Posted by Thomas Antunez on Wed, Feb 10, 2010
What's old is new. Bell bottom jeans, aviator sunglasses, big band jazz, vertical integration...err...yes, vertical integration. By most historical accounts, Andrew Carnegie introduced the concept of vertical integration some time ago, and in the recent past we have seen business process outsourcing essentially debunk the merits of this once cherished business practice. If you need a primer on vertical integration use this Wiki link: http://en.wikipedia.org/wiki/Vertical_integration. Cheap labor abroad drove much of the drive around outsourcing, reinforced by a increasingly educated set of foreign workers who are inexpensive when compared to their US counterparts. However, the economic turmoil experienced from 2007 to the present has stressed companies in multiple ways, so much so that organizations are revisiting the concept of vertical integration (VI) in an effort to improve the quality of their products, protect the supply chain from systemic risks, hedge currency fluctuations, and most importantly, use VI as a means to enhance the customer experience. Which leads me to the point of this blog: Oracle's acquisition of Sun Microsystems.
By all accounts, Oracle|Hyperion is a tiny fraction of the overall Oracle technology stack, albeit a very critical one. Oracle's acquisition of Hyperion was indeed a trendsetter, noted by the quick reaction of SAP and IBM. But setting trends for the sake of setting trends is foolish: Oracle realizes the importance of the Chief Financial Officer and wants access to that office. So, when viewing the impact of the Sun acquisition on Oracle|Hyperion EPM, I feel justified in composing this blog despite the rather small footprint of EPM on the overall stack.
So, what does it mean for Oracle|Hyperion EPM? Well, in the short, term, not much. There is no question that Oracle has made great strides with the original Hyperion software offering, enhancing the customer experience with new features and functions, as well as integrating the technology with ERP solutions for drill back and tighter integration. The Sun acquisition, in immediate terms, is a long term strategic play and the short term effects for EPM are minimal. Things do get interesting in the long term, in my opinion.
The long term effects of the Sun acquisition will be felt throughout the Oracle technology stack, Hyperion EPM included. Here's a very brief sampling (clearly not exhaustive) of what I believe we will see in the next 3 to 5 years.
1. Oracle|Hyperion EPM in a box
- "EPM in a box" offerings will begin to surface, essentially integrating hardware and software for a simple plug a play solution. Today, a customer will make a license purchase of Oracle|Hyperion EPM (think Planning, Essbase, HFM, etc) and will then make a determination on what hardware to use. Most of the time customers select a set of servers which run a version of Microsoft Windows, although they do have the option of selecting UNIX based environments (in my experience I have only really seen Essbase used with consistent success on UNIX boxes). Customers then procure the hardware, install the OS, then have a consultant come in and install the Oracle|Hyperion components. Sometimes we see "fancy" setups that use SAN environments but mostly the work is restricted to the set of boxes designated for Oracle|Hyperion EPM use. In the new world, Oracle will literally deliver a server(s) to your server facility, or even host it for you, that has everything installed and ready to go with a few minor tweaks (security authentication, etc). They may even have that locked down with enough pre-installation contact with the client. So, from a client's perspective, after conducting your due diligence on EPM solutions, all you need to do is take delivery of your box and plug/play. Ease of use cannot be understated in this regard and neither can a "clean install". I have seen companies struggle intensely with bad installs, which ultimately has led many to completely reinstall the software. A bad install can haunt you for a very long time. In the future, a customer will simply identify what they want to buy (Hyperion Planning, Hyperion Essbase, HFM, etc) and the servers will show up at their data center, ready to go. There's just something beautifully elegant about preconfigured hardware and software, particularly when they are engineered to go together (if you don't understand what I mean...look at Apple and their suite of products which unite software and hardware).
2. Reduced strain on infrastructure related issues
- Now that we have EPM in a box, we can expect that infrastructure related issues with the hardware/software should begin to subside in a material way. Today, we constantly see issues with upgrades that occur on the server OS which then cause a conflict with the Oracle|Hyperion EPM software and subsequently require the trained eye of an Infrastructure Consultant that specializes in this area. Tomorrow, the hardware will be specifically built to support Oracle|Hyperion EPM. Any changes to the hardware will be tested in advance of release, to be leveraged by the client and thus avoiding conflicts. Infrastructure related issues can be a huge problem for customers as they fight to keep their end users happy. End users tend to really enjoy the rich analysis that Oracle|Hyperion EPM offers and they place the system at the center of their daily activities. Any downtime is frowned upon and can reduce end user satisfaction quickly.
3. Improved scalabilty
- Today, there are hardware guidelines provided Oracle depending on the Oracle|Hyperion EPM software tools purchased, however, there is this grey area of what to recommend for large scale deployments. Don't get me wrong, no other EPM system, and I REALLY mean no other, can scale the way that Oracle|Hyperion EPM can scale. However, when recommending hardware to a client, certain variables such as user counts, etc must be factored in. Most of the time, consultants will use the Oracle guidelines as a proxy for recommendations and combine that with their own experience. In the future, this new "purpose built" hardware may render experience and recommendations irrelevant. Plainly stated, if you are buying Oracle|Hyperion Planning with full use Essbase and you are rolling this out to 1,000 users, well, your software will arrive preconfigured to work with the right amount of hardware for that task. I love it...let's hope it happens.
4. Tighter integration with...well, just about everything
- Without question, the key benefit I can foresee to clients lies in the ability to unite the organization from a technology perspective. Think about it: start at the top with strategic planning tools like Hyperion Strategic Finance to set out goals and model high level future movements in the business, then use Hyperion Planning to take this down to a detailed level, only to be supported by seamless drill thru functionality to E-Business Suite, which will be linked tightly with demand planning tools like Demantra, etc. In theory, a customer could have set of Oracle|Sun servers completely built out to support their entire supply chain from Design to Manufacturing to Financial Consolidations and Reporting. Today, these systems exist in various capacities, often linked via a hodgepodge of servers throughout the company (and, heaven forbid, with the use of Excel-macros, etc). I know of consultants that have built entire careers designing, building and maintaining data interfaces. Data interfaces can be a huge burden on any organization and tighter integration of the entire Oracle stack, thanks to good hardware/software engineering via the Sun acquisition, can really put a significant dent into this area. This is a huge competitive difference when you think about the longevity, sustainability, and maintenance of the system.
5. Countless other benefits that I have missed...
Obviously I have simplified things greatly. The bottom line is that the Sun acquisition is good for Oracle and good for Oracle|Hyperion EPM. Stay tuned. I'm sure Larry and Co. will be unveiling more and more innovation related to this deal over the next few months...