Skip to Content

I would like to take a moment off the day-to-day blogging and observe an interesting issue that was brought up in a recent industry conversation – that of the unifying force in the IT industry. As we are seeing a transition from Client server (or internet enabled client/server) architecture to a next generation Service Oriented Architecture, we are witnessing a repeat of all the same phenomena’s that we have seen in the last mega shift (from Mainframe to client/server). It starts with a period of uncertainty focused around pure-technology discussions, followed up with a de-facto standardization of an agreed blueprint. What followed last time was an exchange of the technology discussion with a business/application discussion as the main focal point for a significant eco-system/after market evolving around very few players.

What is interesting though is that the previous change was so sharp it was almost brutal to the incumbents. Players like DEC, who did not move well into the new model (separate story, I heard from Hasso on this one – I need his permission to repeat) evaporated completely…well they were subsumed in a company that was subsumed in HP (who not only adopted the model, but became part of the original blueprint for SAP). The most interesting part was that the industry as a whole found a common enemy which moved every participants in the market with a fervor only seen in sharks when they smell a wounded source of protein in proximity. The common enemy of the last wave was undoubtedly the mainframe – everyone wanted to get rid of the mainframe and many companies were hoping to build their business plans on reducing and replacing the maintenance cost for the big beasts, and the people that operated them.

So the question was left open as to who is today’s unifying enemy – that one force that all of us want to make irrelevant to the degree we can agree on everything else in the blueprint. As much as we tried to think of a physical element to replace in the IT shop we could not find it. I believe that some of SAP’s competitors actually had in mind a few years back that we would be that big element to replace, thinking of breaking the enterprise backbone into a small collection of components and letting hundreds of fish swim in coordination to make a single shark…unfortunately that plan didn’t work, as you cannot really replace the application backbone that easily, and SAP has actually figured out the SOA is a great thing for us, and moved from a follower mod e to a leadership position i n the last 3 years. And so we tried to look at the problem again only this time from our customer’s eyes, before we analyzed the vendor community. You see, customers were the first to throw the mainframe in to the water for us all to develop the taste for replacing it, and they were really afraid of the disconnect the mainframe created between them (the CIO or head of IT) and the CEO, who could not figure out why they were getting less and less impact for more and more money.

To really understand the cause for how we get there, one needs to understand Geoff Moore ‘s great model for process innovation (for more on Core and Context see Geoff’s article in last year HBR ). What happened in the past with mainframe was that companies got their IT budget stuck in mission critical context hell (what Geoff calls the Demon) just maintaining the Mainframe and the apps on it. The way from there is either back to investment in core or being outsourced. It is interesting that we are facing the same enemy today, but we cannot manifest it in a physical mainframe box. The enemy is still a fight for relevancy, and the fear is still pretty much the same only with a different name, outsourcing. There is nothing (or no one in this case) that embodies the fear better than Nick Carr . Don’t get me wrong, I have nothing personal against Nick (although some people who read his blog entry about me feel he does harbor some personal grudge against me, or is it just the people who pay him that do?), I think that he believes in what he said– that the IT industry is exhausting its differentiation. Unfortunately, I do not believe businesses have exhausted their means of differentiation, and what we do as an industry is merely enablement of the varying business models that our customers throw at us. Now, if at all, I believe the rate of business model innovation has and is significantly accelerating.

So we have a fight on our collective hands, customers and vendors, to get our system in shape, in agile, recombinant form, before the big bad wolf comes over and deems us all irrelevant (in which case he can outsource the IT department and everything that’s in it to the lowest bidder). The fight is relevant to all of us, because the outsource step will not end in one move, rather it will become a “WalMarting” of our industry, regardless of whether we have a business plan that says we will become arms supplier or we will leverage the lower cost. In a world where you have outsourced the IT of your company, you don’t really have any business flexibility after the contract is signed (think changes to your house plan after the master contractor has signed you in blood on the blueprint…think again). What you do have is a master contractor that is enjoying the benefits of Moore’s law (the original Moore , not Geoffrey), while you are hoping none of your competitors changes the business model and you need to catch up. Catching up is really hard to do. In other words, the next big game is who becomes the industry disruptor, and who gets dragged behind. The disruptions are going to come in shorter intervals and with more brutal force. The disruptions will start interlocking across industries, and will feel more like Tsunamis than earthquakes.

That is why I believe we are seeing an industry that is embracing the next big thing, and a wave of process innovation building around it. But that is all in my next blog on the Enterprise Services Ecosystem , and the great Enterprise Services Forum we ran this month in San Jose.

To report this post you need to login first.


You must be Logged on to comment or reply to a post.

  1. Former Member
    “In a world where you have outsourced the IT of your company, you don’t really have any business flexibility after the contract is signed”. 

    I don’t agree with that at all.  For businesses to adapt and change in the increasingly turbulent environment they find themselves in, they need to reduce inbuilt inertia.  The less “weight” that a company carries the more capable it will be to change direction and maximise its competitive position.  Consider manufacturing; it wasn’t long ago that it would have been unthinkable for a company to outsource manufacturing, now its considered “best practise”.  Nike is one example of this, Boeing another (with its recent deal with Japanese manufacturers for construction of aircraft wings, the most difficult and costly component of the final product).  There are many examples closer to home (in an IT sense anyway) also, with HP, IBM, et al long ago outsourcing PC manufacture.  Far from reducing (or eliminating as you would have it) business flexibility for these companies, it has enhanced it, allowing them to focus on branding and market positioning rather than getting “weighed down” with the day to day issues of product manufacture.

    We are seeing the same trend with IT.  Already PC and network support has been outsourced, in a lot of companies’ system administration also, with some going the whole way and outsourcing all functions.  Outsourcing contracts are not reducing flexibility for these companies; the only thing that is reducing flexibility is the underlying systems that they are running on, who’s proprietary nature makes it costly to embrace change.  This, in fact, should be one of the benefits of the standards based, services architecture that SAP (and the industry in general) are moving towards.  It should facilitate the interchangeability of components from different providers, allowing companies to pick and choose the solution that works for them now, knowing that they can change at any time (and not be too bogged down with integration issues).  In this model the outsource provider will invoice the customer on a monthly “per service” basis with companies using a “tick the box” questionnaire for the functionality they will be using for the following month.  Flexible?  I think so.

    Rather than comparing an outsource contract with building a house, I would compare it to a holiday house.  Do you buy one and go back to the same place every year, or do you rent, and take your pick?

    1. Brad,

      Two observations:
      1) what you describe in most of the examples is out-tasking not outsourcing. If you send a single task out, and keep the ownership on the interfaces in your hand, with clear contracts around quality of service, and replacement possibility for the supplier, you retain flexibility. OutSourcing is when you send the complete responsibility for your IT department or support for a complete end-to-end process, with no flexibility to change into someone’s hands. Not recommended – as many examples of recently broken outsourcing deals can show you that the only way back out is a binary breakage of the ocntract.
      2) some of these companies that slowly outsourced part of their product to others remained in the end empty. The PC case is a great case at hand. Similar thing happened to Compaq with the outsourcing companies. See a book called Innovator’s Dilemma for a full description on how companies “Sashimi’ed themselves to death” that way a slice at a time.

      You have to retain control of the most essential part of your business – the processes that define entity relations and behaviors. IF you don’t do that you will lose competitivenes, through the loss of flexibility that leads your competitors to a faster time to change than you can rewrite the contract.

      1. Hi,
        Usually I would read the blogs and take home the useful information. This one, however made me write a reply.

        In case of SAP, like Shai has put, so far there has been more of out-tasking than outsourcing.
        I came across a special collection of research papers from Mc Kinnsey. (
        They refer to umpteen number of benefits when companies outsource. Ofcourse there would always be some glitches, but the benefits outnumber.
        The sea of change is happening on how businesses run, and the only mantra is adapt or die.

      2. Former Member

        Outsourcing is great if done right, but can be disastrous if it goes wrong.

        GE/Jack Welch was one of the first one to outsource services (he also came up with the 70-70-70 rule) … In the last year or so many US companie outsourced because it was a trend and they were trying to save money by outsourcing their non-core, commodity work. In some cases they didn’t do it right so in the long term they suffered so they had to bring it back in but there are many cases that outsourcing work very well in the long term and saved the business alot of money and in the same time helped companies focus better on improving their core business.

        Also, I dissagree with Shai’s statement that when you outsource you have no control and no way out of the deal; First, the contract is of limited time (1, 2,3 or 5 year) – plan for contingency. Second, you include SLA’s that cover not only responsiveness(quantity of service) of the vendor but also the quality of the service you’re getting from the vendor. Third, add clauses so the contract can be revised every year based on findings from the previous year. Fourth, good legal department advise can add clauses that keep the door always open to get out.

        I’ve been involved in outsourcing IT services for at least 4-5 years and I’ve found out that if you know what you’re doing you get great benefits from outsurcing/offshoring commodity work.

        Outsourcing<>Offshoring; offshoring is the cheapest way of outsourcing. Many companies have outsourced IT services to HP, IBM, GE etc, but most(if not all) of development that HP, GE, etc is doing for you is based in India … so to some extent the ones who have outsourced, have offshored too.

        One last thing, OutTasking, OutSourcing mean all the same thing. What you need to do is: keep the brain(business experts, architects, knowledge area experts) and outsource commodity work (developers, etc). The same developer skills that in US costs $120 it will cost $25 in India.

        good luck,


    2. Kenneth Moore
      Brad – You ARE an outsourcing source yourself (consultant) in a sense.  So I don’t think your input is very insightful because you are biased towards outsourcing (consulting – a form of outsourcing).
      1. Former Member
        Hi Kenneth, to quote Shai “That is far from a fair statement”.  When engaged in debate, normally you need to rebutt points of those with opposing positions, not just come back with, “your points are worthless because you are biased”. Tell me why I’m wrong (and Nike, Boeing, etc…).

        Anyway, my points were focussed on operational outsourcing rather than project (consulting) based outsourcing.  This type of operational outsourcing threatens my livelyhood (or at least threatens to change my customer) just as much as company employees.  I could just hide in a corner and pretend its not happening, or I could read the writing on the wall, engage in the debate, and prepare to adapt to the emerging business landscape.

        What are you doing about it?

        1. Former Member

          Shai, as already rebutted your arguments on outsourcing and cited some examples also.

          As for out-tasking, I believe that needs to be done on a careful case by case basis. You basically want to out-task what is not providing business value, but the issue will always be that you have not control over how the outsourcer delivers their final product.

          Please see this article that clearly points out the issue has always been more of clearly specifying and articulating the work that needs to be done.

          If you DO NOT spell out everything in the outsourcing contract and pro-actively manage it you are doomed to receive a product that is not what you wanted.

          1. Former Member

            Shai didn’t really rebutt any points he just clarified some terminology (some would say splitting hairs, or was that splitting sashimi?). What I call outsourcing (and you too in your last sentence), he calls out-tasking. 

            If we agree I was talking about out-tasking, then I think we are all on the same page.  I completely agree that you need to keep a high level of control of your external providers, but this doesn’t make it a bad thing.  In fact I think its the future of business.

            1. Former Member

              You kind of missed the point of the article. It’s always been very,very hard to clearly define the packaged unit of work that will be out-tasked or out-sourced…That is the hardest part, the management of the outsourcing. We can clearly see that in the past with failed out-sourcing projects. The finished product delivered by the out-sourcer is not quite what was supposed to be delivered and thus alot of re-work had to be done in house to clean up the mess…

              Now, suppose this was a task that added value to a company’s core competence/focus? Could a company risk outsourcing such vital business processes? Can a company afford to save money on somewhat less up-front costs to spend more later on re-work and have irate customers?

              I think there are 2 trends:

              1) The trend will be to more of U.S. sourcing instead of full-fledged outsourcing…The natural competitive nature of the U.S. will allow it to adjust to the cheaper offshoring labor pools.

              2) The cycle of business will eventually find some equilibrium with outsourcing. The luster of outsourcing will wane as more issues bubble to the surface…For example, Dell and others have brought call centers back, etc…

              Now I’m not saying out-sourcing/off-shoring is bad…it definitely has a future. What needs to be talked about is the middle ground. Nothing is all good or all bad. You just need to understand all the issues concerning your decisions…

              1. Former Member
                Three points:

                1. It may be hard to manage outsourcing contracts, its also relatively new. As it becomes more prevalent and both customers and outsourcers develop better skills in this area, it will be no different from any other supplier / customer relationship (and potentially better as it will be by necessity more intimate).

                2. Outsourcing <> Offshoring
                Outsourcing refers to moving company functions outside the business. Offshoring means moving company functions outside the country, and potentially outside the business. I believe outsourcing is the future for the majority of businesses in the future. Offshoring on the other hand will develop an equilibrium point that will alter with time as the different countries’ competitiveness varies.

                3. US <> World
                Perhaps when you say the trend will be towards US sourcing, you mean the trend will be towards “in country” sourcing. I don’t really see Europe and especially Asia offshoring major business functions towards the US.

                As a last point, as this is an open forum I would encourage you to steer away from comments such as “You kind of missed the point of the article”. Inflamatory remarks such as this lead to mud slinging and don’t move the debate forward.



                1. Former Member
                  I’ll respond to your last point first. I’m sorry you consider that an inflammatory remark, it was more of a observation of my perception that the central point of the article was missed. I’m sure you’ve seen much worse remarks before. I was not trying to insult you and would have used much worse language if that was the intent…

                  1) Again back to the article. This article came out in the 80s and it is still “relevant”. The hardest part will always be clearly specifying what you want. This will always be the issue with outsourcing and with general management of internal projects as well. The difference is that if it is outsourced you have no control over how the work is done and you will receive a finished product. How closely the finished product matches the specs will determine the value of outsourcing deal.

                  Also as Shai has pointed out, outsourcing CAN move too slowly for many businesses. You spend all your time creating the perfect specs in order to receive the perfect delivered solution but that simply takes too long…

                  The difference between software/technical project and supplier project mgmt is huge. Why do so many software projects fail?

                  2) Agreed

                  3) No, more along the lines of U.S. companies in sourcing to lower wage areas of the U.S. Offshoring will drive down U.S. labor rates in general and certain U.S. rural areas already have low enough wage rates to be competitive…

        2. Kenneth Moore
          Brad – I didn’t say your points are worthless.

          Personally, what I am doing against outsourcing is what I would do normally.  No job is guaranteed permanently: – I try to keep my skills up, provide great customer service, keep my ear to the ground, and put away enough $$ in savings to stay afloat long enough to find a new job.

          Have you ever called a service number only to reach a person who you can barely understand, who barely understands you and between the two of you, can barely communicate?  Well, I have and it makes me angry.  Poor customer service by wasting my time.  And still my problem is not solved because of miscommunication.  That is the end result of outsourcing.  Poor customer service and waste.

          I think the market/customers will force the “fad” of outsourcing to back-fire for the most part.  Of course, there are a few exceptions here and there. 

          Localization is a better business decision, not outsourcing.  Better market reaction time, better understanding, better customer service.

          1. Former Member
                Outsourcing is not a “fad”. If it was fad then the growth in the companies that are benefiting from outsourcing will be gradualy going down. “Fad” is far from truth because these companies are posting record revenue every year. In America we always think of outsourcing in terms of call center, which is far from truth as well. Companies from U.S are outsourcing numerous technical jobs. Localization was be always be there especially in public sector but i think outsourcing is here to stay. I am independent contractor and i support multiple clients via VPN.  
            As long as your customer is happy with performance, localization will be not a issue.

            Prakash Singh

  2. Benny Schaich-Lebek
    far from measuring me with your presentation style: I like analogies and use them everytime I can, If I can find a good one. The body-system analogy is a good one, but faaar too long.
    Anyway, Carr’s joking on you only lead some reader to recommend your guidance…


  3. Former Member

    I see what you are saying and share some of your outlooks – especially on the timing and impact of future changes the industry will face, but must however question some of your points.

    Take the next TechEd for example: SAP offers free participation for developers during the Bangalore event, but not for Vienna. This I find interesting, as you mentioned the “fight that we ALL have on our hands”. Wouldn’t you say that the equipment for this fight is being unfairly distributed? Why is it more difficult for european developers to gain Netweaver certification through the TechEd?

    1. Michael,

      Fair point. We are trying to establish TechEd in India, as an entry to the market we have to behave like the rest of hte market behaves in India. In the future we will probably need to re-evaluate this policy of no-charge.

      1. It’s a clear trend… I’m probably not alone in my fears that a considerable percentage of “old-world” IT personnel will be outpriced (and outsourced), at least in the short run. We’re already seeing this all over the continent (and in the US).

        SAP behaves like all other behemots in this respect; by going where the prices are low. There was an article a couple of years ago about Microsoft & Sun handing out training material and software for free in India, apparently to “gain momentum”.

        1. That is far from a fair statement. We have grown our presence, education and partners around the world. We do not try to shift from one place to another, we just provide equal opportunity to all.
          If you want to see who is taking advantage and who is contributing to the ocmmunity, check the top contributors in knowledge (for free) on SDN, you will see who is providing free knowledge to all of us…

      2. Former Member

        thanks for your response.

        I appreciate your comments with regards to the global market and SAP’s aim of getting a foothold in India. However, I don’t think it justifies an exclusion of the certification offerings in Europe entirely. The registration fee in India has always been lower than in Europe or across the pond anyway, which makes a big difference already. What I would like to see is cerification opportunities at all TechEds – at prices that reflect the market situation in the individual part of the world.

        I have noticed that the Boston even has now been opened up for certifications as well – which I find commendable. It would be great if the same would happen for Vienna and Tokyo.

        Kind regards,

  4. Former Member
    Dear Mr. Agassi,

    far from disagreeing with you, I really enjoyed the conclusion of Nick Carr:
    “I’m not sure, but I think at the end he’s talking about getting an erection.”


  5. Former Member
    There is a huge difference between outsourcing the IT processes close to “IT sourcing” (people, maschines) and the IT-application. Developing or changing business models is heavily related to the IT-Application stack and less to the IT-Infrastructure stack which provides the foundation of capabilities. We will face a ‘Tsunami’-disruption if new business models (or economic conditions to run IT) are changing and IT-applications are not flexible enough to adapt quickly and efficiently. Outsourcing or Hosting the IT-Infrastructure wouldn’t harm the companies as long as they keep control of the IT-Application stack to model their processes without moving into contractual renegotiations.
  6. Shai,
    I like the strategic cast of your thoughts but I think sometimes you need to go deeper. Outtakes from your words in quotes with my comments following:

    “As we are seeing a transition from client/server architecture to a Service Oriented Architecture, we are witnessing a repeat of all the same phenomena that we have seen in the shift from mainframe to client/server.”

    An SOA revolution with parallels to the past seems to be gaining momentum. Again the brake is fud over new technology and the unhappy prospect of having to adapt all one’s business processes just to keep up. The main parallel is that IT innovators have rallied enough after the last wave to push for a new wave of reinvestment.

    “The common enemy of the last wave was undoubtedly the mainframe. Companies got their IT budget stuck just maintaining the mainframe and the apps on it.”

    Abstracting a level, the common enemy was the inflexibility and unscalability of mainframes. People with steady workloads and unchanging processes were happy to stay with their big iron. Similar now, mutatis mutandis. The increased flexibility of SOA is an opportunity for IT to drive another synergistic feeding frenzy between IT pushers and users.

    “The enemy is still a fight for relevancy, and the fear is outsourcing. I do not believe businesses have exhausted their means of differentiation. I believe the rate of business model innovation is accelerating.”

    Relevancy? The fight is for the centrality of IT as the wavemaker. Outsourced IT is out of sight, out of mind. You are surely right that differentiation goes on and business model innovation is accelerating. The challenge is to persuade innovators to leverage IT as a force multiplier with long-term promise, instead of riding global labor cost inequalities that offer only brief fixes.

    “We have a fight on our collective hands to get our system in shape before the big bad wolf can outsource the IT department. In a world where you have outsourced the IT of your company, you don’t really have any business flexibility after the contract is signed.”

    Right, flexibility goes out the door. But what’s the payoff with flexibility? If the situation changes, I can adapt. The cost? I can adapt wrongly, and break my moneymaker. We need to sell analytic power, nay visionary power, the prophetic power to see ahead and ride the next wave. It’s not enough to give customers a ton of bricks and a drawing board and tell them, hey, build your own apps.

    “The next big game is who becomes the industry disruptor, and who gets dragged behind. I believe we are seeing an industry that is embracing the next big thing, and a wave of process innovation building around it.”

    The next big thing is not SOA but the new ability that SOA provides to adapt to changing circumstances. The vision, as you recently suggested, is that firms are increasingly like organisms, with SAP nervous systems. Market churn is a Darwinian force that drives out unfit organisms. This suggests we need to breed not just flexibility, but responsiveness, power, speed.

  7. I experienced a company outsourcing IT first hand.  And I can tell you that on paper they were very excited about it.  The company re-badged all the IT developers to the outsourcing company.  Most the “good” developers left within 2 months since the sourcing company was just that, a sourcing company, and not a good place to work.  So the company was left with the bottom of the barrel performers, and whatever offshore resources the sourcing company could find to replace the “top performers” who moved on.  Well it has been a year, and the business divisions are up in arms.  No work is getting done (since all the top performers left without knowledge transfering), and now core employees of the company are starting to leave because they can’t get their projects done, which affects their bonuses and their bottom lines. 

    The IT management (which was not outsourced) was quoted in saying they may have made a mistake by outsourcing all the developers. 

    Where does the company stand now?  Fighting to earn back its reputation, and getting downgraded by the market.

  8. Former Member

    I believe there is a common enemy and it is naturally inherent in all industries, it’s just profit.

    The Model: If there is somebody who is able to do things cheaper than I can, with acceptable quality he’ll get the job and I get the profit. With a bit of luck he is not the only vendor so I can retain flexibility. What this purely means to me is, “doing things remotely”. Whether we call it oursourcing or out-tasking depends in my opinion on the domain of the system-stack that we’re talking about. But independend of what kind of things we do remotely, we will have to well-define interfaces and communication protocols. And of course we give away the responsibility for that remote job. So maybe it’s a good idea to establish a control loop to ensure quality.

    The mainframe mapping: Mainframes vanished when a bunch of standard boxes were able to do the same or even a better job but cheaper. Two tiers of application servers disrupted the monolithic central system. The number of interfaces which were created by that is one of our today’s challenges. Because that change happens in the physical domain the hardware vendors had to adapt, or vanish as well.

    The people mapping: If the programming is done remotely then lets say a design document will be the obvious interface. The remote programmers take over responsibility for their work, which means extensibility, maintainability, documentation etc. Many companies today are on the move, doing the programming remotely. This change happens in the domain of software-engineering. That means to me software vendors and software engineers have to adapt. And they better define their interfaces and protocols well!

    Future mapping: It seems that Moore’s law will hold true at least for the next five years. That means tomorrow’s cell phones will have the computing power of yesterday’s notebooks. Computing power will become even cheaper and much smaller and mobile in the coming years. The amount of things that can be “done remotely” will grow with the available computing power not only of mobile devices. The demand on interfaces and communication is already visible. As with the mainframe story, I claim that industrie will take their chance to make profit with doing things remotely in every domain.


  9. Former Member

    Sidestep – request.

    What needs to come out very clearly in terms of skill is more important than just the 100-ft level talk on outcourcing. It would be great if you blogged once on the Solution Architects of the future and the Developer community skill set that the market needs to be looking out for. Lets get the skills in perspective – Its not about outsourcing always!

    Thank you and best regards

    Kartik Iyengar~


Leave a Reply