posted 25 Aug 2004 in Volume 8 Issue 1
KM seeks perfect S&R partner...
In the current KM climate, the role of technology is a contentious issue and is still widely debated by leading KM pioneers. Many feel that technology negatively impacts knowledge-sharing initiatives. Rebecca Cavalôt investigates areas where search and retrieval technology is failing organisations and discovers the secrets to a successful alliance.
The relationship between knowledge (S&R) management and search and retrieval is a stormy one, and what should be a harmonious partnership has hit the rocks due to a pervasive distrust of technology in the KM space. Leading pioneers are citing softer disciplines such as narrative, storytelling and CoPs as key to the future of the discipline. “Embrace the worker,” they cry from the podium: meanwhile, technology has been left to skulk in the shadows, with the implementation of IT tools more of a KM millstone than a milestone.
Despite this, search and retrieval strategies are in place within almost every organisation and a multitude of employees rely on them on a daily basis. The Gartner Group defines KM as “an integrated approach to identifying, capturing, retrieving, sharing and evaluating an enterprise’s information assets”. However, it is imperative to realise that these assets are not simply the tacit, uncaptured knowledge stored in the heads of individuals – they come in many forms, including databases, documents, policies and procedures. Which is where search and retrieval comes in.
Rick VanderKnyff, web manager at the University of California, sees the relationship between KM and S&R as intrinsic. “Search and retrieval is inherent in the ‘management’ part of knowledge management,” he says. Shahnawaz Khan, manager of KM solutions at Srishti Software, agrees that the bond between the two is not only natural, but is also undeniable. He believes that even with the best mechanisms for generating and sharing information and explicit knowledge, all efforts are wasted if that information cannot be retrieved easily and efficiently. So how can organisations ensure that the search and retrieval strategies they employ will support and promote the knowledge sharing, user-led culture favoured by KM’s leading thinkers?
User-led is a phrase that surfaces over and over when discussing S&R strategies in relation to managing knowledge. Michelle Leclercq, global portal manager at Centrica, a utility and services provider, is an advocate of tailored solutions that take the unique nature of each organisation into consideration. “It is important that the search is tailored to the company’s needs and knowledge software, and delivers good return rates on information searched,” she says. Search and retrieval tools need to be optimised in accordance with what the company actually wants to achieve with its KM strategy.
Annie Wang, principle information architect at Aon Corporation, a (re)insurance and risk-management company, believes that this tailoring has to go one step further. Wang has been involved in numerous S&R projects and has found that the most variable facet of any project is the user: those individual employees who are accessing the information. “After each project, we always follow up with extensive user testing and observation, actively seeking the user’s feedback,” she says. This is echoed by Khan, who sees user experiences as an integral part of the construction and development of a successful S&R strategy. He believes that considering the users’ real-world encounters and acting upon their feedback will not only complement, but actively augment a KM programme. Staff who are encouraged to take an active role in modifying an S&R solution will be more likely to use it effectively.
When looking to develop a search and retrieval strategy in a KM context, companies need to adopt a holistic outlook. Content can be managed and organised, but if the actual knowledge seekers are unable to use the S&R tool to access this knowledge, then the whole solution begins to be viewed as a hindrance; often an extremely costly one at that. At Aon, users are intolerant of search and retrieval tools that fit this definition. “Our users demand instant and accurate access to mountains of information,” says Wang. “They get very frustrated if we don’t return good results quickly.” The University of California had a similar problem when attempting to introduce new users to web information that was managed through dozens of independently run static sites, as VanderKnyff explains. Employees knew the information they required was there – somewhere – but were unsure where to begin looking for it. The result? Frustration at the waste of time and resources, which could have been avoided if a more holistic approach had been adopted.
To avoid this scenario, VanderKnyff suggests gathering explicit knowledge and locating it in one specific area. When embarking upon building an enterprise portal, past experiences have taught him to collate information into a single content database. “We took the very labour-intensive step of collecting information and putting it through an intensive process of review, rewriting and editing,” he explains. The hard work paid off as users felt reassured that the information retrieved was not only up to date and relevant, but that their search was not going to miss a vital knowledge source that was hidden away in another part of the organisation.
Of course, large companies often have vast quantities of information in various forms. Organisations are going global and expansion and changing working methods, such as mobile or remote working and multi-lingual environments, are creating new challenges for search and retrieval strategies. Wang believes that today’s organisations need to establish an enterprise-wide search strategy. “You need to make sure that users across the whole organisation can have a consistent search experience and deliver business or country-specific information, as well as enabling information to be accessed across business and organisational boundaries,” she says. Leclercq adds that a deeper knowledge of the type of content present in a company is essential to tailoring an S&R solution. “Not all content is written down or stored in a place where content is retrievable,” she points out. She advises organisations to focus on developing a culture where employees are encouraged not only to generate knowledge, but to store it in a location where it can be easily captured, retrieved, shared and evaluated.
However, by nature we employees are a secretive breed. We want to keep our knowledge to ourselves; sharing is not in our character unless we know we are going to get something in return. When implementing a search and retrieval strategy, management need to take staff attitudes to knowledge sharing into account and make users aware that, in order to source relevant knowledge, they also need to add their own knowledge to the melting pot. “Education is everything,” says Leclercq. “You need to communicate and educate across the business on a continuous basis.” At Aon, issues such as this are addressed systematically through an enterprise-search-strategy group. The formation of a group for this specific purpose may seem a little extreme but, as VanderKnyff points out, “If you can’t find the information, it has no value to you or your organisation. It might as well be locked in someone’s head.” A search and retrieval strategy’s success rate hangs on whether or not an organisation promotes a genuine knowledge-sharing culture.
Once an organisation has been convinced to share knowledge, another cultural challenge rears its head. Users search in very different ways, so a search and retrieval tool needs to understand the search habits and behaviours of its users. This enables the tool that is developed to function more efficiently and retrieve the most relevant information for each particular user. In the S&R jungle, the user is king and each user has their own unique way of hunting for information. Wang believes that a search and retrieval tool must almost be able to read a user’s mind. “It’s extremely difficult to change your user’s behaviour,” she says. “It is far better to try to understand their behavioural patterns and design a solution to accommodate them.” Only an S&R system that is trained to intelligently interpret the user’s query will be able to deliver tailored and relevant matches for that particular user. Khan agrees that S&R solutions need to be designed to accommodate the user’s search habits, such as accepting search queries in natural language. “S&R tools should not overwhelm the users with a whole new syntax to learn,” he maintains.
Ease of use is fundamental to the success of a search and retrieval solution. Khan feels that technologies have evolved exponentially in accordance with user demands, from simple keyword searches to programmes that can actually learn over time. Easy-to-find content is an essential component of a search solution and Leclercq believes that a strong taxonomy is key to intelligent search. However, human input is also desirable in order to be selective. VanderKnyff cites human judgement as the only method of determining which information really meets the business needs of an organisation. “In my estimation, information glut can be just as paralysing as the lack of information,” he says. Dredging through a veritable swamp of information defeats the whole purpose of applying search and retrieval technologies. Taxonomies therefore need to be edited regularly by individuals who are familiar with the system to eliminate irrelevant information.
The human element can be employed as a tracking device to keep S&R technology up to date and relevant. Consistent, expert-led tracking could hold the key to the future of search and retrieval in relation to KM. Users can be casual or expert, frequent or occasional, and job functions are subject to change over time. Intelligence and personalisation of search and its results is where the potential success of S&R lies, in Leclercq’s opinion. Content pre-processing, classification and real-time, expert-led collaboration are the sophisticated S&R features that are currently leading the field. Wang cites ‘real-time query filtering’ as her vision for the future: a query needs to be analysed and routed to the appropriate business rules based on a personal user profile, before the knowledge seeker is offered various retrieval options. She feels that vendors must offer technology that not only considers user profiles, but also accounts for previous search interaction in order to return accurate results and stay ahead in the technology race.
The S&R solutions are out there, but some organisations are still stumbling over outmoded intranets and taxonomies and are unwilling to employ technologies that could meet their S&R needs. The main reason for this reluctance is that measuring the ROI of a search implementation and enhancement project is an arduous task. Organisations that demand a hard-cost/benefit analysis will often find it is difficult to justify investing in technology that meets user needs from a purely financial perspective. However, companies are coming round to the idea of investing in S&R and Wang believes that measuring both tangible and intangible forms of ROI has become commonplace. “Aon use measurements such as user feedback and comments, as well as those such as the reduction in the amount of calls received by the HR benefit hotline to assess the success of a completed search enhancement,” she explains. Khan believes that efficiency, time savings and user compliance are the most important metrics to consider. For VanderKnyff, measurements are not the foremost concern. “In our case, the project was created upon an implicit belief that greater access to organisational information – and a higher degree of accuracy – will help the university cope with increasing workloads and a static level of staff.”
As with any relationship, both parties have to invest time and resources into creating a lasting and worthwhile union. Solution providers and organisations alike need to work collaboratively in order to achieve optimal results when attempting to search for and retrieve a company’s explicit knowledge assets. Knowledge work in general can be unquantifiable in bottom-line terms, but commit to the user and KM and S&R have the potential to be a match made in enterprise heaven.
Shahnawaz Khan, manager – KM solutions, Srishti Software, firstname.lastname@example.org
Michelle Leclercq, global portal manager, Centrica, email@example.com
Rick VanderKnyff, web manager, University of California, San Diego, firstname.lastname@example.org
Annie Wang, principle information architectAon Corporation, email@example.com