SOA is Dead again?

Winchester Mystery House  Picture source: Wikipedia

On January 05, 2009 Burton Group's Analyst Ann Thomas Manes, published a controversial blog post titled: SOA is Dead; Long Live Services.

Few days ago I answered an ebizQ Forum's question: 

How important is SOA to a company's mobile strategy? Only two experts (including me) answered the question.

SOA Questions asked in the same ebizQ Forum two or three years ago where frequently answered by 10 or 15 experts.

Is the lack of interest in SOA (despite the Mobile context included in the question) an indication of SOA's final death? I do not think so.

I have to agree with Mark Twain's saying: "The reports of my death have been greatly exaggerated".

As far as SOA is concerned, Mark Twain's saying could be paraphrased as: The reports of SOA's death have been greatly exaggerated.

Whinchester Mystery House is a well-known mansion in Northern California. According to Wikipedia: "It was continuously under construction for 38 years and is reported to be haunted".

If you would like to know, why it was continuously under construction for 38 years, you could read the article in Wikipedia.

About ten years ago, I worked as a freelance Consultant, for Giga Information Group (which was acquired afterwards by Forrester Research). 

Giga Information's analyst Phil Murphy, used

Whinchester Mystery House continuous construction, as an analogy to Information Technology strategy: 

Every few years we (Information Technology community) invent new technology, new architecture or new concepts and renounce older concepts, technologies, architectures etc. 

However, the term renounce, in this context, could be defined as something we frequently use but seldom talking about. 

For example, The Mainframe which was "dead" more than twenty years ago is still alive and kicking.

More than ten years ago, I asked for Service Based implementation references sites.

The reference sites requested were Mainframe sites, because a client considered Core Systems development initiative, based on Mainframe infrastructure and Service based approach.   

As usual with Leading Age Technology, many reference were too small or in early development stages or irrelevant.

Only one reference site was impressive. It was a Finish site. 
The site success could be partially explained by, their previous successful implementation of Component Based systems.

My Take
Phil Murphy's Whinchester Mystery House analogy is today as relevant as it was 10 years ago. 

We talked ten years about SOA and now we are talking about Mobile Computing, Cloud Computing etc.
In  ten years IT workers will stop talking about Mobile Computing and talk about other concept or technology.

Although, nobody is talking about SOA, it is a frequently used Mainstream concept,architecture and technologies.  

The trouble is that successful and relevant reference sites numbers are still too small.

Those who successfully implemented yesterday's concepts and technologies, will usually succeed in implementing the new concepts and technologies.

Those who fail to realize yesterday's technologies Value Proposition, will usually miss tomorrow's technologies Value as well.  

As far as SOA is concerned, many of the failing enterprises need urgently Consulting services in order to succeed.

  

Customers typology: The Captive

picture source: Wikipedia

The word Captive in this context, is not about someone caught by enemy's forces.  It is about a customer, whose conception about the Consultant's topic are dictated by an expert or someone pretending to be an expert. 

In a previous post you can find an example. The post titled: Vendor Survival Guide: Supermarket, Grocery and Kiosk. 

An example depicted in that post was Data Sorting Software product selection process. The Captive Customer selected a product of Servers and Operating Systems vendor (The vendor is an analogy to a Supermarket). I recommended, a superior product of a small company. Sorting Software was the only product the company developed and marketed (Kiosk).

The "Supermarket" does not care, if you buy chewing gum, as long as you buy meat, fish, bread and milk (Servers, Storage, Operating Systems etc.).
The kiosk's existence depends on chewing gum (Sorting Software). The "kiosk" will try harder, by selling better and cheaper Software.

The supermarket has to sell chewing gum, in order to be a one stop shop. However, its profits on chewing gum are negligible. Therefore it does not care much about selling chewing gum.

Supermarket, kiosk and the Captive Customer
A rational buyer, who is buying only chewing gum, should prefer buying it at the Kiosk and not at the Supermarket.

Similarly, a rational buyer should prefer buying cheaper and superior Sorting software product, from a vendor focused on this product.  To my surprise,  the customer rejected my recommendation and chose an inferior and more expensive product, developed by a vendor selling a variety of software and hardware products. 
It should be noted that the same customer's decisions were often rational decisions.

Why does the customer decided irrationally?
The Customer was a Captive Customer. The Vendor's  Representatives told him that their product is a lot better than the dedicated vendor's product. It was a good enough reason for choosing the inferior product.

Remember, the vendor's employees do not care much about selling Sorting Utility program. Therefore, they did not try to convince him to chose their product.

Main Characteristics of the Captive Customer
The Captive Customer's opinion is that someone else (not the customer) is the oracle. He will do whatever the oracle will say.
The oracle could be a Software and/or Hardware vendor such as: Oracle Corporation, Microsoft, IBM, Apple, HP, SAP, Google etc.  

He could be a Software House, an Information  Technology Research company or even a mythological former manager, who worked for the Customer's organization twenty years ago.
 
How can the Consultant influence a Captive Customer?
changing a Captive Customer's viewpoint is a difficult and complicated process. Success is not guaranteed. You should utilize different method's and different approaches to change viewpoints of different Captive customers.

The following bullets describe potential approaches and ideas based on my vast experience:

• Avoid of confrontation with the Captor

Explain to the Captive Customer what and why you prefer the solution you recommend. Do not criticize the Captor and do not describe his motives and benefits, in case the customer accepts his recommendations (Usually Captors are selling Software and/or Hardware and/or additional Services).

• Refrain from explaining previous mistakes based on the Captor's recommendations

The Captive Customer could have Patricia Hearst's Syndrome. She identified with her kidnappers' ideas. He may be identified with his Captor's views. If he developed Patty Hearst's Syndrome, confrontation with the Captor could operate like a boomerang.

• Present Research, Viewpoints, Case Studies of others supporting your viewpoint 

The data should be with clear cut views and the source should be a person or a company or academic authority, which the customer respects.

For example, many Captive IT Customers respect, research firms such as Gartner, Forrester Research and IDC. In that case, present answers to specific queries issued by you or the customer, relating to the current issue.

You can also present Research Notes, verbal Analyst's opinions etc.

Others could respect a former co-worker or former manager agreeing with your views or even evangelizing these viewpoints. Let them know what she or he thinks (directly by talking to him or her). 

According to my experience, sometime Not Invented Here(NIH) supporting evidence in other language may help you more than local evidence in your own language. 

   

Customers Typology: The Self -deprecating Customer

image source: Wikipedia

The Self -deprecating Customer is exactly the opposite type of The Customer who Knows Everything. He automatically accepts the Consultant advice, without any questions or hesitations.

Probably, you will not believe that the first example of a Self -deprecating Customer I think of, is the CIA. It is about a project named Triangle. Triangle was an Information Technology project of Legacy Modernization and systems expansion after September 11 attacks. 

After  completing the Development phase, the CIA begun the Implementation phase. However, someone who was a Senior Manager, thought that it will be a good idea to rethink if the project fulfills its goals. A committee of 100 leading academic and IT industry experts, investigated the project.

The main recommendation in  the committee's detailed report was: stop implementation and fix the errors in the systems. Afterwards the systems could be implemented again.

It should be noted, that stopping implementation at such advanced stage of the Development Cycle, is very expansive. However, after reading the report in the Web, I understand why the experts recommended to stop and rethink.

The CIA made almost every mistake I could think of. As far as this post is concerned, the self -deprecating mistake was decisions taken by consultancy vendors instead of CIA Managers.

Is not he the ideal customer a Consultant could think of?
It looks like ideal circumstances for consultancy: The consultant's long term employment is ensured, his vision and recommendations are accepted, as well as the deployment methods he recommended.
However, the scenario could be not as good as the consultant could imagine. The following are main drawbacks:

1. The customer takes no Responsibility
In case of failure, regardless if it is due to the consultant's mistakes or unrelated factors, the consultant will be responsible for it.
The customer would argue that he done exactly what the consultant recommended.   

 2. Following blindly is not limited to following the consultant
 The customer may easily follow someone else's advices without any doubt. It could be a vendor or another consultant. The result could be lack of consistency and no systematic approach. 
For example, another consultant contradicting recommendations in related topics, could be executed along with the consultant's recommendations.

3. The Customer is not learning
In the long run this could be the most severe problem. For example, one of the goals of SOA consultancy is to create a Reuse Culture.
Most of SOA initiatives without Reuse Culture fail.  The Self-deprecating Customer could easily preserve the old non-Reuse Culture. 

Usually I prefer to give to my customers fishing rods and not fish.
The Self-deprecating Customer usually prefers fish.   

Customers Typology: The Cusomer who Knows Everything

I worked as an Information Technology Consultant, for many Customers. Customers are different in their Self-Perception, in their view of the Consultant and the Consultancy Process. They also differ in the Enterprise size (Large, Very Large, SMB) and their Vertical Industry (e.g. Telco, Banking, Insurance, Utilities, Public Sector). 

Customers Typology could effect the interaction between them and the Consultant, as well as the probability of successful consultancy.
This post is about a particular customer type: The Customer who Knows Everything.

The Customer Who Knows Everything
This Customer type is sure that he knows more than anyone else. He is also sure that his understanding is better than any body's understanding. 

His point of view is manifested in the Decision-making process: 

1. There is no correlation between the Consultant's recommendations and the decision.

2. There is no correlation between the consultancy quality and the decision taken.

3. The decisions may not be related to information collected and analyzed during the Consultancy process. 

The consultant may be a knowledgeable and experienced expert in a topic. For the Customer it could be the first time he and his enterprise are exposed to this topic.  

The consultant's work could be excellent: assembling data from variety of sources e.g. Analysts' Research Notes, Case Studies, similar enterprises 's strategies, approaches' solutions and conclusions etc.

The conclusions based on this excellent work could be that strategy A or Software/Hardware product A's implementations usually fail, while strategy B or product B's implementations usually succeed.

The Consultant's recommendation could be: select Strategy B implemented by Product B.   

The Customer who Knows Everything, will chose Strategy A implemented by Product A. He could argue: "I know that everybody in the world selecting Strategy A implemented by Product A failed. I know better and understand more than them, so I will be the first to succeed". Needless to say, that he fails like all the others (sometimes even more sever failures than others).

According to Aberdeen Group's SOA survey, Best in Class (BIC) Enterprises (those whose implementations succeed and provide real Value) used more Consultancy Services than others. Laggards (those who fail to provide Business Value), refrain from using Consultancy Services or purchased less Consultancy Services than others. 

I would guess that, the Laggards group includes, at least, few Customers Who Know everything.

How to Consult a Customer who Knows Everything?
The question in this paragraph's heading contains the answer: If someone knows everything it is impossible or at least very difficult to consult.

Effective consulting requires conceptual change. Information, data and brilliant analysis would not change the customer's approach. 
A Psychologist may change the concept, but we are only IT Consultants not Psychologists. 

He will learn only after painful failures. To help him, the consultant should repeat and emphasize his conclusions contradicting the customer's decision, so the customer could blame only himself in case of failure.

However, sometimes even repetitions of the recommendations contradicting the customer's approach, could be in vain: The customer will know who to blame.
The Consultant opposing the wrong conclusions and actions taken by the customer, is to be blamed.