If there is one thing business is good at, it's inventing buzzwords and "solutions that don't work. Fortunes have been made recommending hiring emotionally mature people, reorganizing work to make it more efficient, the benefits of team interaction, the importance of quality, cloning your best people, and so forth.

I'm so impressed with this trend that I'm writing a book recommending organizations hire only applicants who can prove they are alive. I'm totally convinced this technique (I call it "Top Living!) will virtually eliminate the age-old practice of hiring deceased applicants and promoting deceased managers.

Aside from tasting sour grapes about not writing a simple-minded best seller, there is now a trend toward "Talent Management! Theoretically, it could become the best thing that ever happened to organizations, but it won't. Why? Talent Management is more complex than it looks&and, it takes hard work.

Bag o' Magnets
Imagine two kinds of magnets. One kind is red. The other is blue. The red magnets represent all the basic requirements for doing a specific job (i.e., job factors, special skills, and so forth). Blue magnets represent a jobholder's basic preferences, skills, talents, and so forth.

As you may recall from high school, every magnet has a positive pole and a negative. Experiments under normal conditions show a negative pole is always attracted to a positive and repulsed from its twin (negative repulses negative and positive repulses positive)

Now, assume your task is to match the best set of blue magnets with the best red ones. That is, you want the majority of reds to stick to the majority of blues. In other words, you want to match the talents of each employee to the needs of each job in the organization.

But there is a problem, not all the reds have the same shape and strength; likewise, all the blues are different. In the end, you know you will have to choose one set of blues, throw them into a bag of reds and hope for the best&Although you would prefer a perfect match, some repel, some attract, and some just sit there in a state of uncomfortable tension.

Likewise, matching people to job requirements is one big mess of attractions, repulsions and tensions between one person and another and between each employee and the job. What to do? What to do?

I know, let's buy one of those best sellers (may I recommend, "Top Living? ) that recommends sending all the blue magnets to a training class that shows them how to co-exist with other blues; or shake-up the bag and rearrange everything; or, study the characteristics of the most effective blue magnets and buy more of them?

None of these approaches solves the basic problem&Job-person match only works when organizations thoroughly understand the skill requirements of each job and thoroughly understand the skills-sets of each person.

State of HR
It seems so simple. Every organization has a list of roles it needs to accomplish: selling, building, counting, and so forth. The socially acceptable view is anyone with enough drive should be able to do any one of these tasks. But, experience shows differently. Sellers are usually very bad counters; counters are usually very bad builders; builders are usually very bad sellers; and good individual contributors seldom make good managers.

In a perfect world, organizations would know exactly the skills required for each job family. And, they would evaluate each applicant's skills using a library of tests, behavioral interview questions, and assessments backed with strong evidence scores actually predicted job performance.

However, instead of professionally studying job requirements, most organizations tend to use outdated and globally indistinct job descriptions that vary from one manager to the next. To make things worse, every applicant takes the same tests, formal studies are seldom proving scores predict performance, and the mantra of the day is, "hire and hope.

No matter what you call it, Talent Management (TM) under these conditions will fail.

A Better Way
Ok. I have explained why TM will fail. What will it take for TM to succeed?

First, forget pop-psychology and fad-o-the-month nonsense. All that will happen is lost credibility and wasted money. Second, wake up and smell the coffee. Training cannot fix "broken employees, people cannot become "anything they want, 50% of employees fail to meet job expectations, and 50% to 60% of managers (at all levels) fail because they were promoted based their performance as individual contributors. (This data is fully supported by a long history of university research and observation).

An effective TM database must abandon the olds ways and concentrate on building complete and accurate competency sets for both job and employee.

"Complete refers to having a list of key competencies associated with each job and having a similar list of key competencies associated with each employee. Overall, a company with 6000 employees will have 6000 employee data sets, but (through the magic of job analysis) will only need to match them with 15 to 20 generic job families.

Achieving accuracy requires considerable work. Most home-grown competencies sound good, but are fuzzy and indistinct. Fuzzy competencies usually sound like this: "Persist in maintaining a strategic competitive position regardless of economic obstacles and market conditions. Fuzzy statements contain too many elements, too much outside influence and can be measured only after considerable time passes.

In this example, the words "persist, "strategic, "competitive position, "economic obstacles, and "market conditions are ill-defined; the necessary on-the-job behaviors are unclear; effectiveness will take considerable time to evaluate; and, external conditions may influence success from one economic period to the next. A worthwhile competency meets the same criteria as a worthwhile objective: realistic, measurable, time-bound, job- related, and observable.

Clear and concise competencies are the only ways to link people to jobs. An example might be: "abstract reasoning ability (i.e., the ability to evaluate and assimilate large amounts of unclear or indistinct data) or "ability to plan and organize complex projects. In short, a TM database built on fuzzy data contains so much opinion and so little fact that it cannot match a specific employee to a specific job.

Making it Work
I've explained how an effective TM database is a catalog containing two independent sets of data linked by a common competency language. On one side, there is a searchable list of job-families with competencies organized into cognitive abilities, organization skills, interpersonal abilities, and motivational factors. The other is a searchable list of similarly organized employee skills.

Now, let's look at where this information comes from. Building an organizational database starts with clustering jobs with similar functions into job families. Once this is accomplished, a skilled analyst interviews jobholders, managers and senior managers from each family identifying and cataloging key competencies associated with job performance. For legal, political and due-diligence reasons the entire process is supported by confirming surveys, demographic analysis and documentation.

However, collecting trustworthy employee-skills data is an entirely different matter. You don't have to be a professor to know everyday tools like multi-rater feedback and performance appraisals are usually highly inaccurate. For example, there is abundant research showing multi-rater feedback suffers from the "you scratch my back and I'll scratch yours syndrome; while performance appraisals suffer from problems like forced-distributions, no-one-here-is-perfect, and everyone-here-is-perfect. Inaccurate employee evaluations tend to force employees into one of three buckets: 1) walks on water, 2) pond sludge, or 3) has not done anything wrong yet.

The best way to collect employee skills data is to use a personal competency classification system that mirrors the job-competency classification system. Smart organizations will collect employee data from performance on validated hiring tools; job-tailored survey-based feedback; job-tailored performance appraisals; and supplemental validated assessments. It goes without saying employee measurement is a major effort that is done on a regular basis.


Wrap Up
An effective Talent Management program requires several elements. Individual jobs have to be divided into families; a robust system of competencies must be able to link jobs with employees; competencies associated with each job family have to be gathered from job experts; and, competencies associated with each employee must be continuously collected from hiring, performance, training, and supplemental assessments.