posted 1 Jun 2011 in Volume 14 Issue 8
Jeremy Bentley analyses recent research that finds there is a significant gap between expectations of enterprise search and what the technology actually delivers
We’ve all become accustomed to the idea that information should be at our fingertips. In their personal lives, most people ‘just Google it’ to get the latest news, check movie times, compare prices, or find details on a new product. But that doesn’t necessarily work in the workplace where internal information assets can be massive and hugely complex, as well as more difficult to access.
A recent study commissioned by Smartlogic and conducted by MindMetre Research demonstrates clearly that in far too many cases, organisations are simply not meeting the enterprise search expectations of their own managers and directors.
The results demonstrate that, even when applying the arguably weak ‘within two minutes’ on another organisation’s website as a benchmark for a ‘good search experience’, an alarming proportion of enterprises – more than half – are failing to meet even this basic standard with their own internal search platforms. What’s more, over a quarter cannot even meet the lowest benchmark offered in the study – completing a search within four minutes – which nearly 90 per cent indicate is an unacceptable time frame for finding a result.
Mind the gap
At its heart this expectation gap, and the failure of enterprise search platforms in too many organisations, is the evolution of information management in silos and the fundamental lack of a unifying technology that surmounts this problem. Much of the information held by enterprises – for example, proposals, product sheets, research, briefings, reports, policy statements, emails, memorandums, letters, blogs, minutes, client records, articles, white papers, business pitches and other key documents – is typically saved in a range of formats, stored in disparate places, and tagged by different people based on their individual priorities and opinions.
For most organisations that rely heavily on accessing information assets to serve their marketplace or community, generate income, create internal efficiencies, and enable the smooth running of day-to-day operations, overcoming these issues and improving findability is crucial. There are critically important content-management platforms and search mechanisms for leveraging information assets, including Microsoft SharePoint, FAST and Google Search Appliance. However, releasing the full power of these systems requires additional solutions.
For most organisations, taking a semantic approach to enterprise search can be central to enabling findability. Essential to a truly successful enterprise search solution is the capability to work out the context of any document or communication to unlock its full meaning – so that a search does not get clogged by unrelated pieces of information that simply share some of the same terms. That means creating a semantic model that brings together all the key terms that an enterprise would use in its business, and forms a standard for information classification so that documentation can be placed in context when it is being searched.
The semantic model can then be used to drive a metadata classification process, a fundamental step in the findability process. Within most organisations metadata is at best inconsistent and at worst non-existent because metadata terms are usually tagged onto documents manually – with the attendant huge manpower costs. Not only are mistakes very possible, but various groups, departments and individuals within an enterprise each have their own different methodologies and standards for meta-tagging.
In order to be able to apply metadata consistently, repeatedly, economically, according to an organisation’s standards and in a way that can be demonstrated to be trustworthy and accurate, some sort of an automated system is arguably the best solution. Semantic software can be put in place to scan documents and accurately tag them with the correct metadata, so that they can be placed in context and terms can be cross-referenced to ensure that all relevant content comes up in a search. And for organisations that have large-scale archives to be tagged, automation is really the only affordable option.
Failure to be able to effectively search and utilise information assets will continue to be an important issue as businesses and other organisations try to get more out of all the resources to hand in the face of a tough economic climate. And it is not just a concern for IT. It is an issue that management is becoming increasingly aware of and that needs to be addressed.
Jeremy Bentley is chief executive at Smartlogic. For more information visit www.smartlogic.com