I'm in the smaller sliver of the venn diagram who have done health IT and auto dealer IT. This post about DMSs is hilarious and now I'm going down the rabbithole of @Matt Stoller's article.
This quote is gold and very on-brand. There are definitely mafia folks in the auto dealer industry:
CDK executive Dan McCray delivered a mob-style ultimatum to competitor Authenticom: "For god's sake, you have built a great little business, get something for it before it is destroyed otherwise I will f***ing destroy it."
I hated Reynolds & Reynolds - my company had to do screenscraping through another third-party integrator, and it always created such abrasion with the F&I managers.
I love the critique of our ‘little’ industry and your writing style Brendan. As to the behemoth pushing it’s way forward in this arena, at a minimum we are going to see the established three (Microsoft/Nuance, Sufi and Abridge) come under severe price compression due to being compared to the incumbent native solution. That will radically change the competitive landscape for each of them. DAX and the Nuance stack are already suffering from the typical atherosclerosis of those acquired by Microsoft and it shed the native corporate knowledge by doing its typical 1 year purge of cost and overheads.
As to the other two, the ball is in their court….adapt or die!
Bundling is the inevitable endgame when APIs commoditize — you compete on distribution and trust, not just features. The compliance layer is where this gets interesting in health: bundled platforms are suddenly responsible for the entire PHI surface area across every integrated service. That audit trail complexity compounds fast. Really sharp framing here Brendan — the "who owns the patient relationship" question maps directly to "who owns the compliance obligation."
Antitrust is hindered in these cases -- (1) market is oligopoly of 2 - 5 not monopoly -- Sherman Section 1 "restraint of trade" applicable not Sherman Section 2 -- monopolization (2) current economic theory of harm -- raising rival costs -- in this case switching costs -- hard to measure (3) Shumpeterian creative distruction will happen faster than DOJ getting around to sue, win, and remedy
Awesome work Brendan. I love the interweaving and cross industry comparisons in particular.
As a participant in the Healthcare IT market for almost 4 decades (25 years here in the US) it is striking how as I watch the same patterns evolve in almost every business. I recall when SMS (Siemens) dominated health rev cycle when I arrived here in the US in 2002 and were impossible to dethrone/integrate with. And yet. 2 decades later. Hmmm.
I recently wrote up a lessons to learn from Watson Health (and IBM’s repeated failed efforts to crack into healthcare provider space beyond being an infrastructure provider). That pattern is repeated over and over again over the 4 decades I’ve been at this.
Market dominance (and therefore technological lock ins) starts by solving a ubiquitous problem the market has so well, that everyone has to have it. Judy and Karl’s vision wasn’t to help doctors, it was to help health systems CEO’s and boards get tighter control of hospital operations and wrestle it away from the doctors. Obama and the ACÁ just threw a gas can full of money and regulation in the form of Meaningful Use that accelerated the rise of Epic.
At Eclipsys/Allscripts we were too busy with Sunrise trying to solve for the doctor adoption problem by providing flexibility and usability for doctors in configuration. Judy’s ruthless standardization approach satisfied the boards looking for control and predictability, at the expense of any objections from the doctors.
By the time we cottoned on, it was too late. They were on a roll.
Brendan: It's hard for me to square some simple facts.
I know finance has implemented new technologies almost as soon as they were invented. Algorithmic trading and financial transfers were already running on AI and cloud infrastructure long before most other industries even got started.
Semiconductors and manufacturing were ready for AI decades ago. Sensors, robotics, and real-time optimization were the norm. I remember interning at Texas Instruments in the 1990s. Shutting down an assembly line for even five minutes could cost the company millions. Everything was automated down to the last millimeter.
And today? Robots are absolutely crushing it at Tesla factories. Self-driving trucks are becoming more of a rule than an exception.
So it's really difficult for me to believe that real estate and banking face the same API and integration issues as healthcare. But hey, what would I know. 🤷🏼♂️
That's a bit of a red herring fallacy, no? The discussion is not about readiness of semiconductors for AI, assembly line operational processes, or use of robots. It's about how systems of record across all industries share the same pathological patterns of vendor lock-in, poor integration, and monopolistic behavior despite their users' widespread dissatisfaction.
At a certain point, if direct data and independent articles about different industries doesn't persuade you and instead you feel that pointing to arbitrary use of technology unrelated to the core points of the article (systems of record) is more validating, there's not much I can do to change your beliefs. You trust your own internal judgement more than any detailed research.
... and speaking of Meaningful Use (unique to healthcare in your examples), it's important to remember exactly HOW the $30 billion was spent:
Dr. Bob Wachter: "I asked Brailer an unfair question: Given his well-known skepticism about too muscular a federal role, if he had still been ONC director in 2008, would he have turned down the $30 billion?"
Dr. David Brailer: "No, but I would have spent the money on standards, interoperability, a ‘Geek Squad’ to help with training and implementation, and creating a cloud-based ‘medical Internet.’ I never would have spent money on direct subsidies to providers. We’ve built the Frankenstein I was most afraid of."
We could have/should have spent the $30 billion more wisely. We didn't.
Footnote: FWIW I wrote a few articles on interop back in the day. They're dated now but the plot lines are still relevant today: https://dan-munro.com/portfolio/interop/
Love the convo, I'm here for the there will be blood memes. Pls more.
I'm in the smaller sliver of the venn diagram who have done health IT and auto dealer IT. This post about DMSs is hilarious and now I'm going down the rabbithole of @Matt Stoller's article.
This quote is gold and very on-brand. There are definitely mafia folks in the auto dealer industry:
CDK executive Dan McCray delivered a mob-style ultimatum to competitor Authenticom: "For god's sake, you have built a great little business, get something for it before it is destroyed otherwise I will f***ing destroy it."
I hated Reynolds & Reynolds - my company had to do screenscraping through another third-party integrator, and it always created such abrasion with the F&I managers.
I love the critique of our ‘little’ industry and your writing style Brendan. As to the behemoth pushing it’s way forward in this arena, at a minimum we are going to see the established three (Microsoft/Nuance, Sufi and Abridge) come under severe price compression due to being compared to the incumbent native solution. That will radically change the competitive landscape for each of them. DAX and the Nuance stack are already suffering from the typical atherosclerosis of those acquired by Microsoft and it shed the native corporate knowledge by doing its typical 1 year purge of cost and overheads.
As to the other two, the ball is in their court….adapt or die!
Thanks. As an ER physician, I see the need to improve things drastically. Still learning the techno terminology…
Amazing content, thank you for the piece Brendan
Great article and loved the connection to other industries
This was legit Sunday reading
Unreal article, very well done.
Bundling is the inevitable endgame when APIs commoditize — you compete on distribution and trust, not just features. The compliance layer is where this gets interesting in health: bundled platforms are suddenly responsible for the entire PHI surface area across every integrated service. That audit trail complexity compounds fast. Really sharp framing here Brendan — the "who owns the patient relationship" question maps directly to "who owns the compliance obligation."
Antitrust is hindered in these cases -- (1) market is oligopoly of 2 - 5 not monopoly -- Sherman Section 1 "restraint of trade" applicable not Sherman Section 2 -- monopolization (2) current economic theory of harm -- raising rival costs -- in this case switching costs -- hard to measure (3) Shumpeterian creative distruction will happen faster than DOJ getting around to sue, win, and remedy
Great thoughts! Next article gets into antitrust implications
Awesome work Brendan. I love the interweaving and cross industry comparisons in particular.
As a participant in the Healthcare IT market for almost 4 decades (25 years here in the US) it is striking how as I watch the same patterns evolve in almost every business. I recall when SMS (Siemens) dominated health rev cycle when I arrived here in the US in 2002 and were impossible to dethrone/integrate with. And yet. 2 decades later. Hmmm.
I recently wrote up a lessons to learn from Watson Health (and IBM’s repeated failed efforts to crack into healthcare provider space beyond being an infrastructure provider). That pattern is repeated over and over again over the 4 decades I’ve been at this.
Market dominance (and therefore technological lock ins) starts by solving a ubiquitous problem the market has so well, that everyone has to have it. Judy and Karl’s vision wasn’t to help doctors, it was to help health systems CEO’s and boards get tighter control of hospital operations and wrestle it away from the doctors. Obama and the ACÁ just threw a gas can full of money and regulation in the form of Meaningful Use that accelerated the rise of Epic.
At Eclipsys/Allscripts we were too busy with Sunrise trying to solve for the doctor adoption problem by providing flexibility and usability for doctors in configuration. Judy’s ruthless standardization approach satisfied the boards looking for control and predictability, at the expense of any objections from the doctors.
By the time we cottoned on, it was too late. They were on a roll.
Anyway love the analysis!
Brendan: It's hard for me to square some simple facts.
I know finance has implemented new technologies almost as soon as they were invented. Algorithmic trading and financial transfers were already running on AI and cloud infrastructure long before most other industries even got started.
Semiconductors and manufacturing were ready for AI decades ago. Sensors, robotics, and real-time optimization were the norm. I remember interning at Texas Instruments in the 1990s. Shutting down an assembly line for even five minutes could cost the company millions. Everything was automated down to the last millimeter.
And today? Robots are absolutely crushing it at Tesla factories. Self-driving trucks are becoming more of a rule than an exception.
So it's really difficult for me to believe that real estate and banking face the same API and integration issues as healthcare. But hey, what would I know. 🤷🏼♂️
That's a bit of a red herring fallacy, no? The discussion is not about readiness of semiconductors for AI, assembly line operational processes, or use of robots. It's about how systems of record across all industries share the same pathological patterns of vendor lock-in, poor integration, and monopolistic behavior despite their users' widespread dissatisfaction.
At a certain point, if direct data and independent articles about different industries doesn't persuade you and instead you feel that pointing to arbitrary use of technology unrelated to the core points of the article (systems of record) is more validating, there's not much I can do to change your beliefs. You trust your own internal judgement more than any detailed research.
Bundles… EHR…
Learn about them all here: www.offlabelideas.com
... and speaking of Meaningful Use (unique to healthcare in your examples), it's important to remember exactly HOW the $30 billion was spent:
Dr. Bob Wachter: "I asked Brailer an unfair question: Given his well-known skepticism about too muscular a federal role, if he had still been ONC director in 2008, would he have turned down the $30 billion?"
Dr. David Brailer: "No, but I would have spent the money on standards, interoperability, a ‘Geek Squad’ to help with training and implementation, and creating a cloud-based ‘medical Internet.’ I never would have spent money on direct subsidies to providers. We’ve built the Frankenstein I was most afraid of."
We could have/should have spent the $30 billion more wisely. We didn't.
Footnote: FWIW I wrote a few articles on interop back in the day. They're dated now but the plot lines are still relevant today: https://dan-munro.com/portfolio/interop/
Thanks! Yeah, once you see it, you can't unsee it - systems of record are all so similar in this regard