Our friends at Forrester Research recently published a report entitled Enterprise Apps Vendors Have Issues, which reported on a survey of business process and application professionals at 111 enterprises and their beefs about enterprise software. Their top answers in a nutshell: high cost of ownership (91%), difficult upgrades (87%), poor cross-functional processes (86%), apps don’t deliver on business requirements (80%), and inflexibility that limits business process change (75%). This list appalled me for two fundamental reasons. First, there’s absolutely nothing new about these concerns. Forrester might have conducted the same survey ten or even twenty years ago and gotten the same results. Second, weren’t these problems SOA was supposed to solve?

ZapThink has been encouraging enterprise software vendors to leverage SOA principles in their software packages for years. In 2006 we discussed what software vendors had to do to leverage SOA best practices to improve their products. By 2009, we were asking whether there is a future for enterprise software at all. And today, in spite of a full decade of vendors paying lip service to the benefits of SOA, customers are more than unhappy with enterprise software. They’re pissed. And they’re not going to take it any more.

Paying Lip Service to SOA

Rearchitecting an enterprise software package along SOA lines is straightforward in theory. Modularize the core functionality, expose it as loosely coupled, governed Services, incorporate those Services into a metadata-driven composite application framework that enables customers to support flexible business processes with compositions of Services, whether from the software package or elsewhere. Deliver pre-built compositions to the customer, empowering them to manage and modify the compositions to support changing business processes as necessary.

This SOA-enabled product story is the vision behind products from SAP NetWeaver to the IBM Business Service Fabric, to name a few, but virtually every enterprise software vendor has espoused some version of this loosely coupled, metadata-driven story for their own software packages. Sounds good, right? But here’s the rub: if any of these big vendors were actually getting it right, then we wouldn’t be seeing the high levels of dissatisfaction that the Forrester report is tapping into. Why, then, have all the vendors dropped the ball in such spectacular fashion?

It would be easy to say that reworking enterprise software along SOA lines is too difficult or expensive, or perhaps some vendors say they’re making progress, but the real answer is that vendors haven’t reworked their products to be truly Service-oriented because they don’t have to. As long as customers keep paying them, then why change?

The incumbent vendors have a good thing going, after all: customer lock-in. Once an enterprise installs one of these massive packages, it’s enormously expensive and risky for the customer to switch to another. Furthermore, you have to keep paying maintenance. The more lock-in a vendor is able to achieve, the better their bottom line. For these vendors, the real SOA story — that is, offering truly loosely coupled, composable Services, is actually more of a threat than an advantage. They pay lip service to SOA, but when push comes to shove, the last thing they want to do is modularize and loosely couple their offerings to support flexible processes.

Customer Lock-In vs. Customer Loyalty

These vendors, however, are making a huge mistake. They are confusing customer lock-in for customer loyalty. The Forrester report highlighted this confusion: customers only put up with vendors because they have to, but they’re not happy. As soon as a viable alternative comes along, they’re gone. Customers aren’t loyal to their enterprise software vendors, they just don’t have any choice.

The illusion of customer loyalty has caused the big vendors to become complacent. They believe they’ve successfully co-opted the SOA trend, making it one more excuse to buy more software. Of course, ZapThink has been warning you about vendor-driven SOA for a while now. So, is this battle over, with the victory to the incumbent vendors? Not so fast.

There is change in the air in today’s enteprise IT shops. SOA is a part of the story, yes, but it doesn’t go by the name SOA. This change consists of a broad range of shifts, from the move to Cloud Computing to how software innovation is shifting to consumer technologies to the way collaboration is breaking down barriers to global communication. This change will take a few years to play out, but it will put large software vendors out of business and fundamentally reshape the nature of enterprise IT. How organizations leverage technology at the most fundamental level is undergoing a sea change.

Addressing The Five Top Concerns with Enterprise Software

Here, then, are some of the broad changes in IT that the top five concerns the Forrester report identified are driving:

  • High cost of ownership — SOA can lower the cost of integration over time, but cost savings is a SOA benefit that is secondary to business agility. The open source software movement is also driven by cost of ownership issues, but many organizations have struggled to show substantial cost savings with open source approaches. Software-as-a-Service also has a cost of ownership driver, but perhaps the most disruptive trend today that owes its motivation to cost savings drivers is the move to consumer-centric applications in the enterprise, bypassing established IT purchasing infrastructure altogether.
  • Difficult upgrades — Being able to upgrade software to meet changing business needs is an agility benefit, and as a result, organizations who have implemented SOA and who have purchased SOA-centric applications have an easier time upgrading. Leveraging Cloud-based enterprise applications, however, may not ease the upgrade process if the enterprise doesn’t leverage those applications in a loosely-coupled way.
  • Poor cross-functional processes — Wasn’t middleware supposed to address this problem? Set up complex interconnected webs of integration software, and voila! Seamless cross-functional process automation! Unfortunately, today’s software is not up to this task, because the challenges are more organizational than technological. The key to addressing such human/technology issues is via complex systems approaches that center on governance rather than integration.
  • Apps don’t deliver on business requirements — Traditional IT thinking follows the waterfall methodology for software development that assumes requirements are stable, well understood, and properly communicated. Unfortunately, business requirements are typically dynamic, unclear, and difficult to communicate. Agile methodologies help to address these issues, but Agile approaches don’t help much with traditional enterprise software approaches. The shift toward collaborative, Service-oriented, governed applications is an inherently more agile approach that can finally deliver on the promise of the Agile movement, even for enterprise application functionality.

  • Inflexibility that limits business process change — Providing flexibility for automating dynamic business processes is the primary motivation for SOA, to be sure, but there’s more to this story, because automation isn’t the only capability the business requires. After all, most business processes have human input, and the more collaborative a process is, the less automation the business wants. Here again complex systems thinking comes into play: how best to provide flexibility that supports collaborative business process change? The answer involves an architected web of governed human/technology and human/human interactions. Technology plays a role, but a very different role than traditional enteprise applications do.

It is important to note that the common theme is transformation. There are many enablers of this transformation, including SOA, complex systems approaches, governance, collaboration, and Cloud Computing — but no one of these enablers is the whole story. But make no mistake, there’s a sea change in how enterprises leverage technology.

The ZapThink Take

This sea change will shutter large IT shops, sink unsinkable incumbent vendors, redefine the role of the CIO, and turn the practice of Enteprise Architecture upside down. In ten or twenty years, the whole notion of enterprise IT will be unrecognizable from where it is today. And nothing we or the vendors can do will stop this change.

For ZapThink, however, this transformative time is an exciting opportunity. For us it means taking SOA to the next level, helping organizations understand how to deal with change more effectively — both at the business level and the technology level. We’re stepping up our game to rise to this challenge, working with architects, IT managers, as well as CIOs to help them understand both the risks, as well as the opportunities organizations face as they navigate the changes ahead.

We want you to be part of this effort. Over the last ten years, ZapThink has built an amazing audience of forward-looking professionals around the world. If you’re reading this now, you’re part of that audience. We encourage you to join the discussion as we put together ZapThink’s next act, helping organizations deal with change at all levels. It’s an exciting time!

6 comments for “RIP Enterprise Software?”
David W. Wright Avatar

A few things:

1) "Unfortunately, business requirements are typically dynamic, unclear, and difficult to communicate." Actually, that is common but because most companies are not very good at defining their requirements. It can be done, making a whole lot of issues go away.

2) SOA is a solution approach, but not necessarily the only possible one to to the true issue, which is "Business Agility". Developing software agilely is meaningless if the result is still mired in the need to be updated for business change. You need true business process definition capabilities, for one thing. Another is business rules management and its true goal, decision management. Collaborative processing is not needed when people just follow set rules when asked to make a 'decision'; smart systems can handle those, which make up a vast majority of what people think are decisions. Check out the credit and underwriting systems out there to see real examples.

3)Maybe the 'big vendors' aren't getting it, but others are. *** VENDOR NAMES REMOVED *** the latter's HR system was implemented for me by the vendor in short order based on the requirements I used to select them in the first place.

So, my sympathies with companies locked-in to poor software. Don't give up on buying software completely, just be sure it is actually a good product next time. Knowing your real requirements is step one.

Posted by David W. Wright | June 1, 2010
Vince Vaughan Avatar

I think you are dead on with your assessment that software decisions and antiquated practices have created a situation where Enterprise level businesses aren't able to compete.

I don't know if things are all that bleak for vendors though. ** VENDOR NAME REMOVED ** servers are flying off the shelves and into data centers. As you mentioned, cloud architectures aren't a silver bullet for every situation.

The world has changed though. I'm glad someone else sees it. I don't feel so alone any more.

Posted by Vince Vaughan | June 9, 2010
morrg Avatar

Technology, in and of itself does not resolve any business problem, it never has and it never will, in fact quite the opposte is often true as the survey at the top of the article shows. If you could achieve the SLA's you should be able to implement a SOA solution using Carrier Pigeon, parchment and string.
The trick is in understanding the businesses problem and dividing it in such a way that it shouldnt inhibit business change, model the real world and the rest should fall into place.

Posted by morrg | June 14, 2010
Vince Vaughan Avatar


One problem enterprise companies are facing is they are staffed with people who only know their silo of expertise.

I don't know many architects who understand business problems. They talk about technology modernization, rationalization, and achieving business objectives, but don't understand they need to architect solutions that increase sales, brand awareness, and increase a companies capacity to care for both its customers and employees.

Companies have incompetent people in extremely important roles. Because of this, "the rest" has already fallen into place and we are seeing bloated, overweight solutions and processes cripple businesses while small businesses are ripping market and mind share away from these traditionally large companies.

Enterprise software and SOA are just tools they are using to implement the demise.

Posted by Vince Vaughan | June 14, 2010
morrg Avatar

As the old addage goes - "if you do what you have always done, you will get what you always got"!
I think companies that arent getting value for money from their EA group would benefit from a new EA group! There are individuals that understand the problems involved.
Mind you the fact that organisations are structured at the top in a way that hasnt changed for decades doesnt help - which 'C' level executive is responsible for the business effectivness of processes within an organisation? With this problem at the top, getting the required focus lower down is always going to be a problem.

Posted by morrg | June 15, 2010


[...]Sites of interest we have a link to[...]