Experts are Going, Going …

Problem Statement

Donald Travis is the company’s go-to expert on equipment and systems that were installed over the past 30 years. He knows the ‘why’ behind all of the odd configurations, procedures and workarounds. When a weird problem surfaces, he can recognize it instantly and fix it effortlessly. He’s seen similar problems 4 or 5 times over the years. His 30-something protoges might struggle for weeks before they figure out what he already knows. There’s just one problem. Donald retires in 4 months. He’s building a cabin on a lake and has no interest in contract work. Once he’s gone, the next weird problem might shut you down. What do you do?

The Brain Drain Challenge

The scenario with Mr. Travis is occuring against a backdrop of concern about the real threat of baby boomer
retirements and the resulting brain drain. A long list of
reports, studies and surveys agree that a large portion of America’s
skilled workforce is likely to retire in the next 5 to 10 years -
taking important knowledge and expertise with it. At the time of writing, a Google search with’brain drain’ and ‘baby boom’ as search terms gets almost 30,000 hits.

Surveys and studies indicate that a lot of companies, organizations, and industries are very concerned and are currently picking
through a catalog of policy ideas to deal with it.

  • Ernst & Young (2007)* – Survey of HR managers found 68% saw retaining key talent was an important concern.
  • Aviation Week (2007) – Aerospace and defense industries will be able to replace half of retirements before 2010.

* requires free registration

So far, corporate responses to the threat seem to be concentrated on incremental responses that
include:

  • extending the mandatory retirement age
  • creating new employee categories (post-retirement workers, contract managers, alumni mentors, etc.)
  • expanding mentoring programs for younger managers

With these initiatives, the worry (or at least the hysteria) seems to be abating
slightly. Some people suggest that it may turn out to be a bit like Y2K. The
threat is very real but if companies apply enough incremental remedies, they will somehow muddle through. We’ll probably have
to wait and see.

The Individual Departure Challenge

While demographic, society-wide retirement looms as the big threat, key experts are being lost every day for comparatively mundane
reasons. Some would retire under normal conditions. Some are promoted or transferred. Some are reassigned to other duties. Some quit to take other jobs. Sadly, some individuals will get sick, injured or pass
away suddenly. Although these are normal, individual-level losses, they can still leave a big intellectual hole. Handling these routine,
individual losses poses an ongoing challenge and it would make sense to look for generic tools to elicit critical knowledge from departing people and store it where it can be accessed by others. Of course, in a competitive corporate world, the method must also be fast, easy and cost-effective.

Unfortunately, cost-effective elicitation techniques have proven elusive and are non-trivial to implement. A lot of the difficulty stems from the fact that the most valuable knowledge is often the expert’s tacit knowledge. They may not fully understand what they know or even how they subconsciously apply it. This makes it a lot harder for organizations to know whose knowledge needs to be captured. Finally, to make things worse, departing individuals are generally short-timers. They may be perfectly willing to leave their knowledge, but they won’t want to fight to do it. You ask them to write it all in a detailed manual or report, don’t be surprised if it isn’t finished before they leave.

With the inherent difficulty of the task, credible knowledge capture
methodologies are still fairly scarce. It is much easier to do a survey and write
a hand-wringing article than to devise a practical knowledge capture solution. The methodology offered by Larry Todd Wilson and his colleagues
at Knowledge Harvesting Inc. (KHi). is the most complete and comprehensive approach that we have seen. This approach has steps to identify, isolate, capture, contextualize and organize
critical ‘implicit’ knowledge from domain experts.

The following diagram (linked directly from the KHi website) depicts the comprehensive nature of their approach.

Several aspects of the kHi process illustrate the magnitude of the challenge. The first task is to locate and identify the high-value individuals that are at risk and have knowledge that is worth preserving. In a large organization, these people are typically scattered across the operations and it is not trivial to do an accurate inventory. The second challenge is to devise a way to identify and assemble all of the supporting documents and information. An expert’s opinion won’t help if you don’t know what the opinion refers to. The next step is to create a setting for the elicitation that is maximally conducive to effective debriefing and to locate collateral domain experts who can help frame the most revealing questions. KHi has published a marvelous “10-page guide to Eliciting Implicit Knowledge” that provides a detailed roadmap to this process. The extracted knowledge must be packaged and organized for use and that application must be carefully planned, organized and implemented. Finally, the entire process must be monitored, evaluated and adjusted to make sure that the steps stay synchronized and relevant.

iPOV’s activities center more narrowly on the elicitation, packaging and distribution activities. iPOV is continuously exploring new ways to apply informal Flash/web video to achieve better recording and faster application of the elicited knowledge. We believe that our methods complement the KHi approach because there are categories of knowledge for which a video-intensive solution may or may not be ideal. If the knowledge can be expressed in a dynamic, visual form (e.g., it involves physical activity, manipulation of physical objects, dynamic performance of systems, or has a complex emotional content), it is generally easier to show someone than to craft a written explanation. If the knowledge is mainly descriptive, it is usually easier to read an account, look at a diagram or read a list than to watch someone try to explain it.

The iPOV Expert Interview Process

iPOV’s approach to eliciting visual knowledge, is roughly similar to the KHi methodology, with three major exceptions:

  • Some people are intimidated by the use of video. iPOV has to take special care to make the sessions as relaxed, forgiving and comfortable as possible. In particular, we never want to ask an expert to ‘do another take’.
  • The tools and workflows for processing video are very different from the ones that are applied to text and image materials. iPOV draws its processing from its standard video transformations.
  • iPOV republishes the resulting video in a Flash/Javascript package to make it searchable with a Google-like search metaphor. Since video is inherently linear, support for random-access search is essential – otherwise the end-user could spend hours looking for a specific nugget of information. In other words, they would never use it.

For the first point, iPOV follows a simple imperative: Make the expert comfortable. If you want to draw information from the expert, it makes sense to do it with as painlessly as possible. If necessary, iPOV can do additional post-processing to correct for any deficiency that this policy creates. Specific techniques include:

  • Video-record everything.
  • Give the expert the tools and references that they use every day in their job.
  • Work with the expert to follow a loose outline, but don’t try to control what they say.
  • If possible, include someone that can ask sincere, interested questions to draw out tacit assumptions and hidden details.
  • Give the expert plenty of time to develop their thoughts. Tape is cheap and can always be cut away later.
  • Allow the expert can digress to whatever they think its important.

It is surprisingly easy to satisfy these
requirements if you stage the capture session correctly. The key is to set the expert up properly, turn them loose and keep the camera rolling. Once the video is captured, iPOV has a wide range of options for post-processing using its standard transformations. Finally, iPOV has refined a special ‘expert interview’ media packaging that makes the material more useful to potential end-users. Since the iPOV process is narrower in scope, it works easily with isolated individuals and targeted problems. iPOV is eager to work with expert
collaborators (e.g., Aerie Engineering and AvatarMS) to capture knowledge from high-value individuals.

The iPOV Expert Interview Web Package

iPOV’s Expert Interview web packaging actually makes it easier to stage the capture session to glean an expert’s knowledge. The key is the embedded search capability. It uses a filtered search to make weakly structured knowledge easily findable on your Intranet. You can enter a search term and receive a Google-like list of search hits. You can skip from one hit to the next and immediately replay the session at those points. Like any search engine, it lets the end-user decide which results are relevant to their question. This, in turn, means that the expert can be more relaxed and natural during the recording session. The expert can digress, revisit an earlier topic, or even correct themselves and it will have limited impact on the end-user experience. The iPOV packaging system has three standard components and several optional one:

Standard Features:

  • Flash web audio/video record of an interview session with a collection of video clips and a synchronized grammar-corrected transcript.
  • A web interface with embedded search engine. Entering a search term filters the transcript for hits that contain the search term. Clicking on a hit repositions to the correct video clip and plays the video at the corresponding point.
  • iPOV can act as a safeguard for the expert. iPOV normally shares the session transcript privately with the expert before sending it to the organization. The expert can make corrections or minor deletions in the event that they said or did something that later makes them feel uncomfortable.

Optional Features:

  • Hidden, searchable metatags can be added to the script to increase the search accuracy.
  • A list of important search terms can be identified and presented to the user in a browsable directory listing.
  • Individual interviews can be aggregated in a broader technical knowledge portal. This portal adds a search capability that spans multiple expert interviews.

The following examples illustrate typical result from an Expert Interview session. The search field is in the header and the verbatim transcript is partially visible in the the navigation menu on the left of the page. The
search tool filters the transcript to present only the segments that contain the query term. The user can jump immediately to topics of
interest – even in a recorded session that is many hours long.



Lean Manufacturing


Fund Raising


Engineered Couplings
Dr. J.T. Black has written six books on lean manufacturing and the Toyota Production System. He has a direct style that cuts through the hype and clutter and makes important points in a simple way. It’s not just what he says, but how he says it – and how he thinks. Carole Ann Fowler led the initial fundraising that built an art gallery/museum in a university town. Listen to her description of the way she protects the donors’ interests. If you had to suddenly assume her responsibilities, it would make a world of difference to understand the style and approach that donors had come to expect. Glenn Pockrandt was a design engineer for shaft couplings for 30+ years. Listen to his discussion in the last clip in the series. He talks about a nasty problem where variable frequency motors may, under certain types of load, run a coupling exactly at it’s natural frequency – and tear it apart. If that obscure possibility doesn’t occur to you, it will be impossible to figure out why the equipment keeps breaking.

In each of these examples the video recording captured knowledge that would have been difficult or impossible to record with any other method. In the sessions on lean manufacturing and fund raising, the presenter’s manner and attitude (style if you will) is an important part of their message. In the sessions on lean manufacturing and engineered couplings, the expert is able to quickly explain a detailed technical point using a combination of simple diagrams and human lecture. For most technical experts, this is a very natural way to explain this sort of thing.

iPOV’s Expert Interview packaging service is fast and affordable. It is really just a stripped-down version of our restatement transformation. The processing cost starts at $1750/hr of finished video for 1 or 2 hour sessions, but can fall under $900/hr for large-volume projects. If you consider the lower demands for expert preparation and the ability of this format to forgive most recording session errors, iPOV’s processing cost largely pays for itself from the savings in expert time and staff preparation.

Bottom Line

At this moment in history, the need to capture expert knowledge is both high profile and timely. However, the need never really goes away. Even if organizations were not facing an imminent demographic threat, they would need to find better ways to identify and retain their core organizational knowledge. While other organizations explore the issue on an enterprise scale, iPOV has chosen to focus on the narrow, but promising, opportunities that we are finding in web-based video and internet search technologies. We see that as a useful and complementary contribution to any larger problem solution.

Print Friendly
Share
This entry was posted in Applications
Bookmark the permalink.
Post a comment or leave a trackback: Trackback URL.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre lang="" line="" escaped="" highlight="">