Enterprise Search – a 2018 Round-Up

25 Jun

close up of text on wood

By Amy Halverson, Director of Knowledge Management, Research & Information Services, Wilson Sonsini Goodrich & Rosati

More than ten years have passed since enterprise search hit large law firm KM radars in a big way.   Much has changed since then – the technology, the providers, and perhaps most critically, law firm content repositories.  These changes appear to be pushing enterprise search back onto law firms’ to-do lists.  The topic is appearing more frequently in ILTA programming, as illustrated by the recent, and highly recommended, ILTA webinar entitled Enterprise Search Tool Stories, in which three law firms walk through the processes and choices that went into enterprise search upgrades.  The 2018 KM Priorities survey issued annually by Ron Friedmann also shows enterprise search/information governance to be a higher priority for firms than in recent years, with over 45% of large law firm respondents identifying it as a top priority for the year.

Given the above, we thought this would be a good time to look at where things stand now with law firm enterprise search, and to spotlight some of the considerations that arise when upgrading or changing out an enterprise search platform, as opposed to installing one for the first time.  This overview provides a high-level view of the enterprise search landscape, and is not intended to provide an exhaustive list of enterprise search products, providers, or feature sets.  We welcome input in the comment section from readers who have additional information and perspectives to share on the topic.

Types of Search Products and Their Functions

By way of background, the term “enterprise search” in large law firms has broadly been used to describe an internal search application that delivers relevant information to end users from more than one enterprise information repository (such as a document management system, an HR database, and a client-matter system), via a keyword query in a single search box, and that offers users the ability to target specific content within an initial search result using categorical filters, similar to the experience of consumer online shopping platforms.

In talking to peers about their recent experiences with enterprise search projects, Lisa Gianakos of Pillsbury Winthrop Shaw Pittman aptly observed that today there are at least three flavors of search technology products:

  • Search Engines with UIs
  • Search Engines
  • Search UIs

Search engine with UI is your all-in-one tool that provides the technology and the user interface packaged together.  OpenText Decisiv is an example of such a tool that can work out of the box with little customization, although many firms do elect to modify the end-user experience and extend the native search features.  iManage Insight also fits into this category, as it can be used for different DMS’s and has an out-of-the-box UI.

Search engines are the tools that actually power the search, but which have no standard UI or a limited UI and few means to customize.  SharePoint Search (formerly FAST) is one example.

Search UIs are the services that can be used in tandem with the search engine and act as facilitative connectors that deliver the results of the search engine to the user.  SharePoint Search + Handshake is an example of such a combination.  SharePoint powers the search index, while Handshake gives the firm the ability to customize what and how results are delivered and displayed.

Finally, there is the native DMS search.  iManage offers an IDOL-powered search, though with its next release (10.2) it will offer its customers the option of having either IDOL or RAVN-powered search.  NetDocuments relies on Solr, and SharePoint on, well, SharePoint.

Examples of search technology types by product name.

A. Halverson_ table_blog

Requirements

In speaking to KMers who are, or have recently undertaken an overhaul of their firm’s enterprise search systems, I asked what new features their users have really liked or found to be valuable.  Here are a few that you might consider including in a requirements document:

  • Integrated search box that allows users to search intranet content alongside documents and other content repositories
  • Natural language search queries
  • Guided, predictive search that produces easy-to-read top results and pre-programmed best bets
  • Preview view for individual search results
  • Export search results for matters and people to Excel
  • Ability to contextualize intranet content pages by inserting pre-formed search queries that “publish” to the page

If your firm is considering a change or upgrade in its enterprise search system, the first requirement should be that the replacement do at least as much as the current one.  Either track down the requirements that were drawn up for the existing system, or start from scratch.  Does your search allow users to multi-select categories?   Make it a requirement.  Do your users have the ability to preview the source document?  Share or save searches?  Toggle from one data source to another?  Be painstaking; don’t assume anything.   Once you have those requirements listed, build on them to add the new capabilities you want.

Other Considerations

Requirements should extend beyond just the search tool features to include functional needs and dependencies as well.  For instance, if you intend to index intranet html and/or SharePoint pages that include stub pages and non-substantive content, make sure the index can filter out the non-substantive content, because otherwise they will appear in search results and overwhelm legitimate, useful search results.

Other considerations relate to the services provided by the vendor(s) you engage.  When selecting a provider, remember to:

  • Confirm the provider’s level of technical support, and the terms/cost of engaging professional services for improvements
  • For Search UI providers, confirm that the provider offers out-of-the-box templates that are based on experience and feedback received from prior customers. Also confirm that the provider regularly iterates its product based on experience gained from prior installations.
  • Ask about their work process, how do they plan for and implement feature enhancements; what is its typical release cycle. Ask to see the product roadmap and hear how upgrades are deployed.

Should you not have the resources or inclination to undertake the project on your own, your firm can consider engaging a consultant to help you, whether for specific aspects of the project, such as doing user interviews and drafting a resulting requirements definition, to aspects such as technology assessment, solution design and implementation, end-user training, and internal marketing and adoption.

Pre- and Post-Project Tips

Before embarking on a search upgrade, examine the state of your firm’s DMS metadata.  You may want to clean it up before using in a search index (garbage in = garbage out), or may elect to extract metadata  (entity extraction).

After launching your new search, use analytics to assist with adoption and track problems.  Monitor user activity, flag abandoned searches, and follow up with the users to find out what they were trying to find.  This will allow you both to spot functional problems and to do targeted training for users who struggle to use search effectively.  Use reports are also a rich source of clues for what to program as “best bet” search results to particular query strings.

New Flavors

Despite the advances of the past ten years, enterprise search has not changed fundamentally in how it is used – an attorney types in a description of the thing she wants to find.  But in coming years it seems likely that chatbots and voice-activated tools like Alexa, Siri, and Google Assistant will be adapted for the enterprise and used to deliver results of certain types of searches.  In fact, one firm I spoke to, Foley & Lardner LLP, is already developing a chatbot search that is enabled for Alexa voice commands (Charlotte Logullo describes this briefly in the ILTA webinar noted above, Enterprise Search Tool Stories).   So if your firm is not ready to take on an enterprise search overhaul, but does want to improve the efficiency of retrieving certain types of information, a chatbot may be worth consideration.

 

Thank you to Charlotte Logullo of Foley & Lardner LLP, Rick Krzyminski of Baker Donelson, and Lisa Gianakos of Pillsbury Winthrop Shaw Pittman for sharing their valuable insights and experiences with me for this posting, and to Joshua Fireman of Fireman & Co. for contributing to the list of search providers.

Advertisements

One Response to “Enterprise Search – a 2018 Round-Up”

Trackbacks/Pingbacks

  1. Legal tech: a look forwards and backwards - LAC Group | Knowledge & information management services - July 19, 2018

    […] Enterprise Search—a 2018 Round-Up […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: