The Initial, Critical “Contracting-Diagnostic” Conversation

Lots of recommended books in the areas of performance technology and training give sound advice at a kind of “macro-level.”

They often, rightly, advise that one “sine qua non” of long-term success is the direct involvement of operating managers who have effective control of their components.  They correctly advise that one needs to look at and provide for the “on-the-job” situation that is the context for any desired performance.  The also, wisely, advise measuring, both from the beginning of a project, and on a continuing basis.

But the literature rarely says much to the practitioner about how, concretely, to go about doing such things successfully.

 In this post, I am going to describe how one performance improvement manager dealt with an operating manager who came to him for assistance in dealing with an important operating problem.  I will stop, occasionally, as I tell this story, to call attention to some of the moves made by both the performance improvement manager (whose door said “training and development”) and the client operating manager.  In this way it may be possible to highlight some important sequences and junctures in this conversation and to suggest advantageous moves for the performance improvement professional to make in similar situations and conversations.

One morning, Bob, an operating manager, called Bud, a ”training” manager and asked the’s latter assistance in setting up some training that Bob planned to conduct. 

(Bob and Bud are members of an organization of about 3,000, that is geographically decentralized nationwide.)

Bud responded “Sure, when could I come talk to you about it?”

Bob said, “Well, I’m pretty sure about what I want to do.  We’ve been having trouble with the quality of our wage surveys and I want to call in the regional wage specialists to reinforce their understanding of the regulations and standards.”  I think, mostly, I need your help in arranging a meeting room, and maybe with a little facilitating of the question periods.”

Bud: “Well, let me come your way.  Sometimes it’s useful to talk a little at the beginning, even about things that are straightforward.”

Bob:  OK, how about 9:30 AM tomorrow?

Bud:  “I’ll be there.”

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Now, look back at the very first exchanges in this conversation.  Some definite moves have been made and signals have been given.

1.  Bob, the operating manager, initiated this contact and has asked the “training” unit for some help.  That’s always preferrable.  In fact, I would argue that it’s almost always essential.  It has to be the “operating manager’s” problem or objective if any resulting intervention is to result in long-term performance improvement.  Operating managers, not trainers, run the organization.  So if you’re a performance improvement-training professional approaching operating management in order to “sell” something, beware.  Your efforts may be wasted from the beginning precisely because this is your idea not theirs. 

(This latter point does not suggest that you should not be pro-active, in some sense, with regard to operating managment.  But character of your pro-activity is critical.  It should always be about helping operating management identify and respond soundly to problems and objectives that they discern and feel impelled to work on.  To come to them with a problem or objective or intervention that is your perception or idea, rather than their’s, is nearly always fatal.)

2. Bud, the “training” unit manager has responded quickly indicating his interest in working WITH Bob on this project.

Notice that Bud has not quizzed Bob at all on the phone, but has suggested a face-to-face meeting in Bob’s office (a setting and mode of conversation that increase the liklihood that Bud can be useful).  It is usually better to meet in a prospective client’s job setting rather than in your own office, unless useful conversation would be difficult in the former.

3. Bob has questioned whether a meeting is needed.  He has also signaled that he’s “down the road” considerably toward a solution of the problem he perceives.

4. Bud has not argued with Bob’s description of how he sees things proceeding, but simply says that an early meeting is often useful.  This is a plausible, non-controversial reason for meeting.

5. Bob agrees to meet and suggests a time.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Bud is relatively new to this organization, but finds out with only a little research that that Bob’s component is responsible for determing the minimum wage rates that must be paid to workers on various sorts of government construction work contracts.

The sequence begins with the issuing of a request for proposals for a particular government contract.  The contract specifies that geographical location where the work will take place and lists the construction jobs, by title that will likely be required.  Bob’s is responsible for issuing, as quickly as possible the minimum wages that must be paid for each job listed.  Bob’s job is a hot box and a political one at that.  The contracting folks want the list of rates to be paid quickly.  If the rates that Bob’s unit issues are seen as too high, the contractors get upset and complain, often to Congress.  If the rates are too low, affected unions do the same.  So there’s a lot of reason for Bob and his staff to be able to produce sound, defensible lists of rates and to do so as quickly as possible after a request.  Errors are real potential problems (beyond the cost of repairing them), since they make a given list of rates open to legitimate complaint from one of the interested parties.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

It is important and legitimate to do this sort of research and preparation prior to a diagnostic conversation with a client.  You need to know, if you don’t know already, what you can discover quickly about the world a prospective client faces.  Don’t try to impress the client with what you know or have discovered, but instead use this “knowledge” to shape any leading questions you need to ask. (It should also help you understand what the client says.)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

9:30 am arrives and Bud ”knocks” on Bob’s door, so to speak.

Bob:  So what do you think?  Can you help me get a room and a little facilitating?

Bud:  What did you see that first suggested to you that there is a quality problem with the surveys?

Bob:  There were lots of errors in the completed surveys submitted to my staff by the regions.

Bud:  Do the regional staff conduct the surveys?

Bob:  Well, they’re responsible for working with the district offices to get them conducted, but the districts usually collect the actual data. 

Bud: Who in the districts actually collects the data?

Bob: The districts assign the data collection to particular investigators. 

Bud: Are these investigators specialists in this survey work?  Do they do investigations too? 

Bob: They are primarily investigators.  They do this survey data collection as an ancillary duty from time to time.

Bud:  Are errors in the survey data submitted by the district investigators to the regions also frequent?

Bob:  Oh yes.  The Regional Wage Specialists complain about that a lot.  But where are we going with all these questions?

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Again there are things to notice, looking back on this part of this conversation.

1.  Bob asks again whether Bud can help him engage a conference room and arrange some facilitating.

2. Bud does not let his response be shaped by this question, but simply moves to begin to explore the problem situation Bob sees. (This is extremely important.  Too many of us feel obligated to deal with the question asked instead of simply going to work analyzing the problem the client is experiencing.)  Bob (at least initially) responds readily.  Managers like talking about their real work.

3. Notice that Bud goes directly at trying to determine WHO is making the survey errors Bob is seeing.  With a simple series of questions, Bud determines that:

a.  the regional specialists being called into the session Bob plans are not usually those who actually collect the survey date. (Bob has, inadvertently, omitted an important part, maybe the most critical part, of the employee universe at which any intervention, aimed at reducing survey errors, should likely be directed.) 

b. the actual survey data is usually collected by investigators in the district offices (so it is likely that any intervention that is to treat the sources of errors in the survey process needs to provide for errors being made by the district investigators who collect survey data).

c.  the investigators collecting the survey data do so as on an ancillary basis, they are still primarily investigators (this suggests that they might well not treat collecting survey data as as important as their primary investigative duties and may even resent being assigned to survey collection work, unless there is some perceived “honor” associated with being selected for it).

d.  Bob has been willing to talk about the questions Bud is asking, but is puzzled about why they aren’t talking about arranging a conference room and some facilitation (Bud’s next response must show convincing reasons to go on with Bud’s questions, if that is needed).

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Bud:  We’ve been describing concretely, various aspects of the problem you are experiencing.

We’ve already determined that district investigators, not just the regional wage specialists, are a source of some of these errors.  Even if we reinforce the understanding the regional folks of the standards the surveys they submit need to meet, that likely won’t affect much what happens with the people who actually collect most of the survey data.

So you’ve already demonstrated that it might be important to examine this problem situation a little more closely to make sure that any action you decide to take about it has a good chance of producing the results you want.

Do you know or could you find out quickly approximately what per cent of the surveys being submitted by the regions have errors in them serious enough to require that they be repaired?

Bob:  Well, I’d have to ask my staff (they’re organized by region) what their actual error rates are, but I’d say that we actually have to repair about half of the surveys the regions send us.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Looking back:

1.  Bud points out that by talking about this problem area only a little, Bob has been able to determine that district investigators not just regional wage specialist are a source of these errors.

2.  Bud also gently suggests that just talking to the regional folks about the standards they should observe won’t have much impact on what is happening at the district level.

3.  Bud then says that Bob has already demonstrated that a little examination of this problem area is worthwhile and he moves to continue to do so, with a further question.

4.  Bud’s next question introduces measurement.  Does Bob know what error rate is being experienced?  Notice that Bud introduces measurement in an undemanding way.  He asks for the “approximate” error rate. 

This is also important in at least two senses.  First, often managers and organizations do not have measures in place focused on the particular problem of interest.  But second, often, even quite approximate information about the character of a problem can be useful.  Precise information is often costly.  So consultants need to be sensitive to this by being willing (at least initially) to deal with rather humble, even “dirty” information during early diagnosis. 

More, consultants need to develop the skill of helping managers build quick, inexpensive measures that bear on a problem area in the space of conversation.  Most operating managers are not asked to create measures.  Those that they work with are usually imposed.  So, it is often the case that very capable operating folks are not used to thinking in these terms.  They can, however, be important contributors to the creation of useful, simple, low-cost measures if the consultant can provide a little conversational “scaffolding” for this task.

5.  Bob’s response (especially his willingness to make it) is perhaps the most important thing that has happened so far in this conversation.  Bud has made an argument about why the analysis they have done is worthwhile and then has moved directly to continue it.

Bob’s response about the error rate question indicates, at a minimum that he is willing to go a little further.  It may also indicate that he has a beginning awareness of the fact that his original notion of a meeting with the regional wage specialists may not deal adequately with reducing errors in submitted wage surveys.

It appears that Bud has successfully converted a question for room reservation and facilitating assistance into a real effort to examine this problem in sound performance analysis terms.

Bob will likely need to continue to see advantages in this tack, but has at least for the moment agreed to go further with it.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Bud:  So about 50% of the surveys submitted to your national office staff have to be redone in some way.  Do we know, or could we determine quickly, again in an approximate way, what proportion of the survey data being submitted to the regional wage specialists by the district offices, is also sufficiently flawed that it has to be repaired or redone in some sense?

Bob:  Yes, I could ask a few regional specialists. (Presses intercom and says:  “Jerry, call Atanta, Dallas and Chicago and ask the regional wage specialist to give me a quick, rough estimate of what percentage of the wage survey submissions they get from the districts have to be repaired or redone in some sense.  Buzz me in here when you get that.  Thanks.”

Bud:  What happens when your staff gets a survey from a region that needs to be repaired or redone in some way?  Do you send it back to them and ask them to fix it?

Bob:  Well, sometimes we send a survey submission back, but usually the timeline is too close and the contracting staff is pushing us for an issuance, so my folks often make the repairs themselves.

Bud:  And do we know what happens when a survey submission from a district office needs to be reworked in some sense?  Do the regions send it back?

Bob:  Well, I’m sure it varies.  Often the regions do the repairs themselves by phone, if they can, but sometimes they have to ask the district office to gather more local data.

Bud:  And if they do have to send a survey data submission back to the district office, does it go to the investigator who collected the data originally?

Bob:  I don’t think so.  Because collecting survey data is an ancillary investigator duty, I think the district managers tend to spread that kind of work around.  So I think that they usually assign the repair work to a different investigator.

Bud:  So the investigator who collected the original data might not get real feedback on what problems there were in what he submitted?

Bob:  I don’t really know, but I doubt it.  He’ll probably get kidded by the guy who has to do the repair work, but the investigators are so busy with their investigations that a few missing job categories in some submitted survey data won’t attract much attention.  It’s outside what they see as their real work.

Bud:  And what do your own staff people say to the regions about submitted survey data that your people have to repair of complete?

Bob:  Well, my folks talk to the regional wage specialists all the time, so I’m sure they say something about deficiencies they find, but again, they’re often pretty swamped themselves and I’m guessing again that their complaints to the regions are usually pretty general. 

Someone might have submitted survey data for building construction contract, it’s got a number of deficiences, but the most glaring one is that it left off “carpenters.”  So my guy will call him and say, “Hey, Charlie, are they doing building construction up your way now without carpenters?”  He’ll rag him a bit about the worst errors, but he won’t usually go chapter and verse because they’re both so busy.  But, as I said, my staff will usually just fix it ourselves so we can get the schedule issued.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Looking back:

1.  Bud, after introducing the notion of measuring what proportion of survey submitted by the regions to Bob’s staff and getting Bob’s estimate of 50%, moves to see what proportion of the survey data submissions by the districts requires rework by the regional specialists.

Again, notice that Bud asks only for approximate information, since this is more likely to either be available or within the range of fairly accurate estimate.

2.  Bob does his part in keeping this analysis moving by immediately asking someone on his staff to call three regions and get their estimates.  This won’t always happen this way and in some instances this diagnostic conversation would need to be interrupted until Bob had these responses.  In this example, Bud decided that he could go on usefully putting shape of other aspects of this problem.

3.  Bud asks next about how surveys that need to be repaired are repaired. 

One important aspect of this is whether the person who did the work that needs to be repaired is ever notified of that and specifically of what the errors were and what a correct submission should have looked like.  To use performance improvement jargon, is there feedback to the performer about the results of his/her performance?

4.  Bob’s indications suggests there is not usually feedback to the performer, who has made a submission that requires repair, that would equip him/her a) to recognize the original error, and b) to be able to make a correct submission in the future with regard to this aspect.

What happens with survey errors is very like taking the typing errors made by one typist to another typist for correction, without ever letting the first typist see what errors he/she had made or what the text would have looked like if it had been correct.

The lack of feedback loops to performers that would work both to let them identify their errors themselves and to perform correctly in the future is surprisingly frequent in work organizations.  And they look so obvious that sometimes performance improvement specialists are reluctant to look for their presence.  One should always do so.  The provision of adequate (preferrably non-punitive) feedback to performers about their performance is one of the simplest, least costly and most powerful interventions that a performance improvement professional can suggest (and, if needed, help construct).  The impact of such feedback on performance is almost always FAR more dramatic than that of any training that might be proposed.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Bud:  Is there anyway to estimate approximately how much time it takes on average to repair a survey submitted to your staff that needs that?

Bob:  Well, of course it varies widely.

Bud:  But could you make a rough estimate of how many hours it usually takes?

Bob:  Well, I know that it usually takes at least a couple of days.  Let’s say two and a half, so that I’m not under-estimating.

Bud:  So two and a half days…that’s 20 hours, right?  You’re folks work eight-hour days?

Bob: Yes.

Bud:  And what typically are the grade levels of your staff doing this repair work?

Bob:  Most of my staff are in grades 12 and 13, my branch chiefs are 14s.

Bud:  So if we took the midpoint of each of those salary grades, averaged the 12 and the 13 and estimated how much branch chief time we need to factor in, we’ll know about how much a typical submission to your staff is costing both in hours and in pay.

(A schedule of the grades and associated pay rates is readily available and Bud and Bob quickly calculate how much a survey that needs repair costs.  The number is higher than Bob thought it might be. 

While they are working on this calcuation Bob’s staff member buzzes Bob and reports that the three regions all estimate that they have to repair about half of the data collection submissions they receive from the districts before they can be submitted to the national office.)

Bud:  So you can now get estimates from the rest of your regions of what proportion of wage data submissions they have to repair and ask all of your regions for estimates of how many hours are required on average to make the repairs needed.  Then you can determine the approximate pay rates of the regional employees who do the repairs and get estimates of how long repairs typically take and you’ll be able to say both about how much time repairs take and approximately what they cost in lost pay hours devoted to them.

Bob:  That’s very interesting and it is more costly than I thought, but why are we doing all this analysis, why aren’t we working on how to solve this problem?

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Looking back:

1.  Bud is continuing with measurement, this time of the cost of repairing deficient, submitted surveys and survey data, both in terms of the average number of hours required and the associated cost in wages.

He does this together with Bob for the national office staff, and explains to Bob how he can do this same analysis to estimate the likely costs of repairs on these two dimensions of submissions for each of his regions.

The “working together” is important both because it models for Bob, with his active participation, what he needs also to do at the regional level and because it shows him explicitly, and without any condescension, how to make these estimates and calculations at that level.

2.  Bob is surprised at how much the survey errors at the national office level alone are costing, but is still puzzled about why things in the current conversation are not moving to solution.

So, again, Bud’s next response is critical, and must provide Bob with reason’s for continuing with analysis before moving to fashioning a solution. 

[It is important to note that this equation continues to exist at every stage of a project.  The client must always see it as advantageous to go on in the way the performance improvement professional is suggesting.  And the performance improvement specialist must always be able to provide a ready, convincing explanation of both what path is being recommended and of the advantages (to the operating manager) of going that way.]

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 

Bud:  We’ve been doing three important things.

First, we’ve demonstrated that the problem of survey errors is more general than we might initially have thought and it includes the district investigators who actually collect a lot of the data needed.

Second, we’ve been able to estimate what proportion of surveys require rework at both the national office and the regional levels and if we combine these two estimates (which might not be completely licit, but is probably not nonesensical) you’re experiencing an effective error rate of 100%.  That strongly suggests that the problem you have spotted is a truly serious one.

Third, we’ve estimated the cost both in lost hours and in wage dollars that the repair of a given defective survey submitted to your national office staff incurs.  You say it’s higher at the national level than you expected and when you get the estimates from the regions you’ll be able make a comprehensive estimate of what the repair of survey errors is costing the organization.

Being able to show that also demonstrates that it is a problem worth working on, so you should be able more readily to justify the resources needed to map it accurately and to correct it.

Do we know what sorts of errors occur most frequently?

Bob: Oh yes, there’s a clear pattern.  The most frequent problem is the omission of wage data for particular jobs that are included in, and so are needed for, a particular contract.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Bud has given Bob three results of their conversation so far and suggests that not only will it be advantageous for Bob to map the frequency and cost of this problem organization-wide in at least approximate terms, but that the cost he has already calculated for repairs by his national office staff alone is surprising to him.  Bud has also suggested that when Bob can indicate nationwide, the estimated proportion of surveys or survey data submitted that requires repair and that total cost being incurred in doing so, Bob will be in a position to argue that this problem is important enough and costly enough, to merit investment of some resources in reducing such errors.

Again, Bud goes on with a further analytical question, this time about whether some errors in surveys are more frequent than others.

The fact that Bob goes back to work analyzing and that he knows something about the kinds of errors that are most frequent, indicates both that Bud’s reasons have been at least momentarily convincing. 

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

(Bob has just said that, perhaps, the most frequent error is the omission of the data for particular jobs that are needed under a particular contract.)

He goes on now, indicating that he also has a theory or two about why complete data is often hard to obtain.

Bob: …And one reason, that occurs, especially in non-union parts of the country, is that employers are often reluctant to tell us what wages they pay.  Some see this as proprietary information.  Others simply don’t want to take the time and trouble such reporting requires.  There’s also the perception on the part of some employers that there no point in their telling us what they pay because they feel that we weight the wage rates we issue strongly in favor of what the unions report. 

Getting employers to give us the data we need to make sound issuances is a real difficulty.  It’s a place where I think the local investigators would have more success if they would treat such work more seriously.

(Ed: This conversation continues and at the end of it Bob has identified some additional errors that occur and have to be repaired in survey submissions.)

The initial meeting ends with Bob promising to get the rest of the estimates from the regions on the proportion of data submissions that get that require repair, how long such repairs take typically, and what the average pay rate is of the regional staff doing that repair. 

About a week later, Bob has this regional data and has made the needed calculations.  It turns out that the estimated, total national cost of the repair of deficient survey submissions is over $900,000 annually (This analysis was done in 1972.  The cost today would be dramatically greater).

There are subsequent meetings, some of which include members of Bob’s national office staff, some regional wage specialists and some district office investigators who have been assigned to collect survey data more than once. 

In these meetings, the experiences of those collecting data at the district office level, that of the regional specialists and Bob’s national office staff are related and probed to get the best description of the various facets of this problem that can be managed.

It turns out that there are some “skills and knowledge” deficiencies and some training materials are built that provide low-tech simulations of what both the district investigator-data collectors and the regional specialists face and must deal with effectively in such situations.  “Book answers” for each situation are also agreed on and the “rules of practice” visible in them are extracted and described with additional examples in some “resource materials.”

Because it is clear that feedback loops need to be created for both district data collecting investigators and for regional wage specialists so that information about the errors in their submissions are made available to them (including suggestions for how to make the corrections, if this it not obvious), a member of Bud’s (“training”) staff takes the lead in developing one such sheet for each level and revises them with the help of the folks from Bob’s side of the organization at national, regional and district office levels.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Now I want to ask you look back here to the very beginning and to estimate for yourself how well Bud (the “training manager”) and his staff have done in helping Bob (the operating manager client) with the problem he has brought to them.

I, personally, would give Bud high marks at almost every step and stage.  He worked skillfully and successfully with Bob to help him put far more concrete dimensions on the problem situation that Bob seemed to imagine.  He quickly determined that those doing survey work did not get specific feedback on their performance, even when that involved fairly serious error.  Bud gave Bob good clear explanations of what and why they were doing so much analysis and good reasons for going on.  The evidence of whether these reasons and analyses were seen as credible and useful to Bob is provided by the fact that Bob was willing to continue with the analyses. 

So I want to suggest that, to this point, this story provides a favorable and useful model of some of the moves a performance improvement specialist can make while trying, concretely, to take some of the advice of the experts in our field.  

What Bud did cannot be followed or applied mechanically to other situations, but some of his questions, the skillful way he was able to get Bob to analyze the problem before acting, and the way he helped Bob fashion and apply some “quick and dirty” measurements, can likely be drawn on usefully by others in our field.

I have asked you to make an assessment of the work of Bud and his staff at this point because this story has a sad ending.  Sometimes, even if the performance improvement professional has made all the right moves, real success will turn out to be beyond reach.

Here is how my story ends. 

About six weeks after Bob first called Bud, a pilot session is scheduled of a short class-room learning sequence in which the skill and knowledge deficiencies are to be treated with realistic practice and critique cases.  The two feedback sheets will also be introduced and a pilot use of them by the class participants will be proposed to the pilot participants (Bob cannot mandate this second part.  Most of these participants do not work for him.)

The pilot session, mostly, goes very well.  The participative design works as planned.  The participants like and see the value in the “skill and knowlege” practice problems and the “vibes” in the session generally are very positive thoughout the learning sequence.

The learning segment of the pilot is completed and the two survey feedback sheets are produced and their purpose and use is explained. 

Some responses resist an additional item of reporting in a world some participants think is already too full of it.  But the most important signal about such feedback is given to Bud’s staff during coffee break.

There are two union representatives among the participants who are also investigator-data collectors.  This is deliberate, both because Bob wants to make sure there is no union opposition to this intervention, but also because, since the feedback sheet could be seen as a change in the “work process,” it is thought that union agreement about its introduction may actually be required by the union contract (Bob has asked labor relations, but got a fuzzy response.)

These two union representatives take Bud, and a staff member who are facilitating the pilot session, aside and say to them that feedback sheets are a good idea.  More, they say, they would very likely work to reduce errors in submitted surveys at both levels.  The problem, these union members gently confide, is that the use of these written survey feedback sheets produces a body of performance evidence that can be used against investigators who are assigned to to survey work.  All investigators hold the same classification, but evaluation of data collection would be applied only to some.  The first objection, these union representative say is that this is “disparate treatment,” since only some investigators would be evaluated on such work.  But the real objection, they say, is that it creates a body of data that could be used punitively against particular employees and that they (the union) have never seen data that could be used punitively against employees not used in that way. 

For these two reasons, they say very politely, almost apologetically, to Bob and his staff member, they will, when the pilot session reconvenes, publicly oppose even the pilot testing of the feedback sheets at both the district office and regional levels.

They do and some of Bob’s staff members, including two branch chiefs (who see these feedback sheets as generating written evaluation data that could potentially reflect on their performance as well) join in their objections.

The decision about whether to pilot test the feedback sheets is tabled and the “training” session ends.

Bob, his staff, and the field participants, are all effusive in their praise for this pilot on their session evaluation sheets and (a “performance test” of actual learning has not been afforded) recommend that this learning sequence (with minor modification, mostly by way of elaboration with additional case situations) should be conducted nationwide.

The use of the feedback sheets was never considered further once Bob told his superior of the union’s opposition.

This sort of result is something any performance improvement specialist needs to work hard against but must be prepared to encounter.  Such a defeat can occur with even the most enlightened client (and Bob was pretty good). 

The worst thing is not that the performance improvement professional’s best advice and product was discarded, but that the organization felt ultimately (I fear without much real consideration) that it could not institute, over union opposition, something it likely needed in order to have real reduction of the deficient surveys and incomplete survey data it was experiencing at great demonstrated cost.

 The training was given and Bob was satisfied that he had taken sound action and produced an effective intervention.

As far as I know, there was never a move, in Bob’s organization, to begin to collect on a continuing basis systematic data on the proportion defective submissions at the regional and national office levels, the amount of time required to repair a given defective submission or the nature of the errors that made a submission problematic.

 Bob and his organization seemed not to experience any real further pain about an idenitified error rate that was likely largely continuing.  In fact, in a hallway conversation about a week after the pilot session one of Bob’s branch chiefs said “You know, it seemed like a big number at the time…the $900,000 for repairs.  But it’s all fixed cost.  You wouldn’t save any salary cost by reducing survey errors.  Everyone involved would still have to be paid…”

I walked on down the hall, looking for a sufficiently private place in which to shake my head.

I much admire the folks in our field who give good advice at what often seems like the “macro-level.”  Some of them are, as the British say, “clever” as all get out.

But wouldn’t it be nice (and useful) to have a book of the ground-level stories of some of us who have mostly worked at that level and may sometimes have done very well?  I think that would be a book from which all of us could learn.

R. John Howe

“For” Versus “With”: An Important Distinction and a Dangerous Temptation

“Training” is an organizational resource for change.  So are a variety of other performance improvement efforts in which some trainers sometimes indulge.

This often leads trainers, and those who fashion a variety of human performance technology interventions, to see themselves, importantly, as organizational change agents.  They may, in fact, sometimes feel that they have at least a professional responsibility, if not an organizational one, to foster change in the organizations in which they work.

This notion of organizational role may lead some trainers and performance improvement specialists, to look about their organizations for needs they can spot, either personally, or using modest surveys.  They may also feel responsible for looking into the literature, not just for their own professional competence, but scanning for, and alert to, notions and approaches that seem to them useful to attempt to implement in their host organizations.

For example a given trainer may become convinced that either “team building” or improved skills in “conducting effective meetings” is needed his/her organization and may set about, either finding some available resource in this area, or building a custom intervention, drawing on approaches that seem particularly enlightened.  Perhaps the trainer notices (as Kepner and Tregoe did years ago) how inefficiently most people handle information in their organizational “problem-solving” and “decision-making” and “sees” how beneficial improved skills in these areas would be to the operation of the entire organization.

This tendency resides in the context of at least two others.

First, is the fact that training is widely seen to be one of the organization’s “service” functions.  Conceived literally, this suggests that it the training function’s job to relieve operating management of the tiresome task of training employees.  The training unit should intuit the organization’s training needs, and design and implement needed training (and other performance improvement interventions) without bothering operating management much.  Ideally, the central organization will have provided the training function with a budget, and operating management will not even need to contribute (except, perhaps through, a modest annual assessment) to the costs of providing needed training.

The notion that the training function should provide needed training without unduly burdening operations does not obviate a second sense in which the training function is often expected to be of “service.”

There is in many organizations an operating assumption that the training function will respond quickly, and without much question, to management requests for particular types of training and other services seen to be training-related (for example, the coordination of conference room use, or of an organizational auditorium, or of requests for AV services of various sorts).

Managers may come to the training function asking it to arrange such things as “communications” training, “time management” training, training in “handling difficult employees,” “safety” training, or even “leadership” training.  The training function is not expected to “think deeply” about such requests, but rather to have a ready list of resources in such perceived need areas that it can broker and engage quickly.

The organizational tendencies described above often combine to send to trainers and training functions a strong message that it is their job to provide needed training FOR the operating organization and its management without bothering it much.  And because most trainers sincerely want to do a “good job” there is a strong tendency, even on the part of many experienced professionals, to respond to such demands, on the basis of the terms in which they are made, without attempting to probe them much.

In what follows here, I want to argue that the assumptions described above are not only based on erroneous notions about the situaton that actually prevails in most work organizations, but that acting in congruence with them nearly guarantees, not only that the intended objectives will not be realized, but that the relatively scarce resources most organizations have available for training will be mostly wasted.

Since training is a resource for change, and since the sort of change training (and its related performance improvement interventions) can achieve, is a matter of arranging things so that employees perform in desired ways, a critical first thing to notice is who controls the context of any given desired employee performance. 

Note:  In focusing here on the “environmental” side of performance I do not deny human motivation can also be important.  My claim is only that the pervasiveness of contextual factors usually, ultimately, work to corrode all but the most exceptional instances of individual motivation, and since the efficacious “levers” affecting human motivation are complex and often disputed,  the contextual side provides the best pressure point for improving performance in general.

So if we want to mount training and other performance improvement interventions that work, we need to look at how power is distributed in most organizations and, especially, who shapes the context of any performance of interest.

The passage that follows here, immediately, may seem to state the obvious, but may still  be worth saying out loud, since there is lots of behavior in organizations that seems to deny it. 

Overwhelmingly, most work organizations are hierachical.  Power is distributed unevenly through them and, with the coming of computers, is often sharply centralized in the hands of operating management.  There may be organizations in which the rhetoric of  employee “initiative,” and of “creativity” is, in part, real, but in an age marked by such developments as “enterprise systems,” mostly, organizations are looking for employee “compliance.”

(This latter tendency is not entirely a bad thing.  Most of us likely want both our brain surgery, and the resulting adjudication of our medical insurance claim, to be performed following closely established and tested protocols.  There is undeniably a need for creativity in such areas from time to time, but most of us would prefer that it not be necessary on our particular case.  Organizations tend to be similarly conservative about protocol observance.  Sometimes they are interested in ensuring given quality levels.  In others, they may want to reduce their exposure to legal liability.)

 The greatly asymmetric character of power in most work organizations strongly suggests that there are a range of training and performance improvement interventions that need, realistically, to be excluded at the outset. 

 Operating management has repeatedly demonstrated that it is not interested in voluntarily sharing the order of power that would permit real transactional relations with non-managment members of the organization.  This means that interventions that assume such power-sharing are likely unrealistic.  In general, I think it suggests that efforts in “employee empowerment” and “employee creativity,” even, nowadays, many efforts aimed at real “decentralization,” are rather predictably likely to founder in the face of organizational power relationships and should, mostly, not be attempted.

Also included for exclusion here, are interventions that are sourced, mostly, in the perceptions of the trainer.  Even if the trainer is right with regard to a particular need, he or she does not run the organization and so cannot, either guarantee that an intevention that is designed in response will ever be used (this is one place where empty seats in training courses come from), nor, even should the intervention be delivered so that the desired learning has occured by course end, can the trainer create the conditions back on the job that will encourage and reinforce its application in every day work.

Most of the above is by way of suggesting that the optimum consulting paths open to trainers and to other performance improvement professionals who want to help their organizations achieve desired results are pretty sharply circumscribed by the power relationships that obtain in them.  As long as we have a capitalist economy in which power is determined by ownership, notions of democratizing the work place, in some sense, are largely exercises in denial.  Instead, it seems to me, “trainers” need to face the realities of existing organizational power relationships and align their consulting strategies and efforts with this reality.

 What are the implications of accepting the argument above?

 Well, first it means, as I have already strongly suggested, that “trainers” and other HPT practitioners need to recognize that they should not be attempting to make their own judgments about what their organizations need.  And they should definitely not be going about investing scarce organizational resources in the building and attempting to implement interventions that are primarily based on their own notions of what is organizationally useful.  This is one sense in which “trainers” should not be doing things “FOR” their organizations and its operating management.

Second, it means that “trainers” need to recognize in their everyday behavior that operating management runs the organization.  Operating management and its behavior, unavoidably are, or create, the prevailing context in which both employee and organizational performance occur.  Operating management always has its preoccupations: objectives it wants to achieve and problems it wants to solve.  This means (again unavoidably) that efforts to improve employee and organizational performance must be a matter of discovering these operating management preoccupations and of sharply aligning the investment of the limited resources the organization can afford to devote to training and other performance improvement interventions with these management concerns.

Third, this means that diagnostic efforts (i.e., determining what organizational objectives and problems are of concern) must unavoidably start in conversations with the organization’s real operating managers (the real movers and shakers).  As inconvenient as this may sometimes seem, both to ”training” professionals and to the operating managers who in fact run the organization, there is no way to select, shape and implement performance improvement interventions that will impact individual and organizational performance in any worthwhile, continuing way without “bothering” the important operating managers. 

Now I do not underestimate how difficult it may be for “trainers” to arrange the sorts of conversations needed with the operating managers who are the real movers and shakers of the organization.  The “trainer” may not have ready access to such folks and will often need to conspire more than a bit to arrange the sort of conversations needed.  This is a place for the development and exercise of organizational skills not necessarily specific to the trainer’s professional specialty. 

 In situations in which I have identified such an operating manager, but have not had ready access to him or her, I sometimes asked the top manager of my own function (and administrative office head) to make the request for the needed conversation and/or to accompany me (at least initially) in it.  This, in turn, necessitates making one’s own organizational superiors knowledgeable about the need for and desired character of such conversations and requires that they be in agreement with what you are about. 

The operating manager (and anyone in your own function who is to either arrange it or accompany you) will want to know the reason for which you meet and this is difficult to describe, optimally, in terms that will make sense to either of them.  Ideally, you want to have a conversation in which the operating manager describes the major objectives he/she wants to achieve in the coming year and the major problems he/she would like to solve.  It would be best if the operating manager talks about these major objectives and problem areas without regard to training.  This is because, training is only useful on performance problems due to lack of skill and knowledge, and if target operating managers presume a training endpont, they tend to ”edit out” problems and objectives that they assume can’t be impacted by training.  And since “trainers” can sometimes be useful to management in building interventions to improve performance in areas not the result of skill and knowledge deficiencies, it is better of “non-training” problems and objectives are not dropped out of the operating manager’s initial listing.

 What I have, personally, tended to say, when asked what this conversation is to be about, is ”There are relatively limited resources to devote to training in the coming year and we need to talk to you about the big things you want to accomplish so that we can work with you to ensure that those devoted to your organization are aligned as closely as we can manage with your own objectives.”  Sometimes they will respond “You want to talk about training projects?”  I say, “Actually, it would be better if we talked first about the big things you want to achieve without considering at all whether training might be useful in dealing with them.  We can decide later whether training might help you accomplish particular things you want to achieve.” 

You want, if you can, to avoid getting this conversation couched in terms that begin with “training” and that’s hard if you’re from “Training.”  But you need to try.  It makes a great difference how this conversation starts.

 Let me assume, what is not assured, that you have obtained agreement from a top operating manager to a discussion of the sort described above.  How do you prepare?

First, it is useful to notice that, what I have said above about the fact that operating managers run the organization and about the need to recognize that fact, does not mean that “trainers” cannot and should not take any initiatives (notice that we have already recommended that you do so in making the requests for these meetings).  There will usually be adequate space for proactive contributions in your dealings with powerful operating managers, while still acknowledging the realities of organizational power.  There is a reason why you are in the room in such discussions.  You have, or should have, a set of professional skills.  But the sound exercise of such skills does not include “selling” operating managers on either your notions of the problems that exist in your organization or on ”solutions” to them that might attract you personally.

One initiative open to you is careful preparation for this meeting and conversation.  “Trainers” are well-advised to “go to school” on their organizations attempting to determine from published sources what the current situation is in it. 

One particular thing to collect about the target organization is what significant “change points” are visible in it, looking back over the current year and ahead to the coming one.  Such change points can float at various organizational levels.  A new technology may have been adopted.  Some new legislation may need to be accommodated.  Operating management may have either implemented some new policies or procedures or be about to.  A new job of some sort, or a group of them, may have been created, sometimes an entire organizational unit is being brought into being.  Even the recent promotion of a key manager may be a change point worth considerable attention.

Collect also available information on the organization’s performance over the past year or so and any available indicating problems it may be grappling with.  One can often get a sense (you may well already have it) of how the operating manager you are going to meet with is viewed in the organization both by his/her subordinates and superiors.  Does he/she seem to have a pretty free arena of action in the sector of the organization that he/she manages?

It is also useful to know something of what budget resources this manager has at disposal and whether the up-coming year is likely to be a tight budget year or one in which more resources may be available for particular purposes.

It is, as I have argued sternly above, extremely important to be working on “real” problems and objectives as experienced by this operating manager, but it is also appropriate to prepare and kind of “crib sheet” on which you list important facts that you discover in your research, at a minimum the important change points that you have identified. 

Let’s say that you have done the research you can for your conversation with this operating manager.  The next thing is to think about how this conversation should begin, how it ideally should proceed.  It is also important to attempt to “see” how this conversation is likely to appear to this operating manager.

It is important that this conversation occur so that the operating manager with whom you are meeting talks about the real objectives and the real problems he/she is working on, or wants to, in the coming year.  It is not easy for operating managers to be open about such things and many of them will have great difficulty believing that they should be talking about the “guts” of organizational life with mere “trainers.”

You can see how they likely see this.  “Training” is usually seen to be a weak function most frequently housed in another weak function, namely “personnel.”  Why would operating management think it might be appropriate to talk openingly about real organizational problems and objectives with subordinates wh are “not at the table” of organizational decision making? 

 Many operating managers see both personnel and training professionals mostly as folks who are supposed to ”sand off the rough edges” resulting from the impact on employees of management decisions and moves, making sure that the mandated niceties are handled without impinging seriously on operations, and keeping operating mangement out of serious “trouble” as the latter operate.

I do not pretend that there are any “golden keys” available for either producing or conducting such coversations.  But I can describe one approach that has worked in a couple of quite different large organizations.

(In the meeting, the manager should be encouraged to do most of the talking.  The crib sheet should be used to ask more leading questions only if the manager turns out not to have much to say.)

(It can be useful to have a flipchart in this meeting and to work from it, recording key indications the operating manager makes.  Since he or she can see what you are writing, they can quickly correct anything you write that is not what they intend.)

You begin with “Let’s look out over the coming year and list the big objectives you want to achieve and the important problems you want to solve.”

Assuming the manager doesn’t tussle with this question, he or she will begin to list things of this sort.  Record them in summary as accurately as you can.  Ask the manager frequently if you are “getting it right.”  Avoid going into any particular objective or problem area beyond getting a clear understanding of what it is.

You now have a listing of the operating manager’s objectives and problem areas.  It’s sometimes useful to separate objectives from problems.  But the next step is to get some sense of their relative priority.  I often ask the manager, moving one at a time, to rate the importance of each of the indicated objectives on a 1 low and 10 high scale.  Record these ratings.  Next, move to the problems and ask for two ratings for each of them, first at what frequency does a given problem occur and then, separately, and subsequently, to rate the consequences of this problem.  Again, a ”1 low to 10 high” scale is useful.

You and the operating manager now have a prioritized list of his/her objectives and problems for the coming year.  Now ask the manager which of these problems or objectives seems most important for you to work on together (if you have only training skills you will need to search for the highest priority objective/problem area that is the lack of skill and knowledge).  Once you have selected a high priority area for work, it is likely that you’ve done what can be done in an initial meeting.

Suggest to the manager that you schedule a second meeting with him/her in which you can decide how to organize this project.  Say that you believe that useful work can be done in this area but that success will require continuing involvement on his/her side.

Sometimes an operating manager will suggest that the training unit take on one of the indicated problem areas largely on its own, without any investment on his/her side.  Such a suggestion likely indicates that the objective/problem, in fact,  has a lower priority than the manager has indicated (if it had a higher one, he/she would likely be quite willing to invest in solving it).  Managers suggesting that the training unit could work on its own on a given problem area may be displaying good ”organizational manners” (cooperating with  seeming request by training for work by suggesting something that may be marginally useful but likely innocuous).  I have, traditionally, excused myself (and the training function) from such suggested projects by saying something like (“I don’t think that’s an area where we can be very useful to you.”) 

This is the second sense in which it is important not be trapped into doing something “for” rather than “with” an operating manager client.  If an operating manager rates a given objective or problem as a high personal priority, but is then reluctant to invest in it, the original estimate of importance cannot be accurate.  Say to the operating manager that it’s really essential, if you are to be of real assistance, that you and he/she agree on an area important enough for both of you to invest in.

The organization and the conduct of a second meeting focused on how to organize your joint work on a selected high priority objective or problem needs an article of its own, but it is important to continue to work with this client manager for the entire duration of any project undertaken on a basis within which both he/she and the training unit are actively investing it together.

Working “with” rather than “for” operating management clients is an important distinction to be kept constantly in mind on any project.  Working “for” them is a dangerous (and I think wasteful) temptation to be avoided.