered knowledge – not knowing what we know because nothing is telling us what we know.
The list is pretty well endless. The problem of course is putting a value on all of this. Various surveys tell of information workers spending twenty or thirty per cent of their time fruitlessly searching for information, and obviously it is fairly easy to establish what this is costing. Other inefficiencies are much more difficult to value.
The cost of addressing information search problems appears fairly straightforward, but there are some costs that will never be accounted for. Hard costs include hardware, software, training and consultancy. These can be listed neatly in a spreadsheet. The soft costs include the cost of setting up metadata (the information that describes information assets such as documents, photographs, presentations etc.) and time spent by information workers using the search capability.
The value delivered by search technology is almost impossible to calculate. However it is common to hear managers complain that decisions are often made with insufficient information, and one must assume that better access to information would add some value here. At the macro level there are strong arguments showing how knowledge and information capable organisations are valued more highly (higher share price). This should interest senior management if bonuses are linked to share price. It doesn’t require too much thought to show why this is the case. Lack of information is closely linked with increased risk and diminished capability. Easier access to quality information would reduce risk and improve capability and this would show in the overall performance of an organisation.
Most organisations will adopt search technologies simply because they have to. More enlightened management will see the opportunity to deliver a real advantage to their organisation (and possibly to themselves). Search is the other half of the information problem and at the moment global spend on search is less than two billion dollars. Global corporate spend on IT is around two trillion dollars. So we spend 0.1% of our IT budget on search – we’ve got a long way to go to balance the equation.
Finally it is worth remembering that acquiring data is nothing more than a cost (data input, document creation etc.) and only when we convert data to information (by addressing uncertainties) do we get any value.
