Oracle's recent announcement of its "Cloud Box" package, as well as IBM's announcement of NETEZZA acquisition and other announcements, are all related to the trend that surely will be one of the most important events in the near future of information technology: the inclusion of various technological elements under a single "umbrella" (or, in other words, back to the Main-Frame era). Many enterprises' CIOs we have talked to recently report that they engage with over 200 (!) IT vendors annually. And this number is only growing year-by-year.
For example, a typical ERP project in an enterprise organization- the organization should get in touch with about 20 different vendors (if not more) to get "One ERP solution" that includes infrastructure support, databases, operating systems, different application providers for each category. Keeping this in mind, the option of receiving an "ERP solution" within one packaged box while working with a single supplier seems very appealing for CIOs who are looking for simplicity and a single "partner" to work with.
So how did we end up getting back to the same architecture we had "ran away" from just recently? Are we not going to encounter the same problems that pushed us to decide to decentralize our computing, and consequently work with different suppliers in each category? A decision that is reflected in the separation we have made between the layers of hardware, storage, operating systems, Data, applications, management, connectivity? Why do we, all of a sudden, want to get the whole stack in one single box, even if this box is now very nicely packaged and wrapped in nice words like "internal cloud" or "Industry in a box"?
Is it really possible to return to a model in which we have one provider that provides us all the services under its responsibility?
The real reason behind this trend is that CIOs are fed up with the high complexity of managing IT and the search for simplicity (even if that means less choice and less open standards). Business have simply become too complex to manage. CIOs are tired of constantly dealing with optimization issues, the endless worry that the new package will not work properly (for one thousand and one reasons), dreading thoese "middle of the night" phone calls that the servers have crashed again /, the site is not in the air / the organization's main application is not functioning, and the engagement with so many different suppliers.
But before we announce the death of the era of decentralization and the return of the
centralized era, is it also well worth exploring the pitfalls and potential risks of this model:
For example, a typical ERP project in an enterprise organization- the organization should get in touch with about 20 different vendors (if not more) to get "One ERP solution" that includes infrastructure support, databases, operating systems, different application providers for each category. Keeping this in mind, the option of receiving an "ERP solution" within one packaged box while working with a single supplier seems very appealing for CIOs who are looking for simplicity and a single "partner" to work with.
So how did we end up getting back to the same architecture we had "ran away" from just recently? Are we not going to encounter the same problems that pushed us to decide to decentralize our computing, and consequently work with different suppliers in each category? A decision that is reflected in the separation we have made between the layers of hardware, storage, operating systems, Data, applications, management, connectivity? Why do we, all of a sudden, want to get the whole stack in one single box, even if this box is now very nicely packaged and wrapped in nice words like "internal cloud" or "Industry in a box"?
Is it really possible to return to a model in which we have one provider that provides us all the services under its responsibility?
The real reason behind this trend is that CIOs are fed up with the high complexity of managing IT and the search for simplicity (even if that means less choice and less open standards). Business have simply become too complex to manage. CIOs are tired of constantly dealing with optimization issues, the endless worry that the new package will not work properly (for one thousand and one reasons), dreading thoese "middle of the night" phone calls that the servers have crashed again /, the site is not in the air / the organization's main application is not functioning, and the engagement with so many different suppliers.
But before we announce the death of the era of decentralization and the return of the
centralized era, is it also well worth exploring the pitfalls and potential risks of this model:
- Creating a dependency on the IT vendor. Just ask your procurement manager how they feel about working with a single vendor instead of an "open market" with many different options of vendors to chose in each category. The "industry in a box" model means that the organization will work with one major supplier or at least one major supplier for each category. For the procurement manager this trend is a real negotiation problem, since the power now moves to the vendor.
- Connectivity and Integration issues. Most organizations have a heterogeneous application environment with 3-4 major applications vendors. The "boxed" model means that each technological environment will now be more closed off, more proprietary in nature, and less open. This means – connectivity issues. Each environment will be optimized and adapted for the application's main purpose, regardless of other applications' need to connect with it. The supplier will want to give the organization with a box that works flawlessly, with all the different parts "talking" to each other in harmony and are controlled best by a single management tool. What will happen when you try to add a new element into that box? Connectivity problems will appear both on the level of business processes and the Data level.
- Difficulty in achieving an overall monitoring and management view. This will also be a challenge because the "box" providers will include complete solutions with a monitoring tool tailored to the application's specific need. Each solution will include, among other things, a security solution, backup solution, a DRP solution and so forth. This means that receiving information about the organization's overall threat situation will be much harder than before, since there will be a need to synchronize with different information security solutions in order to get to this situation.
- Risk of harming the organization's level of flexibility. If there is a need which has not been addressed in advance by the manufacturer of the packaged solution, the organization can't answer this need (it will require greater efforts to address it). In today's reality, there are many cases in which there are non-standard needs that off the shelf packages do not include. In these cases, organizations can and are doing things that are considered non-standard (adding to the system's components, programming in a way that is not 100% recommended by the vendor etc.). This is almost impossible in the "Industry in a box" model. Boxes manufacturers will not allow to perform actions that do not receive prior approval or "certification" which takes a long time. This means that the level of IT's flexibility in tailoring for the organization's needs will be reduced.
- Significant changes in the status of the IT department as we know it today. The traditional structure of IT organizations' infrastructure department currently includes three separate sections – Network, Storage and Servers. The technological change which we have described (inclusion of all these elements under one box) will require a different structure of the IT Infrastructure unit – from 3 different departments to a unified one. This change will also apply to application teams, and their relationship with Infrastructure teams. If the organization acquires a "CRM box", that also includes infrastructure components, who in the organization will be responsible for the operation and maintenance of the entire box? This is not necessarily a disadvantage, but this is a significant change, and any major change is difficult to cope with.
Vendor's perspective:
IT providers will also experience quite a few changes, some are positive but some are very disruptive and will require a change in the overall vendor's business model.
- Change in the vendor's sales processes. The 'cost of selling' for suppliers in various categories (hardware, software etc.) is a fixed cost. If the supplier will now sell the customer "one solution" which includes the same components inside, rather than selling them individually, the relative cost of selling per each component will fall significantly. The expected cost of selling for suppliers will drop (an advantage that we hope will lead to increased profitability as well as reduced pricing for enterprises).
- Should we expect a decrease in revenues for service providers? It remains unclear whether service providers will experience a decrease in revenues from the sale of services as a result of this trend. If we take as an example the SAAS (software as a service) model, then in this field we can clearly see that services revenue for Integrators have indeed decreased significantly compared with the "classic" model in which an organization actually buys the software and customized according to its needs. A SAAS solution implementation project typically takes between several weeks to two months, almost never includes customizationsm and focuses mainly on training and assimilation. As mentioned above, it is still unclear how this field will affect the overall revenue structure of suppliers.
Who are these providers "Industry in a box"?
NCR have provided a boxed solution long before we called it "Computing in a box" with Teraedta. Among other prominent major manufacturers already offer solutions for Industry in a box we can mention IBM, HP, Oracle, EMC, CISCO, Apple and more. SAP made a major step toward this model recently by acquiring SYBASE.
Take Apple for example. The iphone / ipad industry are actually industries in a box - a box that contains applications, operating system, and hardware. All these parts belong to and are in full control of Apple, which enables it to provide a machine that provides an excellent user experience, optimal performance for applications running on it. But there's no choice, and no openness. The equilibrium exists as long as the box remains closed. Try to open the box and put in elements from the outside - and the balance will be violated.
This example can illustrate pretty well what is expected to happen - as long as consume technology "in the box", we can provide good value. As soon as we try to go out, for example to replace the infrastructure and applications in the box, problems will begin.
This example illustrates one of our essential questions in this area have not yet received an answer: Will manufacturers prefer boxes proprietary technologies, or will they try to adapt to open and common standards?
Recommendations:
- Due to the expected lock-in effect, make an effort to preserve procurement conditions for future deals as well as current deals.
- Organizations should deploy general and flexible monitoring and management consuls that will enable the future connection of "industry in a box" solutions.
- Organizations should start to prepare for the expected organizational changes. For example, measurement of the degree of cooperation between teams and not measurement of each team's performance separately (DBA performance evaluation will be based not by a specific DBMS availability, but according to the overall ERP system availability and performance).
