The troubled launch of the U.S. federal government's healthcare information exchange is a high-profile example of outsourced IT gone wrong. The $400 million project, which was supposed to be a one-stop online shop for Americans seeking health insurance, made headlines for its bugs and glitches.
The initiative was endorsed by the highest executive in the world, had plenty of lead time, and had a relatively straightforward mandate. But, as a recent New York Times article pointed out, deadline after deadline was missed on the multi-contractor project for a variety of reasons -- from government agencies slow to issue their specifications to last minute changes to the Healthcare.gov's primary features.
It's not surprising that the project was problematic, says Peter Bendor-Samuel, CEO of outsourcing consultancy and research firm Everest Group. "When you make such a huge change all at once and you're trying to implement systems and processes for the first time, the unintended consequences that cascade are enormous," Bendor-Samuel says. "We see this all the time in business, albeit on a smaller scale. The only surprise to me is that it worked at all."
-- Adam Luciano, principal analyst with sourcing analyst firm HfS Research.
The rocky start, however, serves as a reminder of several steps any IT outsourcing customer should take to ensure a smooth rollout.
1. Conduct robust capacity planning. "Start by gathering feedback from subject matter experts and key stakeholders across all teams to design a solution with the capacity for planned usage profiles as well as the elasticity to meet unexpected levels of demand," advises Craig Wright, principal with outsourcing consultancy Pace Harmon.
"Ensure that outsourcing agreements include meaningful expectations around agile service delivery performance structures and relevant provisions to hold service providers responsible for quickly responding to changing needs, aggregating their services into an ecosystem-wide, seamless, end-to-end service experience for users," Wright says
Insurance providers should keep this in mind in the coming months as well. "The federal government won't be able to tell insurers how many people signed up for their plans until November, so it will be a huge processing month," says Al Denis, director at outsourcing consultancy Information Services Group (ISG). "Few firms are likely to be prepared."
2. Avoid the Big Bang. "When you're coming up for the first time with a brand new economic system, process and structures -- along with the technology to support it -- it's going to take a while to work the kinks out," says Bendor-Samuel. "This is like any CIO that launches a big project without appropriate testing. It's impossible to believe it would go well."
Instead, outsourcing customers should seek out service providers with experience in agile delivery methods that can serve as a catalyst for transforming delivery capabilities, says Wright of Pace Harmon, including frequent delivery of tested, working solution components and progress that can be objectively measured.
"It is important to clearly and contractually define compensation rules for outsourcers based on iterative delivery of value and through incremental release cycles," Wright says. "Companies should also carefully vet the outsourcers' understanding of the link between objective progress measures and value to the payments they will receive." Outsourcing customers should walk all providers through these rules prior to contract signing.
3. Make customer experience a priority. "It's important to keep in mind the experience of the user when designing systems like the Healthcare.gov Web site," says Adam Luciano, principal analyst with sourcing analyst firm HfS Research.
"This is of paramount importance, and is something that many traditional outsourcing firms may not look at with great depth. A big takeaway from this experience is to try to simplify processes as much as possible and keep in mind the user's experience of the product," Luciano says "Healthcare is complex enough, why do these guys need to add more complexity?"
User experience should also take precedence when it comes to error handling and exception processing. Driving customer inquiries to a service desk via phone or chat is a great moneymaker for the service providers. But if lines are choked or agents don't have the proper information, the result is a horrible customer experience.
"It is much better to establish an outsourcing agreement with incentives for a service provider to deploy tools by sharing in the savings and rewarding behaviors and processes that promote and deliver increased adoption of service functions and features that measurably enhance the customer experience," says Wright of Pace Harmon.
4. Anticipate mistakes. Some insurance providers submitted incorrect information to the federal health information exchange, even posting the wrong rates for their plans. But with poor contingency planning, it's not clear how those errors will be corrected. "Now the question arises, if the error is in the subscriber's favor, will the subscriber be able to take advantage of the wrong rate? If so, who will eat the difference?" says Denis of ISG.
5. Get everyone on the same page. In a multi-sourced environment, like the federal healthcare information exchange project, it's critical to use common metrics across all provider agreements. Such "coordination indicators" encourage cooperation among all parties, says Wright.
"Adhere to common industry terminology and process definitions to the greatest extent possible to avoid incompatible service models and the possibility of critical information or transactions getting lost in translation or stuck in limbo," Wright adds.
6. Set standards. "Apply standards judicially throughout the software development and service lifecycle and selectively apply frameworks such as Capability Maturity Model Integration (CMMi) and Information Technology Infrastructure Library (ITIL) to embrace their value-adding processes and functions," Wright says.
Service providers themselves should define expectations regarding the application and adherence to international standards like ISO 29119 for software testing and ISO 20000 for service management to avoid a situation where a solution or service has never been tested and operational readiness is questionable.