Enterprise Search – part 2 April 20, 2015

https://www.flickr.com/photos/40935731@N00/304135268

Our recent post about our latest Digital Implementers Group workshop, in which we discussed Enterprise Search technologies, has generated a fair bit of discussion. I thought it might be a good opportunity to delve into this a bit more, as well as discuss the role of the digital implementers group.

The Digital Implementers Group

The Digital Implementers Group is made up of diverse group of public sector information and records specialists who get together every six weeks or so to discuss implementation challenges in their information and records management systems. It used to be called the EDRMS implementers group but the group agreed last year that they do much more than implement EDRMS and so we renamed the group. The topics which the group discuss are driven by the group: things like the organisational challenges they are facing, the technologies they are considering, the risks which are emerging.

The blog posts we put up about the group (and our blog posts in general) are not “Official State Records guidance”, although we like to think they contain a lot of useful information. What they do is try to provide a summary of what the group (which contains a reasonably representative cross-section of the NSW Public Sector) is currently thinking in relation to a particular issue. Not all those who might want to can attend our workshops, so the blog is a good chance to share what went on with the sector.

Enterprise Search technologies

For our purposes, we have defined enterprise search as technologies which are able to index content across a diverse set of information repositories within an organisation, and present search results to these repositories in a single interface. This covers quite a broad range of tools, including some which are now being integrated with popular EDRMS / ECM solutions.

We have had quite a few questions come through about the workshop. Some people questioned how we had come to these conclusions about what the opportunity for enterprise search was or was not. Someone even wrote a whole article “debunking” some of the things in the blog post. This is an important discussion topic to continue. Here are some further observations from the workshop and other recent agency discussions:

Enterprise search is a big investment

The budgets of public sector organisations are under pressure. They are seeking to demonstrate value be doing more with the tools they have, rather than spending a lot of money on a blockbuster technology. The cost versus the value of enterprise search is something which will become clearer as the public sector market for these products matures.

How people expected to use it

Our implementers have said that in their experience a well presented and intuitive user interface which provides a dashboard view relevant to the user is much more important – as a default, people probably don’t event want to look to get what they need, they just expect to be presented with a meaningful view of their work processes and/or their current documents.

On the other side of this, some people think that users will increasingly expect powerful search functionality, even if they use it infrequently. Public Sector colleagues in ICT have told us that they think advanced search functionality  should be available as core function in information management tools, rather than a gold-plated extra.

Is enterprise search like searching the internet?

People often expect enterprise search to be like Google. Google is very powerful at finding information about a topic, and so it is often used as a yardstick for user experience and accuracy of search. However, searching for information on the internet and within your organisation are two very different things. The internet search draws out relevant content from an overwhelmingly vast sea of unrelated information. A search within an organisation has to distinguish specific pieces of information from a very large amount of relevant and related information.

In response to the previous blog post, someone emailed the following observation about the main differences between internal and web searches in their experience:

  • Search Engine Optimisation teams competing to make content more relevant
  • Smaller volume of click throughs to determine the appropriate content
  • Search users tend to have a specific result in mind rather searching generally for information

What this means is that even powerful search tools operating across a repository of rich information may not surface the right document in the first few results, which means its usage is going to be either for a specialist role or an occasional one. As someone else recently said, “how often do you look at the second page of Google results?” By contrast, power users may construct complicated faceted search queries using these tools in order to find the information they are looking for.

What’s next

We have focused a lot of effort recently into helping organisations consider how managing records and information better will demonstrate value to its core business. Nonetheless, many organisations still want to know whether a particular tool or capability will assist in meeting obligations under the State Records Act.

The answer is that public offices meet their records management obligations in a variety of ways. Public sector organisations have extremely diverse core business processes, and these impose various requirements. As far as we can tell from our research and engagements, enterprise search technologies can be a major component in a project to clean up chaotic, poorly managed digital information, but is only one tool with a more modest role in business-as-usual, ongoing operations.

We are going to continue to research and develop our understanding of these technologies and their implementation, so keep an eye out here.

 

Picture: Gabriel White, used under a CC BY-SA 2.0 license.

 

 

 

One Comments
Greg Forsythe April 22nd, 2015

This is a good follow up to the first article, and a lot of valid points are made in both. Working for an enterprise search company, I can say that delivering the most relevant content is a top priority. The comment made about relevancy improvement internally is accurate. An organization needs to develop their collections toward a common need or goal. Most of our customers who really need search understand this aspect and have systems of classification already in place.
Companies who need to sort through millions of documents about standards, laws, and research have a much different search experience than those who only deal with a few limited sets of data. The key in any successful enterprise search installation, is knowing ahead of time, what you need to do with the information.

Leave a Reply

You must be logged in to post a comment.