Last week I presented on “Extracting and capturing knowledge found in social software tools” at the Knowledge Capture and Retention in Law Firms conference. Below are the slides I used.

Synopsis

• how firms are using web2.0
• how to extract structured data from web2.0 tools
• quality control
• indexing, archiving and distributing the information

Law firms are using web.20 tools, but the specific types and purposes vary. Blogs, wikis, rss, tagging and social networking tools are all in use.

I referenced a number of examples of how law firms are using social software. If you care to read the original case studies you can access the links below:

• Dewey & Leboeuf are using wikis, blogs, RSS & social bookmarking – link
• The Rosen Law Firm are using wikis for hosting reports, profile pages, communications – link
• Goodwin Proctor are using wikis to manage projects, and client & legal knowledge – link
• Dykema are using RSS for communication and information distribution – link
• Magic circle firms are targeting team wikis, practice wikis and “Pedias” – link link link
• AM Law 200 firms using blogs – link

With respect to extracting knowledge from web2.0 tools, the most important thing to remember is “garbage in – garbage out”. You need the participation of many, while recognising lawyers are unlikely to contribute willingly and that form-filling and wiki-markup is too difficult. I propose that in order to extract knowledge, systems need to be simple and that you need to avail yourself of:

• Tagging / categories
• Favourites
• Social ratings
• Usage stats
• Enrichment
• Data analysis
• Social platforms

In terms of quality control, you need to monitor content, use “track changes” functionality, and watch usage. You also need to visibly differentiate high-quality content (PSL or Partner approved), and make sure that the names of contributors are highly visible.

When it comes to indexing and archiving, the focus for any firm should be high-value / low volume content. Firm’s will never have sufficient resource to index everything, and, much as it may pain them to do so, firm’s will have to prioritise. In order to do so, usage stats are quite helpful in determining what content people find useful. Just as importantly, firms need a process to retire knowledge to avoid poluting their search engines and intranets with irrelevant content.

Finally, I talked about distribution, leveraging a full range channels, including:

• email
• phone
• rss
• newsletters
• meetings

In putting the presentation together I accessed a number of websites, some relevent, some not. For my convenience I used Delicious to keep track of them. You might have a look if you’re interested in learning more.

I thought I might post a follow-on to my prior post on the role of trust and how it affects wiki adoption.

One of the most valuable lessons one learns when working with lawyers is to be your own biggest critic, especially when it comes to presenting an idea. I believe this comes from knowing that if you don’t break down your arguments, consider them from many angles and adjust them accordingly; other lawyers will tear your ideas to shreds.

Personally I find that approach leads one to be overly conservative when it comes to sharing thoughts. I prefer to throw ideas against the wall, see what sticks and adjust them over time given feedback & discussion. I much prefer a hypothesise, test, measure, adjust & repeat approach as the idea improves with each iteration.

Regardless, this blog doesn’t invite quite the level of debate capable of refining my idea, so I’ve had to do it myself.

First off, my representation of risk being flat was incomplete and requires amendment. Traditionally there are several different approaches to managing risk, some of which can change the level of risk:

• Reduction (methods which reduce the severity of the risk >> risk decreases)
• Avoidance (not engaging in the activity which carries risk >> risk remains flat)
• Retention (accepting the risk, a viable strategy where the costs of reducing the risk are greater than if something went wrong >> risk remains flat)
• Transference (transfer the cost of the risk to another party). This is irrelevant for this discussion.

Of course risk can also increase over time depending on the circumstances. My earlier model ignored changes in the level of risk.

So the trust-threshold model should consider risk & trust curves as capable of changing over time. The model below is not prescriptive; but demonstrates an example of how risk & trust might interact when a system update is deployed.

So, with the model updated, let’s look at some objections to the model.

Objection #1 > the value of wikis is self-evident; lawyers who know enough about wikis will use them. Trust thresholds have nothing to do with it.

While I have met very few lawyers willing to use wikis from first viewing, those lawyers do exist, and trust thresholds apply to them. The lawyers in question just started with a level of trust higher than the level of risk. Errors introduced into the system or uncovered during the course of everyday use could impact trust in a negative way.

The key here is that trust is not constant and is affected by the lawyers’ environment. Imagine a negative story about wikis in The Lawyer. Trust might well drop (as some unmanaged risk comes to light) followed by usage.

Objection #2 > this has nothing to do with trust, wikis are just a bad idea

Wikis being a “bad idea” is a purely emotive comment. My example of using wikis to track practice-group utilisation demonstrates their utility. One their utility is demonstrable; people need to decide whether they’re the most effective tools available.

I have seen several firms with gaps in their systems infrastructure where the read-write nature of wikis delivers a tool capable of supporting business needs. At any one time most IT departments will have more projects than they can delivery. In these circumstances wikis can deliver immediate value while a long-term solution is put in place.

An inexpensive, flexible tool capable of delivering value across the firm is not what I’d call a “bad idea”.

Objection #3 > no level of trust will get me using wikis at work.

Hey, I never said wikis were the right tool for everything. It’s entirely reasonable not to use a wiki if you believe the alternatives are better. Wikis would be a poor alternative to a proper practice management system.

Objection #4 > Your graph ignores cost-benefit analysis (feedback from Papa Richards)

Not really, for the same reasons as above. The Cost-benefit question is answered beforehand (like when your firm buys the technology). Risk might enter into the equation at that point, but that’s institutional risk.

I’d love to hear the thoughts of others on this.

I was on my morning run this morning, with my brain in la-la land, oddly thinking about Doug Cornelius’ Wiki While you Work article, when I had a brainwave (well I thought it was a brainwave) about what it would take for lawyers to use social software.

“Trust thresholds! It’s not just about trust; it’s also about risk of each task. You need to earn enough trust to overcome the risk. Therefore, wiki use is dependant on how much a user trusts the content.”

Those weren’t the exact words, but close enough. I almost stopped running, and would have but I had a long way to go and once you stop its murder to get started again. I kept running and diverted my brain onto more simple topics like “why do people throw shopping trolleys in the canal?”

What had popped into my brain was the idea that when using a wiki to execute a task which involves risk, the trustworthiness of the wiki needs to exceed the risk of it being wrong.

First, a story.

My first experience of social software & lawyers was unsuccessful. I was looking to put in place a Wikipedia-style resource for my firm, and met with substantial resistance, and downright refusal to use it. Lawyers being lawyers a number of risks were listed. NOT IN MY PRACTICE! Case closed.

Later I found some practices were using the wiki, but for purposes I had not intended. For example, one practice used the wiki to share information about work allocation. Junior lawyers would edit the wiki page indicating how much time they had spare, and the type of work they wanted.

Until today, I didn’t fully understand why, and I hope the following picture helps to illustrate my point.

The “trust threshold” is the point where risk & trust meet.

My initial proposal was that lawyers use the wiki to execute client work. Trust was low while risk was high. No dice.

However, using the wiki to find more work, the risk was low, and trust was “high enough” to have passed the trust threshold.

A couple more key points:
* The “trust curve” I’ve drawn is not prescriptive. Trust can be lost, and once it falls below the trust threshold, the risk becomes too high.
* Trust curves and the level of perceived risk are unique for each person.
* Usefulness is not a factor that can overcome risk.  It might be enough to get a lawyer to experiment with something less risky, which in turn builds trust.

Ways of increasing trust
I do not think you can change the risk associated with lawyers’ tasks, because cost of failure for a lawyer approaches infinity.

I do think you can increase trust, but it will take time, and some people might never make it.

Some suggestions:
* other lawyers have to use the wiki first (i.e. PSLs)
* the system has to signpost security (i.e. no anonymous users)
* show the lawyers how to recover something that was deleted
* talk to the lawyers about the ways incorrect content can be drafted
* PSL / Partner stamp of approval on a particular page / wiki
* Limit access to a set number of users

So what?

What I draw from this is that:
* Big bang wiki deployments are more likely to fail than not
* You will have to start slowly
* You will have to find something safe
* You may well
* Keep the scope of a wiki low (i.e. single topic, experts only) if you need high levels of trust
* You can measure & aggregate trust & risk by asking people. Probably worth establishing where you are on the scale before you try to wikify your firm.

Credit where credit’s due

Before today, I never read about or heard of trust thresholds. However, in checking Mr. Google, somebody has written about them. In his book Trust in the Balance, Robert Shaw defines the term as the point where trust becomes distrust (I can’t get anything more than an Amazon preview unfortunately).

However, I see no pictures or graphs online, so that baby’s mine! ©

Update: I’ve refined the model a bit in my follow-on post Do trust thresholds exist?

The good folks over at Headshift have launched a new site which hosts a collection of case studies and use cases for Enterprise2.0 tools, including a special section for legal.

If you’re really keen, you can even track their new case studies and use cases as they add them to delicious.

An interesting little post by Toby Ward on Enterprise intranet wikis which carries the same theme as Matthew’s presentation at KM Legal.

Toby shares one story about the usage of wikis at Cisco:

“There are tens-of-thousands of wikis… and the number of wikis about equals the number of employees (65,000)”

I can’t imagine how 2000 wikis in a 2000 person firm would make any sense at all. I struggle to see how it would make sense for Cisco.

He also quotes analyst Janus Boye, who has recently released a “Wiki in the Enterprise” report:

“Wikis often grow out of hand very quickly and consequently many employees simply ignore them. Enterprises also face the risk of an explosive information growth far beyond their capacity to manage that information”

Ooops.

So, some thoughts from an Intranet expert which are mildly concerning. However, he does mention a succesful ThoughtFarmer implementation (circa 2006) where a project manager was able to share a cost-saving technique with international colleagues to the tune of $500k. Not bad, but again:

“Currently, about 50% of Placemaking employees do not edit the intranet. 44% are occasional editors, and 6% are active participants. Although participation is high compared to internet-based social systems, Intrawest is still hoping to improve and look at new ways to get the 94% of low- or non-participators more involved.”

The usage is pretty high, but how much time should lawyers contribute? Sounds like more evidence supporting a case for careful & considered use of wiki platforms and not just for law firms.

More thoughts to come…

Slides from Matthew Parsons’ talk about wikis and blogs at law firms at the Knowledge Management for the Legal Profession Conference.

[kml_flashembed movie="http://static.slideshare.net/swf/ssplayer2.swf?doc=web-20-1208989616266544-9" width="425" height="355" wmode="transparent" /]

Also, over the course of Matthew’s presentation, a number of tools were mentioned. Their sites are linked below for your convenience:

• Slideshare
• Youtube
• Meebo
• Confluence
• SocialText
• Sharepoint
• Wet Paint
• Mediawiki
  

If I’ve forgotten anything, please let me know. Also included is a link to the Tim O’Reilly definition of Web2.0.

One of the things I’ve found about blogging is that it’s more interesting to write original content as oppposed to just linking to articles.

However, over at KM Space Doug Cornelius has uncovered a tidbit from Mark Miller which is too good to ignore, which can help you to determine when blogs or wikis are appropriate for you.

Simple test: one or two people providing content, use a blog; many people providing content, use a wiki.

Genius!

BinaryLaw, a blog I visit regularly have linked to a new Lexis site, CompanyLawForum.co.uk.  

The site allows you to publish opinions, message other users, ask questions within a discussion forum, and  comment upon & rate news items. 

Initially I was going to recommend you visit the site.  However, having gone through the bits, I think it would be a waste of your time (for the moment).  As I worked my way through the interface, I found it very difficult to use and poorly designed.  For example, you can’t click on article titles, instead, you have to click a “View/Add Comments” link which is stuck at the bottom of a post.  Similarly, once you’re in an article, you can’t get back to where you were.  The whole thing feels sloppy. 

Still I’d keep an eye on it.  I’ve provided an overview of each section underneath the screenshot below. 

Opinions Section
Oddly, the section does not show user ratings of opinions, which is inconsistent with the rest of the site (ratings can be seen once you click through to a single opinion).  Also, the site is already receiving commercial spam from India which Lexis should take down.  Having content framed on three sites by makes the site feel constrained and unfriendly. 

News
Similar comments, difficult user interface, but at least user ratings show up on the first page. 

Forum
They’re using standard forum tool so the interface is a little better.  A lot of people seem to be viewing content but there’s not much participation.  Again, spam from India in the forum posts. 

Companies Act
Just links to a LexisNexis login and some dead links.  The spelling and grammar mistakes make me think this is not ready for the big time.

Members Directory
If you’re not logged in you get a “You are not authorized to view this page!” message. 

Several weeks ago I wrote on why your organisation should support Wikipedia.  At that time
26,396 people had donated in the thirty five days since the fundraising effort started (October 22)

Only 14,636 (roughly half) have donated in the thirty one days since my first article. 

The donation drive lasts for only five more days (ends January 3, 2008). 

Help them help you.

Richard Wallis from Talis has just finished a talk on web2.0 search

The theme focused on adding semantic knowledge to baseline search.

Using examples from Google, webcat and Youtube, he makes the point that Librarians are very good at classifying discrete items. Too good in fact, and can go too far (check WorldCat.org for Philosopher’s Stone), while the public are really bad at it (try IT Professional in Youtube).

His view is that using semantic tools can help improve relevance for end punters, but that the tools won’t move into the mainstream until around 2010.

If I can get my hands on his slides, I’ll link them later, but in the meantime you might checkout the Talis website.

Update: Chatted to Richard today and he gave me the link to WorldCat.org.

Ta Richard