Technology Isn't the Product
Jack Bretcher, COO of PartnerMD, on running IT and operations from one seat, industrializing everything behind a concierge practice that is tripling in size, and why the phones and the scheduling stay human.
Key takeaways
- Technology is a tool, not the end game. Start from the business outcome and then open the toolbox, never the other way around.
- Scale smart: you are not ready for location three until unbiased metrics can tell you why location one works.
- After an acquisition, centralize membership billing, IT, finance, and HR. Leave scheduling and clinical workflows alone if they already work. Rule one is "don't break it."
- PartnerMD has deliberately held off on self-scheduling, chatbots, and call centers. Members pay to talk to a person, and after-hours calls ring a live physician.
- Every non-five-star score and every non-promoter routes in real time to the practice manager, operations, the COO, and the CEO. That loop closed last quarter at a 93 NPS.
- The one number he checks every morning is member terminations, the ultimate vote of confidence in a month-to-month relationship.
Most practices split technology and operations into two jobs and two org charts. Jack Bretcher runs both. He is Chief Operating Officer at PartnerMD, a concierge medicine practice headquartered in Richmond, Virginia, with roughly 12,000 members and thirty-some physicians before its most recent acquisition, and a plan to roughly triple from there.
He came to healthcare from e-commerce software and consulting, spent his first years at PartnerMD running only IT, and added operations once the systems side was stable. That sequence matters. By the time he inherited the operating side of the business, the infrastructure underneath it was already built.
“Technology is a tool. Technology is not the end game. The end game is providing an exceptional service.”
The duck on the pond
Concierge medicine sells a relationship, and you cannot mass-produce a relationship. So what do you industrialize?
Jack’s answer is everything the patient never sees. He calls it the duck on the pond: the calm, high-touch experience out front is only possible because of the furious paddling underneath, defined processes, cross-functional teams, and the tools that support them. Get that machinery right and the front-line team can spend its attention on the person who walked in feeling terrible.
“The core, the guts, are what you have to have in place and truly operationalized to excel on the high-touch piece. It’s not one or the other.”
Don’t scale until you know why location one works
PartnerMD wants to grow, and Jack is candid about what breaks when practices grow the wrong way. His phrase is “scale smart.” A team can be delivering exceptional care and still not be ready to expand, because nobody can point to unbiased metrics that explain the success.
“So many people think, I’m really successful in location one, and because we’re successful here, we can go to location three and four — without putting in the hard work to figure out why you are successful in location one.”
Acquisitions sharpen the point. You are buying a practice because it works, so the first rule is the one Jack has carried through every role in his career: don’t break it.
What gets centralized, and what stays separate
The centralization list is shorter than you might expect. Membership billing, IT infrastructure and support, finance, and HR are the services a ten-doctor practice usually does not have at scale, so absorbing them adds value without feeling like a takeover.
Scheduling and clinical workflows are a different story. PartnerMD acquired a practice on the same EHR about a year ago and still runs two separate instances of athenaOne. The practices are in different states, they schedule differently, and their workflows are slightly different. Merging them would buy cross-coverage the two groups do not need, so the tablespaces stay apart.
Where the automation line is drawn
Ask Jack where automation gives his team time back and he starts with where it does not. Self-scheduling looks like table stakes for a medical practice, but PartnerMD promises same-day or next-day acute appointments with no waiting, and delivering on that is more art than science, constant shuffling, blocking, and unblocking that a template with open slots cannot capture.
AI chatbots on the phone or in chat are off the list for the same reason.
“Our patients expect to talk to us. We are a high-touch business, and the live person on the other end of the phone is really important for us.”
Where automation is welcome is the asynchronous, back-office layer: texting for referral coordination and membership billing, with room for AI to route messages or answer basic questions inside those flows. The phone system does real work too, rolling calls between offices so a Greenville, South Carolina patient sees a local number on the callback, triage lines for urgent calls, and after-hours routing straight to a physician’s cell phone. No call centers. A five-doctor office runs with one person up front because coverage flexes across locations instead of adding a third of an FTE everywhere.
Real-time feedback and a 93 NPS
Ten years ago PartnerMD collected feedback on paper forms, like everyone else, and the volume was too small to mean anything. Seven or eight years ago they partnered with a startup, sitting across the conference table from the founders, to build a survey a patient could answer in ten seconds or ten minutes.
The requirement that mattered was visibility. Any score that is not a five out of five, and any response that is not a promoter, automatically goes to the practice manager, to the head of operations, to Jack, and to the CEO.
“If there’s something that’s not right, we’re reaching out to the patient. Even if it was just a nit. That patient feels heard, and if it’s something we can change, or it’s a training opportunity with our team, we act on it.”
Five-point ratings all land around 4.9, which is why Jack prefers NPS: it is harsh enough to flag the office sitting at 91. PartnerMD finished its last quarter at 93.
The vendor story is its own lesson. When two or three tools can solve a problem, PartnerMD picks the company it can grow with and influence, a partner, not a product.
Where athenaOne still falls short
Jack has been on athenaOne for close to fifteen years and calls it the right decision then and now. He is enthusiastic about what the platform is doing for physicians: native ambient documentation, summarization inside the chart, and broader access to clinical data, all aimed at giving doctors time back with patients rather than time in the EMR.
His one criticism is the same drum he was beating at strategic forums a decade ago: the patient portal. It has improved, but not to the level a concierge practice needs, which forces third-party tools, two logins, and split data. Consumer technology and wearables are moving far faster than the patient-facing side of the EHR, and the opportunity to combine wearable data with clinical data is, in his words, huge.
The metric he checks every morning
Lightning round. Most overrated tool in healthcare tech? AI as the cure for everything, powerful when applied thoughtfully, a problem when you expect it to do all the work.
The number he checks first each day: terminations. Members leaving the practice. In a month-to-month relationship, retention is the clearest signal of whether the value delivered matches the price charged.
And the sentence he finished, “in concierge medicine you can automate almost anything except”, landed on patient experience, with a story from a member at his gym.
“Jack, you guys want me to use the portal. I pay you guys a lot of money to talk to people.”
That is the whole episode in one line. Build the machinery so the paddling disappears under the surface, and never let the tool get between the patient and the person they came to see.