Beware Enterprise Applications

by Robin Schofield with Ed Greenaway

Executive Summary

Most companies and organizations are familiar with the problems of custom-coded business applications, such as maintaining financial systems or web applications built in-house once the original developers move on.

In contrast, marketing messages from enterprise application vendors such as SAP, Oracle and Documentum – that sell leading enterprise resource planning (ERP), customer relationship management (CRM) and enterprise content management (ECM) systems – are honed to appeal to executives interested in “industry-standard”, “out-of-the-box” and “scaleable” business solutions.

However, Linksbridge’s experience is whether an enterprise application is in fact a better choice than custom coding strongly depends on the size of your organization and your ability to adapt business requirements to the enterprise application’s design.

Business executives should weigh their options carefully – and also consider a fully outsourced software service – when contemplating whether to embark on buying and building an enterprise application.

The Devil You Know

Most companies and organizations are familiar with the problems of custom-coded business applications, that is software developed in-house or by a software company only for that institution for its specific business purposes.

There was a time when this was the only option available if a company or organization wanted to automate business processes, create collaboration tools, or implement complex web applications.

These systems also conferred significant advantages over competitors when they were successful, e.g. by allowing customers to buy goods and services over the Internet in a way that rivals could not easily emulate at the time.

However, as time has passed, business owners are discovering more and more drawbacks to custom applications.

For example, the original development team moves on leaving undocumented code, while replacement staff members disdain previous efforts and recommend replacement applications that have not been budgeted. Meanwhile, underlying operating systems and databases need updating as they go out of support, breaking the custom code. And, of course, users complain that systems are slow, service is unreliable and interfaces look dated.

Beleaguered executives wonder: Is there not a better way to do this?

The Devil You Don’t

In contrast, marketing messages from enterprise application vendors paint a brighter alternative.

Case Study

A 3,000-person European law enforcement agency wanted to replace its e-mail and records management system, which was going out of support. It picked a large systems integrator to implement a leading ECM system with vendor support. The systems integrator estimated the project would cost $20M and take approximately 18 months to complete.No budget was set aside for application maintenance costs. The selection committee picked the application because the vendor pointed out repeatedly that the system was so flexible that “it can meet any requirement”. This meant users did not have to decide immediately what they did or did not want, or make trade-offs. Instead, the project would iteratively develop functionality until it aggregated into a met requirement. Three years later, the agency abandoned the majority of the project as “unworkable” having spent in excess of $60M.

Companies such as SAP, Oracle, and Documentum that produce leading enterprise resource planning (ERP), customer relationship management (CRM) and enterprise content management (ECM) systems know executives’ pain points well.

They offer “industry-standard”, “out-of-the-box” and “scaleable” business solutions.

It is true that some products have become de facto industry norms, implemented international standards, or worked hard to ensure interoperability with other products or open data exchange.

It is also true that the majority of enterprise applications offer extensive functionality as part of the basic implementation of the software.

Finally, independent testing confirms that only the largest of enterprises need to worry about the ability of leading enterprise applications to cope with user demands.

Nevertheless, the strong qualities of enterprise applications lead executives to three common but poor assumptions:

What is good enough for the big name companies (used as references in marketing materials) must also be good enough for me.
The software can do anything I need it to.
A professionally designed enterprise application will be much easier to maintain than the custom-coded systems it is replacing.

Unfortunately, what is good enough for the big name companies can also mean unnecessary and complicated features for smaller organizations, and a requirement to buy high-end database software, operating systems and computer equipment to support a user demand that does not exist.

As for the second assumption, it is true that enterprise application vendors have created services arms to help clients make their base products fulfill almost any functional requirement. System integrators such as Accenture, EDS and IBM also have large practices that help clients configure and implement ERP, CRM and ECM solutions. However, this assistance comes at a high immediate price in terms of IT consulting, and an even higher repeat cost of upgrading software whenever the enterprise vendor issues new code.

Lastly, Linksbridge finds that many companies and organizations discover that enterprise applications are, in reality, harder to maintain than the custom-coded systems they replace.

This is because developer teams understand the enterprise application poorly, the company or organization has no access to the vendor’s source code, and finding expertise in the enterprise application is challenging and expensive because of the scarcity of professionals skilled in the most popular products.

Also, although many vendors espouse open standards, they create lock-in to their enterprise applications through proprietary features or data standards that make moving away challenging and expensive.

Evaluate Carefully

Instead, our experience is that whether an enterprise application is a better choice than custom coding strongly depends on two factors:

The size of your organization.
Your ability to adapt business requirements to the enterprise application’s design.

We illustrate this in Exhibit 1 (below), where:

The left-hand axis represents the potential benefits to the company or organization of the type of application.
The bottom axis represents the type of application implemented (Custom Build, Enterprise Application or Software-as-a-Service).
The red curve represents the potential benefits to smaller and medium-sized companies and organizations of different types of application.
The blue curve represents the potential benefits to larger companies and organizations of different types of application.

We add “software-as-a-service” in the exhibit as an alternative to enterprise applications for companies and organizations looking to move away from custom-built systems. Software-as-a-service is an application that users access over the Internet without the user (or the company’s or organization’s IT department) having to install anything on their own machine or local server. Software-as-a-service started with applications such as webmail and has grown to include office productivity software (e.g. GoogleDocs and parts of the new Microsoft Office 2010) and CRM software (e.g. Salesforce.com).

What Linksbridge consistently finds, however, is that smaller and medium-sized companies and organizations enjoy better results from sticking to niche custom-coded applications or jumping straight to software-as-a-service, rather than attempting to implement enterprise applications.

This is because these companies and organizations are too small to benefit from the costly functionality and performance offered by an enterprise application and, alternatively, they can satisfy their needs using small custom development efforts to link together office productivity suites or business platforms such as Microsoft Office or use software-as-a-service providers for common business processes.

Conversely, we find that larger companies and organizations have much higher potential to benefit from implementing enterprise applications because their much larger base of users demands dedicated computing power, and their complexity of business requirements warrants a specific implementation for that company.

If You Do Go Ahead

Linksbridge recommends that larger companies or organizations buying and implementing an enterprise application observe rules to ensure projects deliver on time and at reasonable effort. These are:

Only implement an enterprise application if it already closely meets your business requirement with minimal customization.
Simplify and change your business processes to match what the the product does already, rather than attempt to change application.
Do not force every existing application on to the new system, if it is not genuinely a good fit. (System architects will advocate for purity in design where it does not make business sense.)
Have a ranking business sponsor own the project and closely involve real business users in the design.
Implement the new system quickly one process at a time, building on success. Beware of long projects that deliver benefits only at the end of the plan.

About the Author

For more information on this topic, please contact Robin Schofield at robin.schofield@linksbridge.com.

Robin Schofield is a principal in the Seattle office. Robin has over fifteen years experience consulting to leading businesses and non-profit organizations on strategy development, business change, and strategic communications. He has industry experience in financial services, product industries, energy, and the public sector. Robin has a MA from Oxford University, is a graduate of the Royal Military Academy at Sandhurst, and is currently a senior visiting research fellow at London University.

Ed Greenaway, formerly of IBM and Booz & Company, also contributed to this article.