The Auction You Don't Know You're In

You are in an auction right now, and you have never seen the paddle. Every prospective patient in your city is quietly sold to whichever practice can afford to bid the most to reach them. Not the best doctor. The highest bidder.

I spent 20 years as a software engineer before I built a functional medicine agency. The last product I architected went from zero to $500 million in revenue in four years. That work teaches you one thing above all: the market does not reward the best solution. It rewards the one that can afford to buy attention and still make money. Medicine is not exempt from this. It just pretends to be.

The auction is real and you are already bidding

Here is the mechanism. When someone in Denver searches for help with their gut, their fatigue, their hormones, an ad slot gets sold in milliseconds. Google runs the auction. The practice willing to pay the most for that click usually wins the top spot. Same story on Facebook. Same story, slower, with SEO and referrals, where the currency is time and money instead of dollars per click.

Your bid is not a number you set in a dashboard. Your bid is set by math you probably have never done. It is the most you can afford to spend to acquire one patient and still come out ahead. Bid below your ceiling and you lose slots to people who bid closer to theirs. Bid above it and you go broke buying patients.

Most doctors have no idea what their ceiling is. So they either underbid and stay invisible, or they guess, panic at the invoice, and quit. Both are the same mistake: playing an auction without knowing your number.

Your number has a name: allowable acquisition cost

Here is the whole thing in one line. What a patient is worth to you over their life, minus what it costs you to serve them, sets the most you can pay to get them.

I call that ceiling your allowable acquisition cost. Two inputs feed it.

• Lifetime value. The total gross revenue one patient brings you across the whole relationship.
• Cost to serve. Everything it costs to deliver that care: your time, staff time, labs, software, the front-desk hours spent chasing paperwork.

Subtract the second from the first and you get the gross margin one patient produces. Some fraction of that margin is what you are allowed to spend acquiring them. That fraction is your call. But you cannot choose it until you can see the number.

Now the part nobody tells you. Cost to serve is not fixed. When you automate the repetitive work, intake, scheduling, follow-up, the boring stuff, cost to serve drops. When cost to serve drops, margin rises. When margin rises, your allowable acquisition cost rises with it. Your paddle gets longer.

The practices already using AI to run their back office have quietly raised their maximum bid. They are outbidding you for the same patient, and you never saw it happen.

The worksheet: compute your own ceiling this week

No consultant needed. Pull your own numbers. Here is the model with a worked example. Every figure below is an assumption you should replace with your real data. None of these are measured facts. They are placeholders to show the shape.

Step 1. Average revenue per visit. Assume $250 per visit.

Step 2. Visits per patient per year. Assume 6.

Step 3. Years the average patient stays. Assume 2.

Step 4. Lifetime value. Multiply the three: 250 x 6 x 2 = $3,000 in gross revenue per patient.

Step 5. Cost to serve, as a percent of revenue. Add up provider time, staff time, labs, software, admin. Assume it eats 50 percent. Gross margin per patient: 3,000 x 0.50 = $1,500.

Step 6. Pick the share of margin you will reinvest to grow. Assume you will spend one third of margin to acquire a patient. Allowable acquisition cost: 1,500 x 0.33 = $500 per patient.

That $500 is your paddle. You can spend up to $500 to land one new patient and still keep two thirds of the margin. If your ads cost $80 a click and it takes six clicks to book one patient, that is $480. You are inside your ceiling. You can bid. Your competitor who never ran this math is guessing, and guessing loses.

Now watch what automation does.

Step 7. Cut cost to serve with automation. Assume you automate intake, scheduling, and follow-up, and cost to serve drops from 50 percent to 35 percent. New gross margin: 3,000 x 0.65 = $1,950. New allowable acquisition cost at the same one third share: 1,950 x 0.33 = $650 per patient.

You did not raise prices. You did not see more patients. You cut the boring cost, and your maximum bid went from $500 to $650. That is a 30 percent longer paddle in the same auction, against the same competitors.

This is Decision Leverage: the highest-leverage move is not working harder inside the visit. It is changing the number that decides whether you can afford to be seen at all.

What raising the ceiling looks like in practice

I have watched this play out. One doctor I work with, Dr. Piper Gibson, was nearly impossible to find online. We built custom tools for her, calculators, genetics resources, and she went from invisible to ranking. That is Visibility Leverage: once you can afford to bid, you have to actually show up in the slots you win.

Another, Dr. Diane Mueller, hired new doctors and built a waiting list as new-patient calls climbed roughly tenfold. A practice does not absorb that kind of demand by working longer hours. It absorbs it because the cost of serving each patient came down enough to make the growth pay. The margin funded the machine.

The pattern is always the same. Lower the cost to serve. The ceiling rises. You can afford visibility you could not afford before. More patients arrive. You serve them at the lower cost. The margin compounds. That is the whole flywheel, and it starts with one number.

Do this before Friday

You do not need me for any of this. Run it yourself.

1. Pull your real average revenue per visit, visits per year, and years retained. Multiply them for lifetime value.
2. Add up your true cost to serve and subtract it. That is your margin per patient.
3. Decide what share of that margin you will spend to grow. Multiply. That is your allowable acquisition cost, your bid ceiling, today.
4. List every repetitive task your staff does by hand. Intake, reminders, follow-up, billing questions. Pick one to automate this quarter and re-run the math with a lower cost to serve.

The number you get in step 3 is what you are allowed to bid in an auction you are already competing in whether you calculated it or not. The number you get in step 4 is where it goes once you stop paying humans to do what software does for free.

Run the model. Replace every assumption with your real figures. If you want a second set of eyes on your number once you have it, you can work with me.