Recognizing the Problem

In the late 1990s, TVA of Knoxville, Tenn., America's
biggest public power producer, was facing the retirement
of around a third of its workforce, along with the loss of
much intellectual capital, some of it irreplaceable. Many
of the employees who would be leaving literally built the
plants they were operating. If their knowledge left with
them, then it was too late.

To make matters worse, for other reasons TVA was particularly
vulnerable to the demographic time bomb that was ticking.
Because of massive layoffs and attrition, the workforce had
fallen from around 50,000 in the early 1980s to a little
over 12,000. Decades earlier it had ended its internal training
programs. It had laid off younger, less experienced personnel.
And it had discontinued its recruiting relationships with
colleges. So, in addition to losing between 3,000 and 4,000
older employees because of retirement, TVA didn't have enough
already-trained younger understudies to step into the shoes
of the retirees.

The average age of a TVA worker is more than 47 in an
industry in which the average retirement age is 56. (The
average age of an industrial worker in the United States
is around 37.) By 2010, around 60 percent of today's experienced
utility workers will retire.

Then too, even more staff attrition was expected because
power generation is an increasingly deregulated and competitive
industry, and controlling labor costs is one of the prime
ways of ensuring profitability. "We're always looking
for efficiency and productivity gains," says John Walker,
a TVA workforce planning manager -- meaning, not all the
retirees would be replaced. So, even if its crucial intellectual
capital could be identified and retained, TVA also faced
the issue of whether it would have enough of the right sort
of employees to reload with the knowledge.

The most pressing need, however, was to determine what
jobs would be vacated and when, and what knowledge they contained
that had to be conserved or how to cope if it was lost. Before
the inception of the knowledge retention program, the most
worrisome threat facing TVA was: "Somebody is leaving
and we don't know what he knows," says Walker.

The knowledge retention initiative is an aspect of workforce
planning that falls under the category of staffing risk,
and it reports to the risk assessment group in the office
of the interim CFO and executive vice president, financial
services, John Hoskins. TVA's program has three phases that
enable line managers to answer three common-sense questions:

What? Specifically, what knowledge is going to
be lost? This phase identifies at-risk knowledge and skills
-- those likely to be lost soonest and seemingly hard to
replace. This exercise is called a knowledge-loss risk assessment.

So what? This phase prioritizes by asking: What
are the business consequences of losing each item of knowledge?
It weighs the risk of losing this knowledge or skill and
determines how deleterious the loss would be. It cuts the
long list of potential losses down to the few that truly
require action -- the ones most critical to the enterprise.
In effect, it's like developing an intellectual property
endangered species list.

Now what? What can we do about each potential loss?
Plans are established either to retain the knowledge or skill
or to mitigate the effect of losing it. If this or that bit
of knowledge becomes extinct, how do we work around it?

Step 1: Knowledge-Loss Risk Assessment

The first step in the process is designed to identify
positions or people where the potential knowledge loss is
greatest and most imminent. TVA does something few employers
do to find out who plans to leave soon: It asks them. In
the beginning employees were inhibited about sharing plans.
They feared the company would use it to lay them off just
before retirement, or withhold training or advancement, or
slash their department's budget. But as they have become
convinced that the company is using the information for its
stated purpose -- knowledge retention -- increasingly they
have volunteered their career plans. Today around 80 percent
candidly answer the company's annual questionnaires. Now
the process has become automated.

With that information in hand, supervisors rate positions
by computing two factors:

Retirement Factor: The imminence of the risk, as
measured by the time until retirement

Position Risk Factor: An estimate of the difficulty
or level of effort required to replace the position incumbent.

Each factor ranges from one to five. A value of five for
each factor would be the most dire situation. A retirement
factor of five would correspond to a retirement within one
year. A position risk factor of five would correspond to
the loss of unique, critical and irreplaceable knowledge
or skills. The two factors are multiplied to produce a total
attrition risk factor. A total attrition risk factor of 25
would mean the imminent loss of knowledge that is most crucial
to the enterprise. In this case, the highest knowledge retention
priority, action plans with due dates, would have to be developed
immediately to cope with the expected knowledge loss.

The next step is to interview incumbents and their supervisors
to learn the job's knowledge content. Since it is important
to identify both explicit and tacit knowledge, the interviews
include four kinds of questions:

General questions like "What knowledge will
TVA miss most when you leave?"

Task questions such as how to conduct specific tests or operate certain
pieces of equipment.

Fact or information questions focus on what the employee knows and generate lists of
contacts, maps, manuals and other recorded information.

Pattern recognition questions ask about lessons learned and insights about what is
likely to go wrong and how to fix it.

Based on these interviews, TVA compiles a list of potential
knowledge-loss items for the job. These items are analyzed
to determine their importance and decide on the appropriate
course of action.

Step 2: Prioritizing Knowledge Retention Action Items

This phase narrows down the long list of knowledge items
to the critical few that truly require action because they
could damage the enterprise. "We don't have to worry
about thousands of jobs," says Walker. "This process
narrows our focus to hundreds of jobs for the whole agency."

Again, the screening is based on several questions:

What is the relative importance of this knowledge?

How imminent is the loss of the knowledge?

What is the cost and feasibility of recovering it, if lost?

How difficult is it to transfer this knowledge?

As a result of this process and analysis, TVA knows which
issues to ignore, which it can correct with minor effort,
and which require immediate action, either because the knowledge
could be lost soon or the consequences would be severe if
someone with that knowledge suddenly left the organization.
"If someone leaves early [and the departure was not
anticipated], that's the worst case," says Walker.

Step 3: Options to Retain or Mitigate Knowledge Loss

Once the focus is on the knowledge and skills that are
truly critical, specific plans are established to retain
the knowledge or skill or to lessen the impact of losing
it. In some cases the solution is having younger employees
accompany older mentors for a while to learn whatever the
mentors have the knack for, or cross-training someone who's
in a different job. Sometimes the response is simply documenting
a procedure or process that's never been written down. In
some cases new training courses or manuals have to be created
so they include the knowledge that has been, until then,
undocumented "tribal" knowledge. Another way to
mitigate the loss is to engage alternative resources outside
the firm that have the knowledge: find outside contractors
who have the capability, enlist retirees as consultants,
have people at other plants or divisions share the expertise,
or ascertain whether contacts at professional associations
may have the same knowledge.

Results

Some of the results of such analyses have turned out to
be counterintuitive. Jobs that would seem crucial have turned
out to have low or no priority from a knowledge retention
standpoint. For example, although the job of supervising
the operation of a nuclear plant is enormously important
and responsible, the training of such supervisory operators
is thoroughly documented. Everything the operators need to
know in every situation is spelled out in detail. Although
the time to competence for this job is lengthy, what such
supervisory operators need to know is known to others. In
this case, the knowledge retention process has been effective
in that it showed that there was nothing to worry about.

One the other hand, a far more worrisome knowledge retention
hotspot would be an engineer who had spent his entire career
monitoring one piece of special equipment, say a power plant
turbine that runs 24x7. He knows when it's running right
or about to fail just from the pitch of its humming.

In another case, a steamfitter could assess the amount
of corrosion within feedwater pipes by tapping them with
a certain wrench in a certain spot in a certain way and listening
for the sound. So he was asked to coach several other employees
on his method and TVA made a videotape of him doing it.

In another example, an employee knew where all the remote
transformers were in an area. The problem was that they weren't
on any map. He just knew where to drive and no one else did.
So all the locations had to be mapped. His knowledge hardly
seems exalted, and the knowledge management solution seems
simple enough, but had he left and gotten out of touch, there
could have been potentially dangerous problems for the TVA
power grid in that area.

Sometimes there's a way to work around knowledge that
is destined to be lost. One retiring transmission technician
was the only person capable of repairing a particular type
of antiquated switch. In this case, it made more sense to
replace all the switches with new ones than to teach someone
how to care for the older switches.

What Makes the TVA Program Effective

In a complex organization such as TVA, not all knowledge
is equally vulnerable to being lost. The knowledge retention
program is a sensible and creative way to conserve the knowledge
that is most crucial. A number of factors have helped the
program succeed. It addresses a real organizational need.
It's not viewed as merely an HR program. It sets priorities
that are understood by all. The method requires -- but it
also enables -- line managers to cope with an issue they
need to care about. The process allows employees to share
their retirement plans without suffering adverse consequences
and to disclose what critically valuable knowledge they possess.
It also enables other employees to help retain the worthwhile
information -- and make themselves more valuable by doing
so. Like TVA's power grid, the knowledge retention program
keeps the company's intellectual current flowing through
the circuit where it belongs.