Why successful application modernization can be so hard to do

The goals behind an application modernization project are generally very clear – better technology, a more efficient process, more security and better outcomes.

But there are multiple round blocks to meeting those goals. There are disconnects between what agencies say they want and what and how they buy.

First is the question of what is application modernization. The term is too often used synonymously with digital transformation, according to executives we gathered for a discussion around the challenges and opportunities of application modernization.

Our conversation was on-the-record, but not for attribution. See the sidebar at the end of this article for a list of attendees.

Digital transformation takes a much broader view of the technology lifecycle. App modernization is usually a significant portion of a digital transformation project.

But not understanding the nuances between the two can lead to problems and is a barrier to success, the group largely agreed.

The disconnects and misunderstandings result in a success rate of just 30% for application modernization projects, one exec said.

One risk is that the approach to an application modernization project can be too narrow. The customer often doesn’t articulate the goals it wants to achieve.

“If you have a path forward, you have a higher chance of success,” one participant said.

A critical need is a technology roadmap that ties into mission objectives.

“If that’s not there, you are just throwing your money away,” another said.

Industry also can be to blame because they view an application modernization project as a transaction.

“The customer sees it very differently,” an executive said.

A thorny question is how you modernize a system that must continue delivering on a mission, they said, likening the challenge to changing the engine in a car while it is driving.

Many agencies are relying on 20-year-old technologies that have become harder and more expensive to maintain, but the end users are saying: “Don’t touch it. It’s working. Don’t break it.”

Several executives focused their remarks on the importance of a technology roadmap. While acknowledging its importance, many agencies are challenged to create that plan.

“How do you build a roadmap when you don’t understand the technology and, in some cases, you don’t understand the problem,” one said.

Agency leaders often come in with plans for upgrades and replacements, but there isn’t a discussion about the why, executives said.

There also are concerns that this issue will be exacerbated at least in the short term with the new administration coming in with new leaders across the agencies.

“Strategies are going to change, and everybody will have to figure out what that means, so no one is going to have that road map we are looking for,” an exec said.

Putting the administration change aside, industry needs to understand that different agency officials look at application modernization through a different lens.

Chief information officers look at app modernization in terms of security, cost and the platform. From the operational side, the agency thinks about whether the application does what is needed for the mission.

Contractors often need to be the conduit that communications between those two sides and those conversations come back to what the agency is trying to accomplish.

“The outcome can get misunderstood,” one said.

No matter what the outcome the customer is looking for, many said a critical element is being able to measure that outcome.

One participant said the government needs to stop buying software and services and labor all as separate items. The person added agencies should shift toward buying measured, service-based outcomes.

But that shift is hard and expensive because it moves away from the traditional labor-based model and toward a cloud-based, as-a-service model.

The funding mechanism will change as well. One example given was a financial system that moved from labor hours and software licenses to an as-a-service model.

Under the traditional model the government has some discretion on how much it spends. But one said that in a consumption model, “you don’t have a choice, it's a must fund."

Agencies face a technical debt that drives the need for application modernization and also have a workforce debt at the same time.

“When the government doesn’t have the ability to discern the value of the next technology, they rely on the contractors they have in place,” one executive said.

Many of those contractors rely on a business model the is labor-based, which many said makes moving to the model challenging because the contractor doesn’t have an incentive to change themselves,

While the executives were quick to describe the many challenges and roadblocks, they also cited some examples of successes. The key here is also taking a wholistic approach.

One agency has been using a service model for a decade and they are meticulous about documenting their process, what has worked, what hasn’t, etc. A second element is the cooperation across the agency.

“The director, the platform owner, the business analyst all worked together,” an exec said.

Leadership and relationships are critical.

“There has to be courage and trust,” another executive said. “We must have government people with the courage to try something new and the trusted relationship with their contractors to go ahead and do it. And when they run into problems, they lean on that relationship.”

PARTICIPANTS

Seth Abrams, Leidos

Alex Adamcyk, LMI

Jeff Bateman, Veeam

George Batsakis, Groundswell

Robert Carey, Cloudera Government Solutions

Chris Copeland, ManTech

Angel Davis, Presidio Federal

Brian Drake, Accrete AI

Chrystal Hair, A-Tek

Mia Jordan, Salesforce

Dennis Lucey, Akima

Jay Olsen, General Dynamics IT

Mike Raker, Maximus

Nic Skirpan, Bravium Consulting

Barry Snyder, ATi

Don Styer, Serco

Conrad Symber, CGI Federal

NOTE: Washington Technology Editor-in-Chief Nick Wakeman led the roundtable discussion. The November gathering was underwritten by Veeam, but both the substance of the discussion and the published article are strictly editorial products. Neither Veeam nor any of the participants had input beyond their comments.