Global jigsaw: Developing a global search solution

Head of knowledge Gerard Bredenoord and consultant Cora Newell discuss the evolution of Linklaters’ global search tool

 

Key takeaway points

1. Be clear on your business drivers/critical issues and ensure they will be addressed by your search project. Ours were multiple search platforms, inconsistently categorised knowledge and lack of common processes.

2. Using a prototype in the early stages can prove whether your investment in search technologies and the availability of curated approved knowledge can be leveraged effectively.

3. Quit while you are ahead – avoid endless tweaking of the prototype and concentrate on developing a successful business case and managing delivery. Establishing visible practitioner value should secure buy-in and support.

4. Your governance model needs to involve representatives of all essential stakeholder groups. Be sure to adequately support individuals who take on extra burdens to ensure they can deliver.

5. Agreeing the scope of the project is of paramount importance, as well as being clear upfront on what you are and are not doing. It may be better to manage the introduction of the search project in a careful and controlled way and expand the boundaries only once you are familiar with what your search solution can offer.

 

Search solutions in global law firms are critical in enabling all members to easily and quickly locate legal knowledge, freeing up time for client work. Giving staff improved access to a firm’s legal knowledge can also develop confidence in the quality of information returned and increase the level of self-service.

The seeds of Linklaters’ global search solution were sown back in 2007 when the firm’s information architecture strategy was created.

The strategic objectives of the information architecture were ambitious – the integration of all technical content, interfaces and access points into one single platform. They have led to the development of a number of significant IT initiatives, including the launch of the firm’s wiki-based global intranet in July 2009, InSite.

Following the launch of Linklaters’ new global search solution ‘Se@rch’ in November 2011, the information architecture jigsaw will be nearly complete, allowing the firm’s evolving search strategy to take centre stage in future IT developments.

In this article, we tell the story of the development of Linklaters’ global search solution, how it came about and the cogent business reasons for it.

Need for change

The predecessor to Se@rch, Linklaters’ former main repository for legal knowledge, is Knowledge Index (KI), a bespoke system developed in-house in 2005. Covering the firm’s ‘crown jewels’ of precedents, document specimens and deals, KI’s limitations became more obvious over time, particularly given improvements in enterprise search technology.

Neither full-text nor properly global, KI operated on an exclusive metadata basis; all underlying documents were classified into specified fields. Global one-stop search was not possible: users needed to specify both the region and practice area and repeat searches to span numerous regions and practice areas.

In a web environment where users have grown accustomed to the minimalist yet powerful Google approach of freestyle full-text global searching, these limitations became increasingly difficult to swallow.

Users wishing to contribute legal knowledge were also discouraged because of the time-consuming process of completing numerous metadata fields.

Further critical issues were inconsistently-categorised knowledge, the lack of a common process for identifying and uploading legal knowledge, and multiple global search platforms from which to locate this knowledge.

Yet despite the limitations of the knowledge solutions, the firm’s single global desktop, intranet and document management platforms provided a powerful base on which to build.

It became clear from the start that any search solution redesign had to put the needs and desires of Linklaters’ lawyers and other end users first.

The successful launch of the intranet – which had its foundations in wikis, blogs and streamlined searching for global content – helped to stimulate the cultural appetite within the firm for a global legal knowledge search solution. The intranet’s single global home page also showed users the advantages international integration could bring.

The dual aspirations of leveraging the firm’s global legal knowledge resources and skilling up global staff to a consistently high standard became irresistible.

Prototype testing

The search prototype project began in March 2010 to determine which technology solutions were suitable and to seek the firm’s feedback for different interface approaches.

By May 2010, the search prototype steering group decided on a user interface from Okana called Sense. When combined with Autonomy’s meaning-based computing functionality, Sense was considered the right choice. Its flexible user interface caters for multilingual searches and contributions.

A prototype front-end search programme was developed using a pure white screen approach, keeping lawyers’ desires foremost. Key considerations were Google-like freestyle and full-text simplicity, combined with some metadata classifications but keeping screening filters, structure and internal customisation to a minimum.

Social media tools like cloud tags and searches which could be saved and shared were included. Users could also subscribe to RSS feeds and updates.

Rigorous and extensive pilot testing followed. By October 2010, it became clear that the anticipated deliverables were well within reach. A decision was made that there was little benefit to be gained from tweaking the prototype further.

The time had come to concentrate on developing a successful business case and managing delivery of the project to the global partnership. The question was, how could the global search solution get buy-in and support from partners and senior management during the economic downturn?

A small group representing IT, knowledge and learning, and key senior partners was established to put together the business case. It was speedily accepted by the firm’s investment committee due, in large part, to the visible practitioner value evidenced by the search prototype process. Se@rch officially kicked off on 10 January 2011.

Governance model

Consistent with Linklaters’ usual in-house project methodology, a governance model was created to support the delivery of Se@rch. In addition to the firm’s existing knowledge and learning committee (KLC), which is populated by senior and junior members from the firm’s three regions and all practice areas, a search steering group was formed to take responsibility for the project’s management.

This new group, which would report to the firm’s investment committee, included director of information systems and strategy Sue Hall as executive sponsor, global head of knowledge and learning  Suzanne Fine as business sponsor, and Gerard as secretariat.

Sue and Suzanne were already members of the KLC. This close overlap of members of the KLC and search steering group – the two groups also share one regional and all three divisional partners – ensured integration with the practice and provided continuity and focus.

A search project team was also formed under Gerard’s leadership, reporting to the search steering group. Joining Gerard was Se@rch project manager Richard Armitage, an information architect, a few IT technical specialists, a business change and communications specialist, and two individuals who could handle the content migration issues given their KI experience, legal knowledge expert Alexandra Cabannes, a PSL in the Luxembourg office and information officer James Pilgrim.

The search steering group members met formally once a month to discuss substantive policy issues. Along with the search project team, they kept in constant touch to smooth over the many difficult nitty-gritty decisions and the usual disparate and strongly-held views which naturally predominate in a global partnership.

Richard led a small full-time team of about ten members, but there were around a hundred others who contributed to the project in some shape or form, an approach consistent with the firm’s matrix operational style.

It is unusual for resources to be extensively deployed to single projects on a dedicated full-time basis, as the firm believes that project teams built this way result in wholes greater than the sum of their individual parts.

Critical challenges

One of Richard’s major early challenges was agreeing the scope of the project – defining the meaning of legal knowledge. Settling on an agreed definition of legal knowledge proved to be no easy task and put into relative shade the technology issues, which turned out to be more straightforward. The global nature of Se@rch was the major complication – each of the firm’s offices held varying views on what constitutes legal knowledge.

Agreement was finally reached by persuading stakeholders that the initial boundaries of legal knowledge within the scope of Se@rch would be expanded in future when the timing was right.

The powerful nature of the search tool has meant that the groups have had to manage its introduction in a careful and controlled way, letting its usage evolve gradually as lawyers become more familiar with what it can offer.

It is for this reason that Se@rch will initially be restricted solely to legal knowledge, although it is expected that the additional step of integrating the intranet search facility will take place in the near future. 

Another challenge has been the significant focus on content. Ensuring the firm’s legal knowledge is properly identified, updated and classified has been highly labour-intensive.

The different metadata which was decided upon for classifying the firm’s legal knowledge under Se@rch has involved significant fresh mapping work. To relieve PSLs and information officers of some of this burden, the firm recruited additional external help.

However, some of the extra burden on the firm’s PSLs and information officers could not be alleviated by the hiring of external resources. The secondment of Alexandra Cabannes and James Pilgrim to the search project team has been key to achieving the support of these groups.

The size of the task involved was considerable. A minimum of 170,000 records needed converting from KI to Se@rch’s new metadata classification system. But, given that one KI record could relate to multiple documents, it was a massive undertaking.

KI’s notional global basis meant that foreign-language documents required translation, as they could only be described and searched for in English. The multilingual nature of Se@rch means that all documents can be entered and searched against in their original language. English metadata has been added to enable users to benefit from a consistent set of English filters.

Strategic priorities

A main priority has been adopting a business-as-usual approach to Se@rch, looking beyond the rollout date. Appropriate change management support and awareness training will be needed. Search steering group members must also become visible and vocal champions to drive the Se@rch solution.

However, chasing return on investment is definitely not a priority. A focus on quantitative measurement would fundamentally undercut the strategic intent behind Se@rch, ready access to the firm’s global intellectual capital and the creation of confidence in the firm’s search system.

If staff do not have the perception they can find what they are looking for when using Se@rch, it will have failed. But there will be some housekeeping search administrator tools built in to find out which legal knowledge documents are heavily used or, perhaps, not at all. This will help the firm to properly develop its resources.

 The complicated choreography of Se@rch and the skilful interweaving of its four main workstreams – project management, technical solutions, content management and business change, and communications – has finally reached its exciting go-live and key project deliverables.

Our follow-on article next year will discuss the lessons learned from the implementation process and the global impact on the firm.

Developing a new search tool

Do

• Be clear upfront on what you are doing and not doing – define your strategic objectives.

• Be realistic – you need to be able to support your strategic objectives.

• Stagger your key project deliverables if necessary.

• Secure agreement and representation from all stakeholder communities.

• Give proper support so that everyone feels a genuine part of the team.

Don’t

• Overreach and take on too much at once.

• Think short term – bear in mind the evolving strategic roadmap.

• Attempt to rollout everything at once – this dilutes impact and loses momentum.

• Ignore difficult questions and issues raised – if you don’t have an answer, say so and come back when you do.

• Let an overconcentration on return on investment dilute your strategic impact.

 

gerard.bredernoord@linklaters.com; newell@kminsightconsulting.com