“Managing by the Numbers”

A recent article in Business Week is titled “Managing by the Numbers” and it focuses on IBM’s attempt to build a system to assign staff to appropriate engagements around the world. Rather than relying on manual processes, a central planning group is gathering capabilities for each of their employees and attempting to let the computer match skills to opportunities.

I have a lot of experience in consulting at a number of different organizations. If you are interested in the challenges (and opportunities) of running or being part of a professional services organization, I suggest that you read “Managing the Professional Services Firm” by David Maister.

The types of examples given in the book are staffing a web services engagement in the Philippines; there is an expensive (high ranking) consultant on the bench (meaning – unassigned and not currently earning income for the firm) in a faraway country vs. a less skilled (and cheaper) local consultant who could also be assigned to the job – which to choose? This is the type of “problem” that the program is supposed to solve. The implied conceit is that consultants are interchangeable, and you can just build a team out of individual skill sets, have them show up at the work site, and pull off the engagement.

When I worked at one of the large consulting firms, in order to save space, they went to a “hoteling” concept. Since consultants were usually on the road and not in the office, some bean counter figured that it would be cheaper to not give anyone a permanent office and just have them occupy whatever space was available on the occasion that they had to work in town. The company did attempt to link your phone to your location and sometimes even had a nameplate ready for you, along with a little cart for your office supplies, so you were able to get started working with a minimum of effort. The company only had to have office space for the people likely to show up, which was maybe 25% of the total staff on a given day, saving them in rent money.

What is interesting is that both of these examples treat consultants in a similar manner; as cogs in a machine, or interchangeable parts. This is a terrible assumption; consultants are human beings, and they don’t automatically function as a team just because you put people in the same room and give them an overall objective (a consulting engagement). In the case of technology consultants, which is most of what IBM offers, the problem is even more acute; even a casual reader of Dilbert or viewer of the classic movie “Office Space” knows that technology staff often have poor or difficult interpersonal skills, and so throwing together random staff in a room and assuming it will work out smoothly is just bananas.

As far as the office hoteling went, it was a disaster. Many times I went to the office and I knew absolutely no one; every time I sat next to a random cluster of refugees. The only constants were the executives (they had permanent offices); a few times I attempted to converse with them but it was pointless as they viewed me as a random intruder. There was no camaraderie; no team spirit; I just started working from home if I wasn’t on site (which was rare).

Why would a business that is ostensibly about people (the famous line that all their assets walk on the elevator each night) not even think for a minute about the basics of getting people to communicate, work together, and focus on teamwork? The article never really touched on this “broken” aspect of their business model.

Another element of this business model that is terrible is that it ignores the reality of time and space. If you are running a job as a consulting manager in say, Memphis, your staff comes in from around the country on Monday morning, bedraggled, and everyone finally is sitting in their seats around noon. No one works on Friday at a job site anymore; they all start leaving on Thursday in the early afternoon. Thus, best case, you get 1/2 day Monday, Tuesday, Wednesday and 3/4 of Thursday. This totals up to 3 1/4 days out of a 5 day week; but of course you charge the client for 5 days anyways.  In addition, travel costs money – typically you’d add 15% – 20% onto total charges (client billings) for travel – and while this money is expensive to the client and costs the same to them as “consulting dollars”, it doesn’t add anything to the bottom line of the consulting firm.

The assumption also is that these consultants, who arrive on the job site not knowing one another, have some sort of common knowledge base and training, or understanding of the processes needed to run the engagement. This is usually wishful thinking; central training is not very robust, and just keeping up with the barrage of memos, expense policies, and personnel reviews consumes a lot of time (and adds nothing to the quality of the engagement).

What is the alternative to this model? One alternative is the “local office” model, where a local city like Chicago would hire their own staffs based on the needs of their customer base. These consultants know each other, and learn to work like a team. The local office invests in training unique to their needs, and staff can have local mentors. Travel is less painful; 5 days of work is received by the client rather than 3 1/4, and the staffing levels are much less variable (sometimes people don’t make it in at all in bad weather).

The “local model”, however, has problems – it is impossible to staff locally in a manner that leverages everything that is on the web site for services that IBM offers.  For example, the web site touts services from supply chain management to specific types of technological consulting.  On top of this, IBM claims industry expertise in a variety of areas, meaning that they need to provide 1) specific technical expertise 2) specific industry expertise 3) project management / consulting partner leadership in order to successfully complete the consulting job.

Thus in many instances the local office will be forced to bring in staff from outside the geographic vicinity in order to complete the job (as advertised; they could offer a more limited menu of services, but this typically isn’t done), and some other bean counter figured that it would be more efficient to leverage idle staff sitting elsewhere rather than to leave potential billing dollars unused.

I don’t blame IBM for attempting to put some rigor into their exercise of deploying consultants around the world in a more systematic manner; however, the real question is – what is accomplished by this activity, and is there a better way to meet the client needs (primarily local staffing, which may leave some holes in the skill set?)

Cross posted at LITGM

6 thoughts on ““Managing by the Numbers””

  1. Its like baseball. The best teams develop their own players from their farm systems. Other teams try to improve by hiring stars off competitors – that’s sort of like hiring consultants – and the results are about the same.

  2. They want to measure the unmeasurable.

    Managers have to justify themselves to strangers (upper management, accountants, investors, government regulators etc) who know little about the details and nuances of any any particular field or unit of a company. These strangers judge a managers worth by numbers, usually monetary ones. Therefor, there is an intense desire to create a system of metrics that can measure everything about a business function in an objective manner and communicate it anyone.

    Unfortunately, many vitals elements in business, such as whether person A gets along with person B, cannot be measured and thus reduced to numbers.

    One of the most important lessons a scientist learns in training is to understand exactly it is they measure and whether the measurements actually have predictive value. Unfortunately, to many people believe that number generated by a hunch or an inaccurate measure is better than no number at all.

  3. “Good management is rare.” I would say that they are extinct.

    What companies need are leaders that have a vision and a plan, not a bunch of yes boys and girls that only look on how they can save their own hide. Unfortantely the last 40 years our education system has produced mostly illuminati liberals that are afraid of tarnishing their image and looking like non-inclusive snob. We need to start cleaning up the system by giving control of the schools back to the states and maybe in 20 years we will start producing some decent workers and freethinkers.

  4. Part of the issue here is the heirarchical model is still in place at IBM. No surprise, they’re a dinosaur corporation. Heirarchical models were appropriate for Agricultural and Industrial economies, and roughly idealized themselves in the Feudal Enclave and the Modern Corporation.

    I put it to anyone reading this that the proper model for an IP & Services economy is not heirarchical, but some form of networked and decentralized decisionmaking. I think the central management idea in an IP&Se cannot be optimized via the heirarchical model, but it needs to be something more decentralized and less structured — when it comes to IP and services, the knowledge you need, the right decisions you need, are not going to be found efficiently by filtering them through a chain of people, up the chain, then down the chain as directives. One stupid person, one limited mind, blocks the flow and spread of the “right” idea. That is, the heirarchical model places all manner of roadblocks and obstructions in the way of the knowledge and decisions/suggestions getting to where they are needed.

    So some work needs to be done on optimizing that model. I don’t know exactly what form it will take, or how decisions will be made, but it does need something for dispute resolution between nodes, when one “node” (i.e., person) with relevant knowledge says “x”, and another node says “No, z!” There may already be some computer/hardware networking techniques which are applicable here.

    I would also suggest a term for it which I think carries the idea of who will be in these organizations — Kith.

    I think the typical kith needs to be fluid — everyone would be in different kiths attacking different problems at different times… and you might be called in to help with a solution to one narrow aspect of a problem because you know one or two people in a kith, who know you know about those types of problems, then go back off and do something with some other group you deal with more regularly.

    Hence, these kith are loose organizations of people you know whose skills you are familiar with (which is why I think the term “kith” is quite appropriate).

    Any individual may be in a few of them at once in substance, and dozens, if not hundreds, at once in some lesser form. There’s probably an “inner kith”, the ones mainly tasked with a problem and its solution, and an “outer kith” — associates of inner kith members who might be called upon for information on a specific aspect of the problem.

    And that name highlights the problem with IBM’s approach. It’s got the idea of matching up skills right, but it fails to value the relevance of working with people you know in solving problems — I have a friend who is smart as hell, but, since anytime he walks into a room, he’s the smartest one there, he tends to be rather arrogant about what he knows. So I know from my experience with him how well I can trust his opinions — because I know what he knows about and I can tell when he’s merely extrapolating from what he really knows.

    It also bears a resemblance to the local office model, as well as (via that “outer kith”) the means to approach the problem with the local office suggested above.


    I do think that the historical system of isolating and unsocializing the smart ones in an effort to “normalize” them (i.e., limit their capabilities down to the “average”) is going to bite society in the ass — we really do need to work on making the smart ones more socially able and capable of functioning in a group. They will always have difficulties because their approach to things is always going to be different from the norm, but that is the reason they are so useful, too.

    Anyway, that’s my US$.02… don’t spend it all in one place ;oP


  5. Good comments.

    I don’t know if I fully understand the kith definition but in practice it more or less fits my model when I staffed a consulting job… I either knew you and could command your loyalty (and knew you’d work hard towards success) or you were “dead to me”. It all is about getting people on your team and working towards a common goal.

    Another problem with the IBM type model is that you’d have to be a FOOL to give someone a bad rating… my plan was, just get them off MY job as fast as possible, and give them an OK rating. There was no upside for me to write this person up and try to drum them out of the firm, but there sure was a down side, since I would have to spend a lot of time on process and since no one else gives bad ratings, either, I would have to contend with the fact that a lot of crappy people walk around with stellar marks. Thus I am not just taking on this person, I am taking on everyone else he / she worked for previously, as well, and they aren’t happy about this either. Thus there is a ton of downside on getting rid of bad personnel (from the firm, not from my job) but no up side at all.

    I didn’t have incentive to share with others unless there was a chance I’d work with them again – the firm is too big and there were too many distractions – but I’d go way out of my way for someone I knew, especially since they could help ME later, or put in a good word when the ax was about to fall.

    The plan / assumption that people are centrally trained was really a joke, too… some were good, some bad, it all depended on what they brought to the table when they started. There is no common “firm” manner, since training in that vein was a local office specialty. Central training (when you went to another city) was basically a junket.

Comments are closed.