'Most companies are surviving on trial and error'

by Abhinna Shreshtha    Oct 18, 2010

John ZachmanWhat impact does a CEO really have on the business? Is it possible to simulate effects of decisions like hiring, layoffs, etc on the enterprise? To many, such questions are at best rhetorical, but what if these are answered? This is what we tried to do in an exclusive interaction with John Zachman, the man who developed the concept of enterprise architecture.

The field of enterprise architecture began when Zachman, then an IBM employee, questioned why architectural ideas and methodologies were not utilized in building enterprises. A radical concept for its time and one that has still not seen widespread adoption, Zachman’s unique views on business, IT and the enterprise could well lead us to rethink age-old beliefs and take a reality check on today’s business scenario.

Business-IT alignment has become one of the most used clichés now-a-days, are we any closer to finding a solution?

The world has not figured this (business-IT alignment) out; alignment is perceived to be a big problem that no body knows how to fix. This issue has been around for 20 years at least and the fact that people are still talking about it shows how little work has been done. If you see a priority list of any CEO in the last 4-5 years, alignment will always come in the top 10 issues.

Here’s what your typical CEO says to their IT team, “I have spent a lot of money in IT but I don’t think what you guys are doing with that money is in line with the plans I have for the enterprise.” When your CEO thinks this way, something is seriously wrong.

The problem is that enterprise strategy is not clearly manifested – Where do I see the company 15 years down the line? How do I get there? These are not questions that CEOs these days seem to be asking themselves.

How did the idea of using architectural ideas in an enterprise crop up?

Back in the 1960’s, at IBM, we were working on how we could move from ideation to implementation in a systematic manner. Of course, at that time we had no idea about enterprise architecture. Then in ’72-73, we began thinking about how architectural rules could be applied to this problem. The first sketch of what would become the Zachmann framework didn’t come out till the 80’s.

The Zachmann framework was learnt by looking at buildings and airplanes and the descriptive representations needed to create them. It turns out that you need the exact same representations to create enterprises.

For example, in construction you have a bill of materials, i.e. a list of the raw materials, sub-assemblies, components, etc. and their quantities needed to manufacture the end product. Well, even enterprises have a similar thing called an inventory structure, or the logical model, which is nothing but a bill of materials. An inventory of the enterprise includes its customers, products, vendors, orders, locations, etc. You can’t manage these without a bill of materials or, in enterprise terms, a logical data model.

There are more examples I can give you. There is a direct one-to-one relationship between every component of architecture and the enterprise. Now what happens if you want to knock down a wall in a building? If you don’t have the blueprint you won’t know whether knocking down the wall will bring the building down on your head. Similarly with enterprises you can make a few changes and the entire thing might just implode on you. If you want to change something that exists, and you don’t have any architecture, you either create the architecture or just blow down the whole thing and start from scratch.

Enterprise architecture is basically about converting ideas and strategies to reality. How difficult is the journey from conceptualization to reality?

The process of converting an idea into reality needs to go through a set of transformations – you got to name it, define it, create the semantic constructs of the concepts, then you have to represent it; because you are dealing with engineering designs; specify it, based on the implementation technology, then you configure the tooling and then you substantiate the idea.

By doing this you are ensuring the logic is consistent with the concept, the physics is consistent with the logic, the tooling is consistent with the physics, and then when you finally substantiate it, in all probability the substantiation will be consistent with the idea.

If you don’t follow this process it will be a miracle if it is aligned. The problem is that this is a very long process which puts off most business heads. Everyone wants quick results, but building an enterprise needs time and planning. If you don’t want to take the efforts then don’t expect it to be aligned.

So, if we put it in simple words, the Zachmann framework is like a blueprint of the enterprise?

Yes, it is a schematic representation of the various components in an enterprise and their relation with each other. These include inventory, network, process transformation, business concepts, etc. The framework also tells you how a change in any one component or artifact will affect other artifacts related to it.

How guilty is a company’s management for not introducing clarity at the conceptualization phase?

Here’s the thing – the business heads of the world don’t get very impressed when their enterprises are not aligned with their idea. On one hand they say that they want their business aligned with IT and on the other, they want to start nailing boards immediately because they want to see something today.

Basically, they are arguing the short-term and the long-term at the same time. If you want your business to be flexible, integrated, dynamic, interoperable, reusable, then that’s a long term process, it’s not going to happen overnight. If you want it to happen in days, your IT team will go ahead and do it, it’s not that they have a choice, but then don’t expect it to be all these things; just expect it to be implemented.

Now, I am not saying that there is something wrong about the short-term, it all depends on what you want. But what I can guarantee is that at some point of time you are going to wish your enterprise was engineered. But it costs too much and it takes too long, so people don’t want to do it.

To give an example - Rick Wagoner, the ex-CEO of General Motors, wanted to change the way GM was operating. The problem was that he did not know what needed to be changed. So it was a case of ‘let’s shut down this division and see what happens’, ‘let’s do this and see what the result is’. We all know what happened to GM by following this strategy.

Most of the enterprises are surviving on trial and error today; there is no order. If you increase the complexity of the enterprise, for whatever reason, or if the market suddenly changes, you need have some architecture in place to figure out how to adapt to the new scenario. Of course, if you don’t think this is an issue, then that’s a different thing. But the characteristic of the information age is complexity and change and enterprises are more complex than any man-made structure. Business heads need to understand that enterprises are not a commodity but an evolving organism.

Having served in IBM during John Akers’ tenure as CEO, what are your views on the tumultuous time Big Blue faced back in the 1980’s?

While in IBM, I was shifted to Los Angeles in 1972 to work on an important project. John Akers, who moved in as director of western region soon afterwards, was one of the smartest guys you could ever meet. I had the opportunity to meet him a couple of times back then. . A couple of years after I first met him he became the CEO of the company, a post he held till I retired in 1990.

My opinion is that he was the best CEO IBM had since Thomas Watson Jr. But Akers was the one in the hot seat when everything collapsed. They attributed the demise of IBM to him, when he was just the guy on watch when it went down. My personal opinion is that he had a pretty good understanding about where he wanted to take IBM. It took him 20 years to move IBM to a center of production environment and he knew that it would take another 20 years before IBM moved from a center of production to a mass customization environment.

Interesting really, one of the smartest guy you could ever run across and here he was being blamed for the demise of IBM, but his perception and his strategy was bang on.

You mentioned tradeoffs between long-term strategy and short-term profits. Is this an adverse effect of the business strategy prevalent these days?

It is popular these days to consume value to drive the stock prices up but the question is - are we also investing in the future? This is what CEOs and company boards should be asking themselves. If you are not consciously making long term investments then it will create problems in the future. These problems might not be apparent right now; 5-10 years down the line, a time will surely come when the enterprise will repent not making investments in the past.

The management needs to make up their mind whether they want to pump stocks in the short term or create value over the long term – do you want to consume value or create value.

Recently, Rakesh Khurana, a professor at Harvard Business School, was talking about the Jack Welch (ex-CEO of General Electric) rules of success, basically shareholder value became managed earnings became earnings quarter to quarter to please Wall Street. But Welch was consuming the future for short-term success. If Jack Welch is the epitome of the modern CEO, he is also the epitome of the short-term strategy.

Whatever happened to concept of making an investment that was not realized till one or two CEO’s down the line? Long-term investment is what made capitalism the envy of the world, but the trend is more towards short-term profits now-a-days. Khurana feels this skewing towards the short-term strategy is a watershed moment in the history of American capitalism and I agree with him.

Of course I understand the significance of short-term success; you got to keep the stock price up too, but if that is your only aim then you are not making any business in the future. Here is what enterprise strategy should ideally be – a number of short-term decisions that are a subset of an overall long term strategy. In the long-run, enterprises that cannot accommodate the idea of architecture are not going to be in the game.

Could another reason be that today’s enterprises have become too hierarchical and nobody wants to take ownership for what is perceived to be a very long-term issue?

There is no single reason why enterprises fail. One is the CEO issue – the cycle of enterprise architecture is longer than the tenure of CEO. Of course, an enterprise can continue with its enterprise architecture independent of the CEO, but if you want to engineer the enterprise to what the CEO wants then he needs to be involved.

Similarly, there is no reason why a company cannot continue with its enterprise architecture project if, say, the CIO leaves suddenly. You just need to ensure that the project is well documented and you have all available information regarding each and every artifact and process in the organization. Enterprise architecture does not need to be dependent on any one person.

You have compared enterprise architecture to a science, but any discipline needs time to develop and mature. How long before we see enterprise architecture mature to a science?

It is 2010 and enterprise architecture is not even a discipline yet. Yes, I have defined the fundamental constructs behind it, but there is still a lot of work to be done. The Zachmann framework is an ontology, the underlying laws of physics for a discipline called enterprise architecture or maybe enterprise management; whatever it will be called in the future. You need to understand that the whole concept is less than fifty years old. We don’t have a thousand or even a hundred years of knowledge behind it. Think of it like string theory. My work till now just provides a base for future generations to work on and develop a discipline out of it.

We are still researching on the periodic table, research goes on forever. I call what we have done so far Enterprise Physics 1.01 and I think this is a beginning.

With a number of different ideas about enterprise architecture around, is there a case of too many cooks spoiling the broth in the offing?

Right now, the argument in the industry is what enterprise architecture is and I will be very dogmatic and suggest to you that the Zachman Framework is enterprise architecture. Because it is the same architecture used for buildings, airplanes, locomotives, for every single thing that humans have created. It has everything that you would require and it defines everything.

However, many people are arguing about their own versions of the definition. You may be doing something very good, and there are many who use enterprise architecture ideas in their organization without realizing it, but as long as you are not building this model you are not doing enterprise architecture.

If we would stop arguing about what enterprise architecture is and start figuring out how we can use it to build enterprises that are lean, dynamic and can accommodate the complexities in the future, we would be heading somewhere.

What does the future hold for enterprise architecture?

The aim is that soon you should be able to do with enterprises what you can do with airplanes or buildings i.e. you take a wire and trace its path in 3D and using graphics in the architectural representation of the airplane down the wing, up the fuselage and into to the tail. Similarily, you should be able to take an employee in an organization and trace him through the various levels of the company. We can simulate performances of airplanes in 3D, why can’t we do it with enterprises?

This is where we should be heading – a stage where we can actually simulate how certain changes or decisions will affect the enterprise. For example, I move my cargo from Mumbai to Singapore, or I fire certain employees. I should be able to check a simulation and figure out how it will affect my company. Basically, the aim should be to remove guesswork from decision making.