Regarding the long-suffering demand that HR be granted a seat at the table, I’d like to address this issue using my work levels goggles. Is it really about HR needing to acquire this competency or that competency, or is it about organizational structure?
I’m Talking About a Strategy Table
If we assume that “the table” is a place where strategy, not current operations, is discussed, then seats at the table must be filled with people capable of thinking strategically. Unfortunately, most competency models don’t have a valid or reliable definition of “strategic thinking capability”. I’d like to offer one.
Strategic Thinking Capability Defined
Using work levels terminology, this means people capable at level 4, i.e., those capable of mentally managing,planning, integrating, balancing and coordinatingmultiple serial pathways to deliver goals with delivery times falling betweentwo and five years.
Level Four Work Example
An example of level four work I offered in a previous post was:
Expand our sales footprint beyond the US by establishing a sales force in Mexico which should be responsible for 15% of total sales at the end of four years. (i.e. Integrate multiple serial pathways: recruiting, staffing, facilities, technology, Mexican human resource law and customs, customer identification, sales process, product offerings, marketing, advertising, operations, delivery, warehousing. etc. to deliver a four year objective)
You’ve also heard the complaint that no one will hire you without experience, but how can you get any experience if no one will hire you? HR is entrapped within this classic, catch-22 dilemma.
If your only experience with HR has been in an organization where HR was relegated tocutting paychecks, planning picnics, and offering soap to smelly people, then you might not be able to appreciate the value a strategic HR role could add to your organization. Picnic planners need not be in your strategy sessions.
We’re back to work levels, not all HR roles are created equal.
You Get What You Design and Pay For
Assuming your organization is of sufficient size and complexity to merit a strategic HR role, in order to attract strategic HR talent, you must create a strategic, level 4 HR role, embed it with level 4 work and then pay accordingly. No organization is going to pay $120K for picnic planning, and conversely, no strategic-capable HR professional is going to take a $40K, picnic-planning role.
If You Build It, They Will Come
To grab a line from the movie Field of Dreams, if you build it, they will come. If you built a role that was tasked with creating an integrated recruiting, screening, selection, compensation, development, performance management, managerial leadership system to build a talent pipeline capable of staffing our organization to meet our strategy 3 to 5 years down the road, you wouldn’t dream of excluding that person from the table. And your applicants fora rolethat was scoped at thislevel of complexity and specificitywould well deserve a six figure salary.
To Seat or Not To Seat
If your organization is of sufficient size and complexity to merit a structure containing a truly strategic HR role,then it would beimperative to have them at the table. If HR in your organization isn’t structured to add strategic value, save the seat for someone else.
I’m OK. You’re OK. Let’s fix the system.
Is HR in your organization strategic? Should it be?
Some people resist the notion that cognitive capacity is constitutional – that its maturation cannot be affected by any intervention. They may have what they consider to be evidence to the contrary, that with proper managerial assistance, an employee’s cognitive capacity can be raised.
I wish to explain here how I understand RO would explain such occurrences. Consider an employee with Str-II judgment capacity in a Str-II role for which they lack skills and knowledge. Let’s say the role has to do with building staff capability within a business unit. The question is how the manager can support the employee. We’ll start with looking at results of having no support and then the consequences of increasing support.
You interview the employee about building staff capacity or you ask their manager about their performance, and everything shows the employee to have Stratum-I capability for this work. You ask them what must be done to improve staff capability and they say, “Hire better people”. Anything else? “I guess we could train people better.” Anything else? “Hmmm. It might be useful to do a survey the capacity we currently have. I hadn’t thought about it.” The answers are declarative. They have no experience in this subject matter. It has never occurred to them how hiring the right people and doing the right training are both necessary and must be combined to get the best capacity. But if you assess their judgment capacity by asking them about retail, where they come from, they will tell you about the importance of the right mix of price, style and quality and can tell you in situations new to them how to get that combination. They can tell you how to combine service and quality to maximize revenue. They would show cumulative processing and prove themselves capable at Str. II. But what they accomplish in their current role is not as much as one expects from a Str.-II-capable employee.
First Level of Support
Now they get the first level of support. In the most passive form, it is exposure to facts and methods; the employee observes peers approaching work in a particular way and copies it. But a more effective and efficient approach is to send them on training programs and for their manager to give them coaching so they start thinking about how to combine hiring with training and even how to combine a number of elements to maximize the effectiveness of training. Through either approach – the more passive observation and copying or the more directed training and coaching - they acquire templates and methods that enable them to apply their cumulative-processing ability to their current work.
Their performance has gone up a stratum, not because they have just acquired Str-II judgment capacity – they already had that – but because the experiences they have give them the skills and knowledge that enable them to apply their Str.-II judgment capacity to their current work. And this would also show in an interview of them regarding building staff capacity as well as on their performance. When an observer says, “We have raised their ability by a stratum” we need to be careful. You have raised their performance and their applied capability by a stratum but you have not raised their judgment capacity. That was and remains Str. II. They could and can exercise judgment using cumulative processing.
As an aside, consider their over-promoted colleague in another department, one who is actually only Stratum-I capable. Exposed to the same training, the second employee will learn that recruiting is important and training is important. But they can only think of these as separate interventions. They are in danger of a) recruiting people who are in roles that do give employees the time to take training or b) designing training that is not appropriate to the people being recruited. Our Stratum-II-capable employee, on the contrary, will implement recruiting + training as a combined intervention.
Second Level of Support
Enter the next level of support. Through training, coaching, software or some other support, the employee learns a serial method:
This method produces better results than before. It may even produce better results than you would get from someone capable at Str-III who hasn’t learned this method, particularly if the serial method is learned richly. With the manager’s Str-III guidance, our Stratum-II capable employee may acquire variations of the methods refined to solve particular types of problems, and if the variance in the environment is small enough, this density of skill and knowledge may practically substitute for judgment. An interview with them might rate their performance at Str III working on issues related to development of staff capability. Someone assessing their judgment capacity by listening for structure could hear serial processing in “How do we build staff capacity? Well, first we survey the capabilities we currently have and will shortly need and then we try to determine the gaps between them. That tells us how to hire people who can be trained to fill the gap completely, people with the needed judgment capacity who might just lack some skills but who are interested in learning those skills. So we can then train the new hires to be fully capable of doing what we need.”
But all they have acquired is a method, and perhaps a way of describing the method, something they could have learned when they were Str-I capable. They can follow the method step by step. But they cannot design the type of survey of current capabilities that will facilitate the type of description of the gap that will facilitate the hiring of just those people who can, from the type of training you will provide, be trained to fill the gap. (Again, sufficient density of skills and knowledge may in some circumstances practically substitute for that judgment.) The method will likely improve the results of the cumulative processing the employee does, but it would be an error to call it “serial processing”. RO’s focus on judgment capacity sometimes puts skills, knowledge and even performance in the shadows. We sometimes do not probe sufficiently to determine whether the observed series is created by or repeated by the employee being assessed.
The method, richly as it might be learned, is still followed mechanically. When the employee matures into Str-III capacity, they have insights about it. “That’s what my manager was trying to get me to understand!” At that point, they can fully use the serial method with serial processing and apply it effectively in new situations.
There is another possibility here for how the employee and manager might work together. The employee submits their plan to their manager who then tweaks it and enhances it, describing how to:
If this is what is going on, it is important to describe this situation accurately. The work is the manager’s work. The performance is the manager’s. The 18-month task of getting the staff capability we are looking for is the manager’s task.
Third Level of Support
So the MoR (manager once removed, manager’s manager) says to the employee, “It’s great that you are building our staff’s capability. But you know at the same time we are going through a series of moves in our marketing organization and another in our sales department to increase the number of customers. I need your staff capacity building to be integrated with these other two processes. As you are a high-potential employee, I want to work closely with you on this and have you coordinate all of these processes. I will walk you through it all step by step.”
You can call this “support”, but be clear that you are supporting outcomes or supporting the organization but you are not supporting the employee. Depending on how detailed the “step-by-step” walking through is, the employee might be able to get away with Str-I declarative judgment, Str-II cumulative judgment or Str.-III serial judgment. There is no sense in which one could say that the employee’s performance has risen. The employee’s performance is at whatever level it was before. The parallel processing is all done by the MoR. This may be a useful for the organization. Certainly organizational performance will be enhanced; it is better for the organization if the employee’s work in building staff capacity is coordinated with marketing activity and increases in sales. If you can first bring in new clients who can be served by existing staff capability, then new clients who can be served by unskilled new employees, and then new clients who will most benefit from newly-trained staff members, profits will be maximized. But we would be deceiving ourselves to label the support as supporting the employee.
If the MoR is “walking individuals through the task by moving them through each step of solving a problem”, then this task is the MoR’s, not the employee’s.
Here’s where things net out:
In this case, we must recognize what the work is that the employee is doing. If the task is to create a process through which to serve customers in sector X, and if by “process” we mean a series of steps, each leading not only to the next but also to the steps after the next, then we must recognize that the employee is not working on that task. They cannot work on that task. It doesn’t fit in their head. If the task is to provide service, and if that service can be provided by filling in the blanks in a general process already developed by their manager, then their manager has developed a method that allows the Str.-II capable employee to produce output previously requiring Str.-III capability. (As an example of this general concept, there is software that allows the Str.-I-capable employee to do insurance sales work that previously required Str.-II capability.)
I suggested understanding support as coaching, training or use of an employee to conduct sub-tasks that makes better use for the organization of the employee’s abilities. The point of this blog is that it is useful to understand who does what work within different categories of support:
While output may be increased by any of those methods, none of them increases the employee’s ability to exercise judgment, only to make better use for the organization of that ability to exercise judgment.
(This statement is part of a Participation Agreement that all of the Society's core affiliates sign.I hope that you will read it carefully.Your comments are welcome. Please post comments you feel will contribute to a quality public dialogue below. You may send any questions or private comments to me personally at
While GO Society’s associates and fellows agree to support the Society’s purpose, we come to our organizational work from different disciplines, different sectors, different functions, different roles and levels of organizations, and different interests in improving organizational effectiveness.
What we hold in common is appreciation, knowledge, and experience in a systems approach to designing and managing organizations that is based on foundational concepts about understanding and organizing complexity in work settings developed by Lord Wilfred Brown and Elliott Jaques. This approach, over the years variously called “The Glacier Project”, “Stratified Systems Theory”, and “Requisite Organization” includes well-defined, researched, and tested concepts of levels of work complexity, levels of human capability, accountability, and effective managerial leadership practices.
We recognize the discovery that organizational systems have a direct and substantial impact both on the personal achievement of people at work and on the capacity of organizations to create wealth for society, and that as a consequence of this the design and implementation of such systems carry strong moral and ethical implications.
While we all appreciate this systems approach and these concepts, we again differ as to which and how many of the concepts we emphasize in our individual practices and how we may integrate them with other theories and skills that we use in our organizational work.A few examples:
In summary, we all have and celebrate our own educational foundations, skilled knowledge, and experience in improving organizations, and in this document we all agree to our common science based principles and concepts described above in general terms.
We also agree to participate in a continuing dialogue on the continuous evolution of these ideas both in a private area of the web and in face-to-face meetings, helping GO Society associates and fellows to come to broad agreement on what specifically is included and is not included in this commonly held requisite approach, and what we should encourage all Society affiliates to endeavor to master.
Many organizations have struggled with the problem of silos, how to coordinate the work done by different functions. Common solutions such as “dotted-line” relationships or matrix organization often leave employees confused about who their real manager is or what to do when confronted with conflicting orders.
RO provides a very useful strategy-driven solution to the problem, TIRRs (Task Initiating Role Relationships) that makes an employee accountable to serve the interests of a department they do not work in while still having the employee accountable only to their own manager.
A TIRR can give Employee D authority to have Employee E do something (consider advice, provide a service, coordinate their work with others, or take an action immediately) or to ask or tell Employee E not do something (not step outside of policy or not interfere with a strategy). What is critical about this arrangement is that it is Employee C, the manager of Employee E, who holds E accountable to work within this arrangement.
A question came up recently in a project I was working on regarding who needs to determine what the relationship should be between Employees D and E. My colleague and I agreed that that the conversation could start at any level:
We also agreed that ideally, B, C, D and E should all have their opinions solicited and considered. The question was, if B and C came to agreement, need they get A’s approval? I put the question to a number of colleagues, and the most articulate answer came from Michael Anderson:
I think that A has to see the results of whatever B and C come up with, to ensure that the overall pattern of cross-functional relationships is consistent with how he/she wants to integrate work between B’s and C’s functions . If sufficient context was provided in the first place, this would probably be  a formality and  an opportunity to reinforce some key messages and have a deeper conversation. If B and C got it wrong, then clearly the A has to intervene and override.
So, at the end of the day, A has the last say. I guess that makes him/her the decision maker, but at the end of what ought to be a rich and inclusive process.
One way of framing the debate is: Is the design of TIRRs the work of B and C, or is it A’s work? In principle, I’d say that it’s intrinsically A’s work, because that’s the point of integration. If that’s the case, the debate is more about how much authority A chooses to give to B & C and how much he/she chooses to hold back. And that’s more a judgment of what works in a given situation, rather than a pure point of principle.
And that is what makes sense to me. The conversation can start anywhere and it is important to obtain and consider the advice of all of the parties involved, but the working relationship between D and E must be decided upon by A.
La Global Organization Design Society ha inaugurado una sección sobre minería y manufactura pesada en su sitio en Internet. Entre las numerosas aplicaciones mundiales de la Organización Requerida, el trabajo realizado en ACINDAR, en Argentina, es una de las referencias más significativas. Los visitantes de habla hispana tienen a su disposición presentaciones y entrevistas en video a varios de los protagonistas principales de este proceso, comenzando por el CEO Arturo Acevedo. Quedan cordialmente invitados.
The requisite organization (RO) model of human capability has four components:
Originally (as I recall), -T was defined as negative temperament. And capability in role depended on the absence of –T. The concept behind –T is that there is no particular temperament required for any role but there are temperamental issues that are counterproductive. At core, an individual cannot just will –T away. It is most often raised in relation to issues like alcoholism, uncontrollable rage or other dysfunctions that result in abusive or anti-social behaviour; but it also applies to other dysfunctions such as anxiety that is so strong that an employee becomes too nervous to deal with others. Dealing with -T typically requires some kind of therapy.
Most technical terms within RO are used in the same way by most practitioners. But –T is used in several different ways, so I’d like now to explore three interpretations of the concept of –T. As always, I am presenting my own point of view on this issue and am making no claim that this is the correct point of view.
1. –T as antisocial behaviour
When someone with some familiarity with RO tells a story about an employee who treats others abusively, they will often describe the behaviour as a “minus T” problem. I find this to be an unfortunate use of the term for two reasons.
First, -T is an aspect of a capability model whose purpose is to explain what underlies behaviour. The behaviour itself is not –T.
Second, and more important, capability may not be at the cause of the behaviour. The employee may consider their behaviour to be acceptable. It is only after the employee has been directed, coached and held accountable by their own manager to treat subordinates with respect that we should seriously consider whether continued abusive behaviour is a capability issue, an indication of -T.
I have been called in by managers to coach a subordinate who treats others abusively; typically, I find that the manager brings me in so as not to have to deal with the unpleasantries of holding someone to account. My intervention, then, is not with the employee but with the manager, reminding the manager that it is their own accountability to explain to the subordinate how they are to treat others, tell them what they must start doing and stop doing, and hold them accountable for behaving in that manner. Almost always, this intervention solves the problem. I doubt that I am the only consultant with such experiences.
For me, abusive behaviour is, itself, not –T nor is it proof that an employee is plagued with –T.
2. –T is inability to behave in the required manner
Elliott Jaques changed his definition of –T several times. The last definition I am aware of is that –T is the inability to behave in the required manner. I have heard that he changed from previous definitions so as not to encourage managers to engage in amateur psychoanalysis. But I have problems with this definition, too.
There are several reasons why one might not be able to behave as required.
The ability to behave in the required manner depends, in part, on cognitive capacity and on skills and knowledge. These are already elements in the model. The elements of a good model do not overlap with each other. This leads me to expect that there is a way to further clarify what underlies the ability to behave as required.
3. –T as an overpowering value
I think of a value as an attraction to or repulsion from a gerund (the form of a verb ending in “ing”). To say that one values art is to say that one is attracted to owning art or looking at art or making art. It strikes me that this is what is at play in –T. In the early 1990s, I said to Dr. Jaques that I thought of –T as the inability to will oneself to do what one really wants to do. I was thinking of issues like alcoholism where one cannot will oneself not to drink or extreme anxiety where one cannot will oneself to interact with others. He said that sounded right to him.
For me, this puts –T in the realm of values. One is so strongly attracted to drinking alcohol or so strongly repelled by interacting socially that one cannot simply choose not to drink or choose to socialize. The value is so strong that it overpowers the will.
This third formulation of –T reduces the capability model to three parts:
But it puts a somewhat different perspective on the third component, values. I always understood the RO perspective on values simply as the need for an employee to value to work in the role sufficiently to commit oneself fully to it. Part of this notion, as I understood it, is that if there are parts of a role you do not like (e.g. certain detail work or social obligations), you can do those disagreeable aspects unless you are bound by –T. In other words, if you value the analytic work in a role sufficiently, you will be able to engage socially in teams, make presentations, attend social functions etc. unless you are so socially inhibited that your condition would be categorized within –T.
I now have a different point of view on this for two reasons.
First, I believe that while values may be so strong as to be compelling or so mild as to be resisted they may also be in a middle range. This middle range consists of values that one can choose to resist, but not on an ongoing basis.
My second issue in regard to what one will do in order to have the work one wants relates to a basic motivation I learned about from Mike Jay, utility. Some people seem wired to do what is required to obtain what they want. If they value B highly and do not value A, they will still do A in order to get B. Others are wired with low utility. I will pretty well do A in order to have the experience of doing A and am unlikely to do A (if I disvalue it) in order to get B.
I find the issue of values under-developed with RO. It is as important an aspect of capability as cognitive capacity but has not received the same level of exploration. Perhaps it is an important next step in the development of the field.
Wilfred Brown and Elliott Jaques spoke and wrote about their original concept breakthroughs as “The Glacier Project”.
In my previous blog, I used a scientific/engineering approach to address the question of why an employee’s manager should have greater cognitive capacity than the employee. (I fully recommend Harald Solaas’s further posted comments on the topic.) In this blog, I’d like to complete the analysis and show why the manager should not be too much more capable than the subordinate. Specifically, why is it optimal for the manager’s capacity to be just one stratum higher? I will then use this as a case study to address Nicolay Worren’s posted concern about “the specific scientific studies that support the various hypotheses of RO theory”.
As before, I will use a scientific/engineering approach, using Jaques’s characterization of engineering as “an art grounded in a science”. “Place the manager one stratum above the subordinate” is an engineering template. The scientific grounding comprises the laws of cause and effect that justify it.
So let’s explore this question in the context of a Stratum-I-capable employee and see the difference it makes whether their manager is capable at Stratum II or III.
Why should we not allow gaps?
As I noted in my previous blog, because they have current potential capacity at Stratum I, the employee is capable only of declarative processing at the normal adult level. That is:
As previously noted, if the manager’s current potential capacity is at Stratum II, they would also be capable of cumulative processing at the normal adult level. That is:
The manager’s greater capacity gives them a higher perspective on the employee’s abilities and ways of solving problems thus enabling better coaching.
But if the manager is capable at Stratum III, they are also capable of serial processing, of getting to Step C by taking Step A that leads to B that leads to C.
This excess capacity serves the manager well in their own role, but it tends to hinder their managerial work in three ways:
In short, such a gap between manager and subordinate has a tendency to lead to:
What is the quality of the evidence?
Nicolay Worren’s post raises two issues.
The first is that “it is quite common today to find organizations where the manager has a cognitive capacity below his/her subordinates. I agree it is not ideal, and probably is even dysfunctional in some cases, yet some of these organisations are highly successful despite this dysfunction. How would you explain that?” Most of the engineering templates in RO specify what is optimal, not what is necessary. So working outside of those templates does not predict total organizational failure but rather a suboptimal combination of effectiveness (reaching strategic targets), efficiency and trust. I expect this is true of engineering templates in many fields. The typical result of a car’s not being tuned up to specifications is that the car runs inefficiently or does not reach top acceleration, not that it won’t run at all.
In some cases, a Stratum-III-capable manager may be skilled and personally motivated to explain things in a way that work for a Stratum-I-capable subordinate. But even a dysfunctional III-I relationship need not be fatal to the company. A company in the right business at the right time can do very well even if it is poorly structured, staffed and managed. (As a client in the financial sector said to me about the good times, just as they ended in fall 2008, “In a hurricane, even a turkey can fly.”) And a company with a good strategy will beat its competitors if they are no better managed and have weaker strategies.
Worren’s second issue was in regard to scientific evidence. He said that he feels “the reader should be informed about the specific scientific studies that support the various hypotheses of RO theory. If one tried to search for the ‘science behind the rule’, as you said, I think one would find that some of the RO hypotheses are strongly supported, others partially supported and that yet others lack any empirical support at all.”
In the case of the template discussed here, the scientific research might be conducted in an experiment in which work is given to pairs of people:
The tasks might be set up to simulate manager-subordinate relationships and the object would be to see how the dynamics differ.
I am not aware of any such research so I asked Ken Craddock. He replied that he expects that such “experiments have been done within firms - but the results are proprietary”. He suggested that the US Army has conducted such research, but I have not personally tracked it down.
In my case personally, while I have been recommending this template to clients for over 15 years, it is not because of any scientific evidence I have been aware of. To me, this illustrates the point that scientific grounding is an aspiration for RO, not a fully-accomplished mission. The difference between RO and most other approaches is not that we have, in fact, given a scientific grounding for all of our templates but that we value such grounding. It is not enough for us to say that companies are more successful when they place managers one stratum above their subordinates; we seek to find the underlying science. I find it understandable that we have not yet been able to do the scientific research to substantiate our hypotheses regarding why our templates are effective. Perhaps we can do a better job of a) explicitly linking templates, when we recommend them, to research that has been done and b) clearly stating when we are not aware of any such research but are basing our recommendation on our own experience.
Here is my theory, managers don't skim books because they are busy, they skim because most of the content is not useful to them. In my own case I usually skim books mercilessly but if I’m reading a genuinely deep book (e.g. The Secular Age by Charles Taylor) then I read each sentence very carefully; even re-reading pages that were particularly important.
How do we write for managers so they get value from every word?
A History of Bad Examples
I think part of the problem is our own experience with books. Many of the books we read are written for novices. A popular history book like Guns, Germs & Steel assumes you don’t have any real background in the subject. But if we are writing about management for managers then they are deeply immersed in that world and a style suitable for novices won’t work for them.
Another bad example is books written by one technical expert to other technical experts. In this case it’s natural to exult in the jargon, shared mental models and obscure references of the field. Imagine one computer geeks writing to impress his peers, there is no way he would stoop to writing in plain English that could be understood by the masses. Managers are expert in the practice of management but not RO. The book you write on RO for managers is not the one you would write for RO consultants.
So as authors we may be tempted to write for novices or for technical insiders because that is what we have seen before. We need to push those bad examples aside if we are to write a book managers want to read.
The Manager in Mind
The easy solution to writing for a say stratum IV manager is to have someone at that level in mind as you write. Think of the process not as writing but as explaining it to them. Think of it almost as a conversation where you explain something, imagine the question they’d ask, and then go from there.
You sometimes hear that someone writes in a ‘conversational style’; I’ve never seen that presented as being a bad thing.
Diversity of Readers
The challenge of course is that there is a diversity of readers and not all managers, even at the same stratum, will have the same background knowledge and interests.
Here are a few tips with dealing with a diversity of readers:
It's worth remembering that the normal style of books evolved in a pre-Wikipedia era. Anyone who feels they need to know more about time-span of discretion can find it online—but it's nice if they have the micro-explanation so they don't have to drop your book right at that moment.
Writing up a case study on a successful project can serve many purposes. It can be a stand-alone marketing piece, a chapter in a book or part of your training material. However, if you are going to write a case about a client you normally need their cooperation. How do you get that?
First, Do You Need Client Cooperation?
When a consultant is brought into a client they implicitly or explicitly agree to protect the client. If the client has learned how to do something well then the consultant should not share that with competitors; if a client has done something embarrassing the consultant has an obligation to keep that under wraps. There is a pact of confidentiality.
In some cases it might be alright to do a disguised and anonymous case without a client's explicit consent. You will have to be the judge of whether the disguised case breaches the implicit pact of confidentially—but err on the side of caution.
A disadvantage to having a disguised case is that it may feel less credible to readers than one about a named organization. An advantage is that in an anonymous case you may 'smooth out' the story line more easily than in a real case. For example, if one of the HR managers retires mid-way through the case that may be irrelevant to the tale. In an anonymous case you may simply refer to the 'HR manager' and skip the unimportant retirement. With a named client you may feel compelled to mention the manager’s retirement and say a little about their replacement.
The test as to whether you need client approval is simply whether anyone is likely to get upset. If the case is disguised but the client thinks “I know you are talking about me!” then you may have a problem. Think hard about the people involved and how they may react when they see the case you've written. It's always better to ask for permission even if you are sure no one would recognize the organization nor learn any trade secrets from your case.
Who Has to Agree
If you do want to name the client, or have them actively involved in helping you with the case, then for RO interventions you probably need to the approval of the CEO. RO projects are usually big enough and deep enough that you need to go right to the top.
Obviously the case will have to be primarily flattering to the CEO or there is little point proceeding. If you do get the CEO’s support you may think that's all you need but because RO projects range wide you should seek out the support of everyone involved. You won't need their formal agreement if the CEO is onside but you should make it clear that you will look out for their interests and make sure no one looks bad.
It is also a good idea to ask to be introduced to the PR people. They have a stake in anything written about the organization and will appreciate being included. Furthermore the can be of real assistance, perhaps even offering their editing skills. They may want to re-use the content in their own way and you should encourage them to do so.
Who Controls the Content?
The potential downside of a case write-up for the client is straightforward—no one wants to look bad. People can be quite sensitive to nuance so this is not a minor issue. Furthermore RO projects involve big change and that usually means resistance to change and no one will want to be portrayed as one of the dinosaurs who was working against the better interests of the organization.
One way to minimize this fear is to go back to making the case somewhat anonymous. The client may still be happy to be helpful, but won't worry as much about the downside. “Somewhat anonymous” could mean that the organization is named but the individuals are disguised—you can tailor how much is revealed depending on the situation.
The other step is to promise the client you will show them the case before anything is published. That gets us to the question of who owns the final content. The answer is straightforward; they have the authority to veto and the ability to make suggestions for changes but the ultimate content is yours.
Vetoing does not usually create problems; the bigger problem occurs when someone has suggested some changes that you don't like. Either they are badly written or if the PR department is involved they may insert marketing fluff. The trick is to take some of their changes to show you are responsive; but not to cede control over the writing.
What's the Up-Side for the Client?
Most people are happy to be supportive of your efforts to write a case. However, you should be clear about what the upside is. Here are some points:
Understand what will appeal to the CEO and other key players, make the pitch, and allay their fears by promising they can review and veto the content.
Cases are great but you will normally want client approval. While they are likely to be supportive don’t take that support for granted and make sure all your ducks are aligned so that no one is upset by your case and no confidentiality is violated.