The following Bloomberg deep dive into one of the big local government IT titans you've never heard of does a great job of describing how "lock in" works:
A multiyear license for a city of 20,000 people might cost $1 million up front, then bring in $200,000 annually for software support. Moore’s predecessor as CEO used to joke that he had three daughters and a fourth kid named “maintenance revenue.”
This golden child came to haul in hundreds of millions of dollars each year, as clients proved wary of upgrading for fear of a healthcare.gov-like fiasco. The chief information officer of one of the biggest counties in the US—who, like other sources in this story, spoke on condition of anonymity to share an unvarnished perspective and avoid damaging a relationship with Tyler—says its courts stuck with Odyssey’s 2014 suite for nearly a decade because newer editions they tested were buggy and they didn’t trust the company to handle the transition smoothly.
...
Requested fixes require buying support time, and Tyler billed the county $220,750 to complete Gillum’s and the other holdout’s data transfers. “They’re fully aware that they’ve got you by the balls and you’ve got nowhere to go,” the staffer says. (Tyler says that it’s looking into the county’s software concerns and that Gillum’s office didn’t participate in necessary training or project meetings that would have reduced invoices.)
Put more politely, it’s very unlikely that Tyler’s sky-high 98% client retention rate is purely because of product satisfaction. There’s clearly a measure of vendor lock-in, too. A common refrain among customers is that they feel they’re in too deep to back out. After investing years and millions of dollars in a fraught implementation, can they afford to do it again with a lesser-known alternative?
Source: Bloomberg "Tyler Tech's Odyssey"
Jennifer Pahlka provides a great succinct summary of what led to this moment in a recent testimony to the US Senate Committee on Homeland Security and Government Affairs.
How did we get here? A little history helps explain. Starting as far back as the 1960s, but particularly in the 1990s, when companies like Amazon and Google were emerging, leadership in government (both Democrats and Republicans) mistook what ultimately proved to be a massive digital revolution for a mere tactical shift in the tools of implementation. Tools are things you buy, so leadership saw digital as a problem of purchasing. Instead of recognizing that no institution, public or private, would be able to operate effectively in the coming decades without basic digital competence, and therefore hiring people who understood this brave new world, our government developed extensive processes and procedures for buying digital technology as if it were simply a commodity. Today, as we bemoan the lack of expertise in highly specialized, complex domains like advanced software, it’s worth noting that the inner workings of procurement seem as specialized, complex, and mysterious to the layperson as the inner workings of an AI model. Government is clearly capable of developing capacity in specialized domains. We just picked the wrong ones.
We have treated digital much like we treat pens, paper clips, or vehicles that the General Services Administration buys for agencies: we don’t need to know how it works, we just need to acquire it. Once we’ve acquired it, other than perhaps a maintenance contract, we’re done. Today, though it takes us a painfully long time to do so, government knows how to acquire static software. What we need to acquire are capabilities.
