Property/casualty insurance carriers embarking on massive projects because “the system is too old” are making one of the most common technology mistakes in the industry, according to an executive of a technology company.

In addition, many insurers make the mistake of putting checklists of desired features on requests for proposals from tech vendors without paying attention to the quality of the providers in delivering on those features, said Michael DeGusta, chief executive officer of ClarionDoor, a cloud SaaS platform for distributing insurance products with rate, quote, bind and issuance capabilities.

DeGusta, who delivered his remarks at the InsurTech NY Spring virtual conference earlier this month, contrasted the RFP process for tech projects to the experience of buying a car. “If you went to buy a car, you wouldn’t say, ‘Oh, this Porsche has an engine, has doors, has a cupholder. But wait, over here, we have a Ford Fiesta [with] an engine, doors, but two cupholders, so it must be better!'”

“That’s not how the real world works. Quality is really what matters more than long checklists,” he said.

Still, DeGusta gave a list of his own—a list of technology mistakes to avoid. “The common thread of all these mistakes…is they don’t really feel like mistakes,” he said, warning attendees not to be surprised if one or more of them was happening at their companies today.

Along the way, he gave suggestions to fix the mistakes as well. For example, after noting that the vendors who win the jobs after an RFP process are often those that are better at “jumping through the hoops” of the RFP processes with hefty documents responding to the carrier checklists, he suggested an alternative. Specifically, he recommended that carriers contact every vendor that might be interested and tell them—concisely—what business problem they are trying to solve. “Give them one hour to make whatever presentation demo they want. Nine times out of 10, it’ll be pretty clear which two or three is worth spending more time with,” he said.

Here’s some of what else DeGusta had to say about the first two common mistakes and a few others.

If it ain’t broke, don’t fix it. “It’s actually only bad technology if it’s actually stopping you from pursuing your business needs,” he said, noting that an existing system being too old or technology “being generically, amorphously better somehow isn’t justification to spend years and millions of dollars distracting from the real priorities,” DeGusta said. “Companies really need to ask themselves what is the actual business problem you’re trying to solve,” he advised.

The RFP. Calling RFPs “the most overdone and underthought processes ever created,” DeGusta observed that they take up massive amounts of time for all parties—immediately adding months and months to a project before it even gets off the ground.

Another drawback is that RFPs are written based on what worked in the past. “If you really did do that RFP for a car, it would probably ask questions like, ‘How many cylinders does it have?’ You’d miss the Tesla all together,” he said.

Five-Year Implementations. DeGusta believes that nothing good happens after 12 months when it comes to technology, and said that any carrier that’s in Year 3 of a five-year implementation is doing something wrong. The fact that companies are making “unnecessarily huge, scary, multiyear technology bets” is what’s causing the slow pace of change in insurance technology. “And then two years in, it’s already a year behind. But at that point everyone’s too invested and too afraid to admit it was a mistake, so people keep pushing ahead.”

Insurance Companies Aren’t Tech Companies. DeGusta stressed that insurers should focus on their roles of developing innovative insurance products, launching them and managing risk rather than becoming technology companies. “No one tries to write their own spreadsheet app, and now people are finally starting to realize that maybe Amazon is better at managing servers than we all are. We need to keep pushing that attitude further up the stack and making sure we’re enabling insurers to focus just on what makes them different, not on building technical infrastructure,” he said.

Other tech stumbling blocks for insurers that DeGusta touched upon were being attracted to buzzwords (like blockchain proofs of concept at companies that are behind the curve on basic technology projects), trying to find multi-tools that claim to solve all problems at once and neglecting the foundations for API software.

“So much progress has been made in insurance technology, but certain habits seem hard to break—and just keep leading insurers into painful experiences with their systems and their software vendors time and time again,” he said.