C R M I T - Solutions On Demand

Any professional who has experience with these financial, HR, manufacturing, and other business apps will be intrigued–and suspicious. ERP is notoriously complicated, so the prospect of no-hassle implementation and management is hard to ignore. Then reality sets in. What about data security, scalability, regulatory compliance, systems integration, and customization?

Software as a service is starting to gain favor for several reasons: It’s relatively easy and inexpensive to implement; it’s flexible; it doesn’t require as much infrastructure; and its costs are more predictable. Gartner predicts that 25% of new business software will be delivered as services by 2011, up from 5% last year.

Gartner estimates that software services accounted for 8% of CRM revenue last year and will jump to 12% this year.

There are a few fundamental truths about the IT industry one picks up along the way, but which can sometimes become hidden by the rapid innovation in technologies, and the complexities of delivering business outcomes in the real-world.

Any IT system that doesn’t respect the business operating model will always fail (although it may cost lots of money and take years to do so)

Just like any management discipline, IT strategy, planning and execution is more about spinning plates than juggling balls – it’s as important to know when to leave something alone as it is when to intervene

And so for all the inherent benefits of Software as a Service (SaaS) – the economics, the ubiquitous reach, the enhanced usefulness and the enhanced customer services – perhaps it is how the SaaS model helps or hinders these 3 truths that will ultimately determine its business value relative to other IT delivery models.

What’s interesting is we don’t have to look too far into history to see where we might end up this time.

Before we go a little further I want to declare a bias. No matter how one slices and dices a major business, it is just that – sliced and diced. Sliced and diced means management control in the slices and dices (the business units) – with varying degrees of freedom to act for sure – but still freedom to act by definition.

So, the evolutions of departmental computing and the networked PC helped the 3 truths because the slices and dices could act as they needed to (enabled by IT), they could communicate with other (and more importantly the centre could communicate with them), and they could be left to it without expensive ‘corporate IT’ getting in the way.

However, over time, the need for enhanced corporate compliance, cost control and re-engineered cross line of business processes to deliver efficiencies all kicked in, which is pretty much where the corporate IT function has been busy in bringing centralised IT control to operating models for the past 20 years.

What’s truly interesting is that individual and business unit SaaS use – which is currently on the rise – fans the flames of many of the old corporate tension points the CFO and CIO have been spending years trying to address.

So are we going to see central control kicking back-in again?

Intriguingly, the world isn’t like the world was when we had the networked PC and the departmental computer for one underlying reason – the Web. Today, the world has an unprecedented level of connectiveness – through globalisation and through the Web – and SaaS is based on the Web model from the ground up – i.e. we get to support the slices and dices and enable ease of communication (and therefore control, orchestration and information access) across units, employees, partners, customers, suppliers, the corporate centre and so on.

And while just as with any approach there are many challenges with the model, to my mind, the predicted rise of SaaS has been underestimated not only because of the obvious benefits, but perhaps more so because it serves the 3 fundamental truths of IT so well.

Are ‘per-business’ package and custom solutions no longer valuable? For sure they are valuable and for sure they will likely be of value in many circumstances for many years. But I wonder whether SaaS – particularly SaaS that supports Web 2.0 and Web 3.0 evolutions – is pointing the way to IT for the 21st century. And that its predicted rapid rise has been if anything undercooked.

When I get around to buying a new car, I know what is important to me. Being 6 feet 4 inches tall the first criteria is whether or not I can sit comfortably and see myself driving many miles, and then small things become important, like whether I can see the lights at an intersection. The comfort criteria is independent of the car and must be applied to all cars that I try, although I do like the image of relatively fast cars. When you look at what I drive, it might surprise you, I have a Ford Focus SVT which is good for 140 mph and 0-60 in about 6-7 seconds I am told. Many people have asked why I do not drive a SUV or similar and the reason in many cases for me is that they typically have less room for a driver of my proportions than my Focus.

When I bought my car, there was a lot I already knew. Having driven many cars, I knew where the steering wheel, radio, rearview mirror, etc. were and I knew that I could drive the car without a specific driving course, although to push the limit and understand how it would go 140 mph on a sweeping right hander would probably require a day at some race track. My wife on the other hand has to transport kids to the school and to soccer practice so her vehicle of choice is a minivan.

Now that I have shared with you some of the reasons how my family chooses cars, you may be wondering how this relates to SOA. Well, last week in a meeting with customers and prospects alike, I was getting questioned on selection criteria for ESBs and SOA infrastructure technologies.  And in nearly every conversation, the same criteria was clear - how many messages per second, how much memory, etc. etc., and only a few really came up with a statement such as "We are facing challenge X and would like to solve it?" This is the most important question because just like my car challenge of "sitting comfortably for long periods of time" outweighs the need to do 140 mph (which I have yet to do), if I bought a car that does over 150mph, it would probably be gone by now as it might not meet my primary needs. And lets face it, how fast do any of us really need to drive? Even the cheapest car on the market will do 80mph all day.

Using standards such as Web-Services, JMS, etc. also means that developers can apply the same disciplines to one technology as another - in much the same way that most drivers can switch cars. If it did not, Hertz and other companies would not be in business. Which brings me to the second theme of questions and comments I was getting, "What do my developers have to learn to be successful?" They have to learn to transition from one environment to another but still be able to apply the same core software engineering practices. For instance, you would not write all your business logic in one Java class so don’t do the same in one ESB process. In CORBA you would not make every object accessible, so why do you try to make every Java object a web-service?

SOA What? Buy the technology that makes sense for you and your business, e.g. if you are a distributed manufacturing organization build/buy a SOA that can be deployed and managed in a distributed manner.  If you are dealing in high volume transactions or events, you will, like a racing driver, need to look into performance. Once you have chosen your SOA technology, remember to still stop at red lights, and not to redline the engine at every gearshift. The speed limits may have increased, and you can break them later, but they still exist. And just like I would get special tuition to drive safely at 140 mph, remember to call on the experts of your SOA to drive at the limits.

Let us know if you have questions on SOA Products, we belive that we can help you (sales@crmit.com

I've been thinking about the SOA lifecycle.  What's interesting about it is that while most people acknowledge that SOA changes the traditional Software development Lifecycle (SDLC), we still have no clear definition of what the change actually is.

The most common attempt I see is to think of the SOA lifecycle as "design time, runtime, change time".  But the problem is that even without SOA, everything IT builds still changes - so the notion of "change time" is really just part of the regular SDLC.  This model doesn't tease out the changes caused by SOA infrastructure.

Here's my latest attempt at clearly articulating how the SOA lifecycle is different than the traditional SDLC.

In a traditional SDLC, you can broadly say there are two phases:

• Pre-production (design, development, QA, etc.)
• Production (deployment, operation, etc.)

In a SOA lifecycle, though, there's a new lifecycle phase which fits between these: I call it pre-consumption.

What is pre-consumption?  It's a hybrid of pre-production and production - part of what's being built is in production (some of the services) and part is in pre-production (some of the consumers):

• To service providers pre-consumption looks like part of their production phase since their services are complete and operating in production.
• To service consumers pre-consumption looks like part of their pre-production phase since their consumer application is still being built.

No wonder we have such hard time defining the SOA lifecycle - to any one individual it looks just like their regular lifecycle.  But, when you look at it overall for any given application, you realize that different people are in opposite ends of the lifecycle at the same time.  This is the nature of pre-consumption.

Original Blog Entry from Progress Software Blog < Here > (Thanks to Daniel for Permissions)

Will 2008 be the year of SaaS?

We’ve been talking about SaaS for a long time, so it could be that it is entering fad status. There are articles everywhere about Using SaaS for ERP and CRM, but the licensing has still not caught up.

Organization’s perspective and acceptance of SaaS as a cheaper and more agile method to deliver value seems to be on the increase. SaaS should allow large companies to leverage leading edge business application capabilities immediately, without the same level of understanding in-house about the technical complications upgrades or concerns about IT infrastructure demanded by traditional software delivery approaches. It should free them up to think about what differentiates them from the pack.

There seem to be claims that SaaS will cure everything that’s wrong in an organization’s IT. There have even been efforts to create support groups for SaaS consumers. All of these seem a bit fad like to me, but there is always a hubbub of activity around the kernel of value for any fad.

For those businesses that choose to go it alone, the combined costs of hardware, software and skilled technical staff can be a real drag on profitability. Add in the cost of ancillary (but critical) functions such as disaster recovery and backup, and the total cost of ownership can quickly become burdensome.

But a hosted services model slashes up-front investments as well as ongoing service and maintenance costs, putting top-tier software solutions within reach of even very small companies.

All this activity, however, doesn’t mean the SaaS wave is poised to engulf traditional licensed software. SaaS’s share of the business application market today is more like a drop in the bucket. And enterprises have been slow to embrace SaaS, raising objections over reliability and availability.

Yet the arrival of the big enterprise-software guns, the emergence of integrated business communities in the cloud, and increasing desperation on the part of IT to minimize application deployment and maintenance hassles, suggest that SaaS is on the verge of much faster adoption.

By contrast, CRMIT and other SaaS providers incrementally swap in new functionality, streaming new innovations to all customers at once, keeping the UI as consistent as possible.

Yet the main attraction for SMB customers — letting a service provider shoulder the burden of software deployment, maintenance, and availability — can be a showstopper for large enterprises accustomed to maintaining full control. Not to mention that the huge cost sunk into existing CRM or ERP licenses becomes a whole lot tougher to justify.