Preamble
This article was written for the 8th Anniversary issue of The Data Administration Newsletter. The anniversary theme brought to mind annual reviews, when we take time out to assess our performance –
and prepare our excuses if it has fallen short of expectations. Surely the most commonly proffered reason for failing to achieve data management objectives must be “lack of management support”.
Here are some thoughts on tackling that problem.
The critical success factor
Three years ago I moderated the closing panel discussion at the DAMA International European Conference. The panelists[1] were some of the most prominent and
experienced people in the field, so I took the opportunity to ask the perennial question: “what does it take to make data management work?” Predictably, I received the perennial answer:
“management support”. But the panelists went on to agree that the key to winning such support is genuine empathy with business managers, and that achieving such empathy is a seriously difficult
challenge. Richard Barker, who had moved from a technical management role (he was a Director of Oracle Europe and the original architect of the OracleTM Designer product) to a senior management
role in another organization, was particularly emphatic, stating that he did not truly understand the needs and attitudes of general management until he was part of that community.
I can only echo Richard’s observation. In the last ten years, I have spent much of my time working with CEOs and line-of-business managers on strategy and planning issues, as well as managing my
own consultancy business. Looking back to the time prior to that, when my focus was solely on data management, I am more than a little embarrassed at the naiveté – and indeed arrogance – of
some of my assumptions about general managers’ wants and needs.
Last year, I spent some time reflecting on the lessons I’d learned, culminating in a presentation with the same title as this article[2]. It included
observations about communication (managers don’t really understand the architecture diagrams), the nature of management (not just making decisions) and the preference for projects over permanent
infrastructure. But a single lesson stood head and shoulders above these: what managers want most of all is support for their key initiatives.
I can offer only one justification for devoting the remainder of this article to something so obvious: in my experience, the lack of management support for most data management groups can be
attributed to the neglect of this simple requirement. Managers who feel unsupported do not give support in return.
Why wouldn’t data managers support other managers in their organization? The problems seem to relate to four factors:
- Lack of understanding of individual managers’ goals
- Belief that data management objectives are important in their own right
- Desire to exercise professional skills
- Lack of perspective
Let’s look at them in a bit more detail.
Understanding managers’ goals
Most managers are focused on a small number of key objectives. Success comes from achieving these without badly messing up in any other area.
Individual managers’ objectives are often surprisingly constrained in terms of scope, authority and time. Even quite senior managers are likely to be less focused on the “bottom line” than on
their specific contribution to it. They are not likely to be expected to achieve goals that depend on the cooperation of other managers outside their direct reporting line. Nor can you expect them
to plant seeds that will bear fruit only after they have moved on, or been fired for failing to meet shorter-term targets.
The key to understanding management objectives is listening. Unfortunately, when we succeed in getting some of that all-too-rare time with senior managers, there’s a natural inclination to fill it
with our data management message. Remember the salesperson’s rule: one mouth, two ears. People who listen are more likely to be invited back, so it’s not a zero-sum game.
It’s natural – and appropriate – to listen actively with a view to identifying opportunities to use data management ideas and techniques. But there is a fine line between looking for opportunities
to help and pushing for evidence to support a predetermined approach. Just because “recognizing data as a key resource” or “making decisions based on objective information” is included in a
formal plan or policy, you should not assume that these are key goals for managers. Time and again, I have watched managers roll their eyes as someone lifts a phrase from the business mission or
strategy to support their own agenda. In one case the CEO of a large corporation was quoted in the press as stating that uncoordinated data was a major problem. Needless to say, the data management
team (not to mention associated consultants) featured the quote in their presentations for years after that. But the CEO’s words did not translate into the level of support for data management
that they seemed to imply. Why? Because an observation made once is not the same as a key objective.
At this point, you might protest that there are certain universals: most managers, for example, would be interested in opportunities to reduce their expenditure. Not so: firstly, expenditure may
not be a key goal for managers who are tasked with (for example), building market share or developing new products, as distinct from those who are managing high-volume routine operations. In some
organizations (particularly in the public sector) managers may be concerned to spend all of their budgets in order to qualify for similar funding in the future. If you are offering an opportunity
to reduce costs with no significant effort or investment, you have a reasonably compelling proposition, but proposals are seldom so simple. Managers are well used to – and thoroughly
suspicious of – salespeople who offer long term benefits in exchange for up-front investment. Don’t make assumptions about what managers want.
A manager’s key objectives are, practically speaking, non-negotiable. “You ought to be more concerned about x” is not a sustainable argument, unless you’re prepared to take it to his
or her boss.
You will have managers’ attention if you can demonstrate your willingness and ability to help with established objectives, preferably by supporting the strategies and tactics that they have
already developed. Remember that managers will usually have devoted substantial effort to their approaches, so that an alternative proposal, even though it addresses the right objectives, is likely
to be interpreted as “telling them how to do their job.” (In such circumstances, the most effective approach is not to present a fully-worked solution, but to make them aware of the relevant
tools and ideas so that they can design the better solution themselves.)
“But wait”, I hear you say: “individual managers’ goals are not the same as the goals of the organization. As you’ve said, they’re limited in time, scope, authority…” This is true: every
organizational structure is a compromise, but it represents the best compromise that the CEO can come up with. If you think a manager’s objectives are out of line with the organization’s overall
goals, then consider raising the issue with higher management. But don’t, in ignorance of the big picture, seek to impose your own view of how the organization should run.
Earlier in this section, I foreshadowed an alternative line of attack by qualifying the key goals focus with the phrase “without badly messing up in any other area”. In general, I don’t
recommend “hygiene factors” as the main target for data management efforts. The manager will need to be convinced that the risk is great enough to justify action, and, in general, will not value
such “defensive” actions as highly as those that further the key objectives. If you can establish, for example, that there is a genuine risk of prosecution for failing to provide correct
information or failing to meet privacy guidelines, and that your proposed actions will ameliorate that risk, then you should of course put that proposition to the manager. Just don’t expect to be
thanked.
Finally, in the light of the above, it has to be said that the traditional “big” enterprise-wide data administration approach may have little appeal to many managers, particularly those without
an enterprise-wide brief. However, individual sub-disciplines such as data quality management, data modeling, introduction of standards and data warehousing may be eminently useful in progressing
their agendas.
Data Management as an objective in its own right
A data manager recently complained to me that he had been unable to win support for data management initiatives across different lines of business (LoBs): the heads of these LoBs had a high degree
of autonomy, and were predictably not excited by proposals requiring cooperation. I suggested that only the CEO would be interested in initiatives that involved multiple LoBs. “But what if the CEO
doesn’t support us?” was the response.
Look at the implicit assumption here: data management is a good thing whether the CEO thinks so or not. (Alternatively, the data manager may simply have been concerned to keep his job, but in this
case, as in others, I sensed a commitment that went beyond self-interest.) Many data managers, I suspect, find themselves in this position: believing in the value of data management but being
unable to convince their organizations to make the necessary investment. Should we attempt to do what we believe is right in the absence of the necessary support? Is it reasonable to fight the
management of our own organizations?
My view is an (almost unequivocal) “no”. As a business manager, I expected my colleagues to offer ideas, and to argue vigorously for them. But if a decision was made not to proceed with their
proposals, I would have been outraged if the proponents had continued to fight a “guerilla” war – to pursue their own agenda – as I have heard advocated in some data management forums. I believe
most CEOs would agree with me, despite the attention given to “skunk works” in the popular management literature. Who’s right? The CEOs are – because they are the ones accountable to the owners
for the organization’s overall performance.
Under what circumstances might we make an exception? We recognize the role of “whistle-blowers” when management is acting illegally or unethically. Generally, failure to implement data management
would not fall into this category. But what if the benefits of data management were so well established that only a willfully ignorant organization would work against its shareholders’ interests
by failing to invest? “What if” indeed. The argument is academic, because data management simply does not have that standing. There is no compelling evidence for the value of data management
functions in providing competitive advantage or preventing organizational failure despite the rhetoric of members of the data management community. If there was, we might have less trouble
convincing management in the first place.
The answer I gave to my data management client with the potentially unsupportive CEO was this: if you’ve had your opportunity to present the data management story and he doesn’t buy it, then
accept that, at least for now, this organization is not going to implement data management across the LoBs. The CEO has made a decision (rightly or wrongly, but on the facts available to him) that
LoB independence is more important than data management. Look for other ways to help the business.
Before we leave this topic, let’s return to the “skunk works” argument. Isn’t it true that, as George Bernard Shaw said, “all progress is made by unreasonable men”? Isn’t there room for the
one-eyed supporter of an idea who regards failure as only a temporary setback, and who presses on regardless of management’s indifference or opposition? Not in my organization, there isn’t –
unless the possible fruits of the idea are central to the organization’s mission. Managers value innovation only when the costs and risks promise a commensurate return. That’s why “skunk works”
stories are typically about new products and not internal infrastructure.
Exercising professional skills
If the previous point was about loyalty to the higher calling of data management, this one is more selfish. Most professionals want to use their skills in a worthwhile cause. “To a child with a
hammer, everything looks like a nail” and few professionals in any discipline are immune to framing approaches in terms of their own specialization. If you see yourself as a data modeler, then you
naturally seek out opportunities to do data modeling.
Conversely, managers’ focus is on the goal: the means is secondary. “Do whatever it takes” is the essence of the manager’s approach. In my experience, this difference in perspective is the
single greatest source of friction, disrespect and distrust between managers and professionals.
In the data management sphere, it has meant that data managers who are fixated on particular approaches have missed opportunities to make major contributions to the business – because business
managers have found another way to achieve a “data management” goal. CRM applications, data warehouses, and EAI initiatives[3] all address the very goals
cited by data management advocates as critical – but are often assigned to other groups. In fact data management groups have sometimes turned away from these “compromise” approaches because they
did not fit the prevailing data management philosophy.
A particular challenge for data management groups, then, is to decide whether they are primarily the providers of a particular set of skills, to be deployed as management sees fit, or the solvers
of particular types of problems. If the former, they should not be too choosy about how their skills are deployed. If the latter, they need to be prepared to do “whatever it takes” and recruit
accordingly. A group that chooses the “both” option and positions itself as “solving certain kinds of problems using certain kinds of skills” risks being marginalized unless it can establish
that their skill set is the only effective one for solving the problems.
Perspective
When I run consulting skills courses, we begin with a competitive exercise involving the building of paper planes. There is often some debate over the selection of the winner – debate that can
become quite intense and threaten to disrupt the remainder of the class. The appeal I have to make, of course, is to “keep it in perspective”. Tomorrow it won’t matter who won the T-shirts or
who was the better plane builder – but when you are intensely focused on a goal or a task, it can be extraordinarily difficult to maintain that perspective. (Sports fans, take note!)
As professionals focused on data management goals and tasks, it is very easy to lose, or never have, a sense of how important they are within the bigger picture of the organization. On the one
hand, we want to make the strongest case for their importance; on the other, we need to be balanced advisors rather than fanatics or prophets of doom.
Data management people have a reputation for being one-eyed. Regardless of other business priorities, they can be expected to push the data agenda. “And why not?” they ask. “We’re the data
people, and we’re expected to highlight those issues. Others can make the final call when all the facts and opinions are in.”
I don’t buy this argument. First, it characterizes the data people as lobbyists rather than advisors. (I hope my surgeon will at least alert me to other options rather than pushing for surgery
because she’s a “knife” person). Second, it creates a vicious circle: data people’s unbalanced advice is rejected or compromised, hence they feel they have to push even harder in the future.
Finally, it sets the data management person against other stakeholders – against the spirit of “support for manager’s key goals”. The term “data bigot” may have its place as an amusing
self-description in the company of data management peers: but few managers want to employ people who are bigoted in any way, except, of course, towards their own objectives.
That said, and to return to Shaw’s “unreasonable men” quote, I believe that there is a place for single-mindedness in our profession, and that we need some fanatics if we are to move forward. We
need the Chris Dates, Larry Englishes, Clive Finkelsteins, Bill Inmons, Ronald Rosses, and John Zachmans[4] to take a “full speed ahead and damn the
torpedoes” approach to developing and promoting their particular ideas. But in their cases, we are talking about their core business – not a support function. In an earlier article[5] I criticized some supporters of the Zachman framework for being too zealous; but I would not make that criticism of its originator. We need him and his colleagues to
push it as far as it can go – and beyond. But those of us working as employees or consultants need to be able to take all of these ideas with the proverbial grain of salt, adapting them so they
support rather than dominate our organization’s goals.
Finally, data management people frequently assume that I have used my strategic planning sessions with senior managers to promote – or at least make them aware of – data management goals and ideas.
Indeed I have, but I have also learned that data management issues (and even IT issues) are a relatively small part of the challenge facing most executive teams. To bring in the ideas from time to
time when the context suits is add value: to bring them up constantly is to demonstrate a lack of empathy with their agenda.
What about me?
The above might seem to paint a depressing picture for data managers. This may be true for those who are on a “mission from God” with a clear vision of what they believe is right for their
organization, and a determination to implement it regardless of whether the organization wants it or not.
But for those who are prepared to see data management as a set of principles, ideas and techniques that, strategically applied, can directly help organizations and individual managers achieve their
goals, the news is not so bad.
I recently received an email from a data architect who had consciously re-oriented her efforts towards supporting managers rather than carrying out the data management mission. First she
acknowledged that there was indeed a change to make. But she went on to say that her level of influence had grown substantially, that others asked for her contributions, and she was looking forward
to going to work in the morning. And, of course, her improved standing in the organization means that there is a greater likelihood that they will eventually end up with some version of big data
management – if, of course, it helps them achieve their goals.
The third Edition of Graeme’s book Data Modeling Essentials (written with Graham Witt) was published by Morgan Kaufmann in November 2004. Graeme is as outspoken and annoying on that subject as
he is on Data Management.
[1] Peter Aiken, Richard Barker, Michael Brackett, Michael Ferguson, and John Zachman.
[2] Keynote address at the (Debtech) Fall Data Management Summit, New Jersey, September 2004.
[3] And, historically, Y2K initiatives
[4] I chose these people as examples because of their influence in their particular areas rather than as particular illustrating fanaticism. I deliberately
excluded data modeling…
[5] “What’s Wrong with the Zachman Framework?”, TDAN, 1st Quarter, 2005.
