Packaged Versus Custom Apps: The Debate Rages On

It wasn’t that long ago that packaged apps ruled the application delivery landscape and custom development was decidedly the second choice. Today, the decision is not so cut and dried, as firms struggle to find the right balance between the quick time-to-market of packages and the competitive distinction custom development can create. In the midst of this shift, a new option — rent (SaaS) — is gaining traction. Most enterprises already support a mix of packaged and custom applications — but with fast adaptation, customer experience, and process integration the top priority for most enterprises, do firms need a different mix of custom, packaged, and SaaS apps to maximize customer value?

Next month at Forrester’s IT Forum in Las Vegas, a panel of experts will debate the pros and cons of custom-developed applications and packaged applications in our Application Development & Delivery track. Attendees will have a chance to vote on the future of applications. But we decided the debate was too juicy to sit on it for another month. So, on April 25, from 2 to 3 p.m. ET, Forrester will host a Tweet Jam — using the hashtag #ITF11 — to answer the question:

“Which is better at delivering customer value: packaged or custom applications? Why?”

We asked our panelists to get the discussion started, and here is what they had to say:

Dr. John Bates, CTO, Progress Software

Neither and both! The pace of business change is so rapid that a packaged app is often outdated and thus inadequate before it goes live! However, packaged apps have great time-to-market — which is also critical. Building your own app has huge risk and horrible time-to-market. So ideally, you need a packaged app you can rapidly configure. I call this a “Solution Accelerator” and I believe it is the only viable path for agile businesses.

George Lawrie, vice president and principal analyst, Forrester Research

Conventional wisdom is that custom development is a high-risk luxury you should use only for a firm’s most critically differentiating processes. As the digital content of firms’ offerings increases as a proportion of their value, Application Development & Delivery Professionals must make shrewder choices about where to apply their custom apps effort. I have, in my research to date, found no magic “technology” that overturns conventional wisdom that packaged apps are less risky, take less time to implement, and are less likely to become obsolete.

John R. Rymer, vice president and principal analyst, Forrester Research

Slow down, George. There are many, and increasing, exceptions to this conventional wisdom. Applications packages struggle to keep up with new requirements in fast-moving industries and are often expensive and risky to customize. And conventional applications are expensive and difficult to scale due to their reliance on relational DBMS. Bottom line for me: Packages work well when implemented for narrow purposes and without customization so they can be incorporated as services into custom compositions.

Tom Grant, senior analyst, Forrester Research

And yet vendors still push back when their customers ask them for more customization options or when they ask for better survival rates for their existing customizations when they upgrade. As long as there’s no negotiated settlement between vendors and customers about how much and what type of customization to allow in packaged applications, there will be convincing arguments for some custom app development. Some vendors offer configuration as an alternative to customization — metadata- or model-based inputs that are more likely to survive an upgrade. But it’s still early in the penetration of this new approach, so it’s hard to tell if it really solves the problem for customers.

Mike Gilpin, vice president and research director, Forrester Research

George, you’ve shared with me some preliminary results from your latest research into where and why firms are investing in custom development. One figure I found particularly interesting was that among your respondents, of the two-thirds of respondents who were running one to four instances of ERP packages globally, a third run 100 or more custom apps, also globally. External customers or suppliers use well over half of these apps, with differentiated innovation or customer experience being the main advantages these custom apps give. Given that, where do you and John agree on what’s happening with custom app development today?

George Lawrie, vice president and principal analyst, Forrester Research

Both John and I see that custom apps can be more cost-effective, in operational terms, if they deliver services better targeted to what the business actually needs. We also agree that custom apps typically do a better job of satisfying today’s requirements, although they may not anticipate future requirements as well as a package that the vendor designs to support many different firms of different sizes and industries. And custom apps may be easier to align with your long-term architecture strategy, since you control it all.

Mike Gualtieri, principal analyst, Forrester Research

I’m with Dr. Bates. “Neither and both” are better at delivering customer value. The real problem is not so much the packaged apps but the way firms buy and implement them. Many firms believe vendor promises that their apps can sing, dance, and do the dishes. Yet many packaged apps have horrendous usability. They are loaded with features aimed to sell more packages rather than deliver real value. Other packaged apps provide beautiful, highly configurable experiences. But the real problem is the way firms make the decision to buy good or evil packaged apps and when to do custom development. They need to help in evaluating build versus buy versus rent.

John R. Rymer, vice president and principal analyst, Forrester Research

Yes, and as George’s data shows, it’s often customer-facing apps, increasingly for those mobile touchpoints, that are the focus of custom development today. But don’t forget that your employees are becoming more like your customers every day, especially the ones who arrived more recently —– they have much higher expectations for what software can do for them than their predecessors. They simply won’t tolerate what I call “work prevention tools” (sub-par user experiences in internal apps)! Or even if they do, your customers ultimately are the ones to suffer.

Mike Gilpin, vice president and research director, Forrester Research

So our research shows there are still areas where it makes sense to do custom development or configuration, especially for customer-facing experiences or where employees’ experiences are especially important — like when they are serving customers. Ever experience a slow check-in at a hotel that you could tell was caused by bad software? Like, you’re thinking, “What in the world is taking so much typing just to check me in?”

Mike Gualtieri, principal analyst, Forrester Research

Oh yeah! And the converse is true — the customer experience can be great because of good design for employee-facing apps and processes. I recently had a great experience checking into a hotel! The doorman asked my name — he had a little headset connected to the front desk. By the time I got there, my key and my welcome packet was ready to go!

Mike Gilpin, vice president and research director, Forrester Research

Thanks, Mike. So better design, innovation, and customer experience, are all strong factors driving firms toward custom development in certain key situations. Thanks to you all for your help in getting this conversation started!


Neither or Both with a whole of Governance

I agree on the neither or both with the caveat that governance is in place to keep the package apps from running amok. The all to common problem with package apps is unchecked integration and extension. The vendor says looky here I have this wonderful API and before you can say spaghetti your agility has left the building.


Decision Services

One way to square this circle is to use explicit decision services to automate business decisions in support of standard packages. Often the need for flexibility/customization and agility comes from a need to rapidly change decision-making in ways that are specific to the company. Externalizing these decisions as Decision Services and using a business rules management system to manage the logic in these services allows the use of standard packaged processes/screens/applications while still allowing a custom interaction. Essentially the decision making is custom. These services also allow custom analytics (predictions based on the behavior of the company's own customers for instance) to be embedded. Many situations have either standard and stable processes or common components that lots of companies can use and these should be purchased in packages to manage costs. But flexibility, agility and customization should be done using components such as Decision Services.