The healthcare sales map was wrong. So we drew a new one.

Pick any surgeon in your CRM. Where does she operate?

Not what her email says. Where she actually operates. Which building. Which floor.

If you cannot answer that in ten seconds, you are selling into healthcare with a broken map. Most of us are. I did it for years before I noticed.

The wrong desk

Hitesh Vyas, Sagar Pandey and I worked in the GTM team at Innovaccer. Innovaccer sells a data platform to health systems, payers and life sciences companies. It unifies patient records across a system's hospitals and affiliated clinics, and six of the ten largest US health systems use it.¹

That product has the most frustrating problem in Health IT baked in. The deal is only as big as the network you can map. A health system is not one building. It is a hospital, twelve satellite clinics, forty affiliated practices and an ACO. Miss a branch and you size the deal wrong and pitch the wrong sponsor.

The reps were good. The messaging was good. A lot of outbound still hit the wrong desk. A nurse manager filed under a health system 300 miles away. A surgeon who operates at three sites, showing up at one, never the one that buys. A CIO with no facility, just a parent.

Korzybski was talking about language. He could have been talking about a ZoomInfo export.

Every B2B data vendor faces the same two buttons. Link her to the health system's email domain, or find out which clinic she actually works at. Guess which one is cheaper.

Why every healthcare sales map is drawn the same wrong way

Most B2B databases attach a person to a company by their email domain. It is cheap, it scales, and for software companies it is roughly true. One company, one domain, one org chart.

Healthcare is the one industry where that assumption fails on contact. A health system can run forty facilities on one domain. An affiliated practice can have its own domain and still sit inside the system. A physician can work at three sites and be employed by a fourth.

The incentive is to have the biggest database. Nobody gets paid for knowing which clinic buys. So nobody finds out.

The healthcare data vendors have the other half of the picture. They know the buildings well. They know a lot less about the people inside them, and none of them help you reach anyone.

Getting punched in the mouth

In February 2024, Hitesh Vyas, Sagar Pandey, Sushant Mishra and I started Dmand AI, then called Launcheazy. We did not start with a product. We ran GTM for seed and Series A healthcare companies. Inbound, outbound, the works.

We got our clients meetings at Providence, Ascension, Mayo Clinic and a long list of others. So the outreach worked. The data made us earn every one of them. Every week, the same wall. Not messaging. Not channels. Data.

Our stack looked like everyone's stack. One tool for facilities. Another for contacts. A third for sending. A spreadsheet to keep the three from lying to each other. There are four ways out of that:

1. Buy a facility database.
2. Add a contact database.
3. Add a sending tool and a spreadsheet to reconcile all three.
4. Map the entire US healthcare universe yourself.

We are number four. Not a healthy decision. The right one.

At some point you stop calling it a wall and start calling it the job. While serving our clients, we ended up building our own map of US healthcare just to do that job. This year we turned the map into a self-serve platform, Dmand AI. Sellers and founders at Health IT and services, MedTech, pharma, medical staffing and consulting companies now use it.

Health systems have org forests, not org trees

Here is the thing nobody tells you before your first healthcare deal. Health systems do not have org trees. They have org forests.

An IDN or health system sits at the top. Under it, divisions. Under those, operating groups. Then regional hospitals, then hospitals, then the outpatient departments, satellites and clinics where care actually happens. Somewhere in the branches are joint ventures, management companies and practices acquired last quarter that have not updated their website. Every layer has its own budget, its own procurement and its own idea of who signs.

Borges wrote a one-paragraph story about cartographers who made a map so detailed it was the size of the empire, and therefore useless. Healthcare is the one place where the joke stops being a joke. If your map does not go down to the clinic, it does not tell you who buys.

Both ends of the bell curve land on the same answer: just find out where they actually work. Only the middle builds a probabilistic affiliation model from email domains, LinkedIn headlines and vibes.

So we drew a new map of US healthcare

Dmand AI is a sales intelligence and outbound platform on top of the complete US healthcare universe. Not just the big systems. All of it, in one parent-child tree. Where it stands today:

• 997K+ sites of care, from IDNs and hospitals down to outpatient clinics, physician groups, ASCs, FQHCs, labs and pharmacies.
• 7.4M+ HCPs, from physicians and nurses to every licensed cohort, each tied to the site where they actually work, not to an email domain.
• 11M+ healthcare executives, affiliated at the site level, so the CIO shows up under the hospital she runs, not the parent.
• Waterfall enrichment for emails, direct dials and LinkedIn, checked across several sources in order.
• One workflow on top: 360-degree profiles, network discovery, AI research columns, multi-channel sequences and a unified inbox.

You build the list, enrich it, find a warm way in and run the sequence. One tab.

Meanwhile, every sales team keeps glancing at the new outreach tool with AI in the name, while the data under the list waits.

The map is still wrong. Just less wrong.

The hierarchy is complete from the top to the last child: IDN or health system, divisions, operating groups, regional hospitals, hospitals, then outpatient departments, satellites and clinics. What keeps moving is everything around it. Practices change owners. Physicians change sites. Executives change titles. We refresh affiliations continuously, we have corrected hierarchies we got wrong, and we will keep doing both in public. When our coverage numbers move, you will see the math, spreadsheet included.

That is why we started writing. The seller's side of healthcare is badly documented. There is no shortage of advice on cold emails. Most of it arrives as a cold email. There is almost nothing on how an IDN actually decides, why a physician group changing owners changes your buying center, or which titles at a 25-bed critical access hospital sign the contract.

We sit inside this data all day, so we started a newsletter called Selling Into Healthcare. How healthcare buys. Data notes. Build notes. Field notes from real outbound to providers.

If you sell into US healthcare, subscribe below. And if you have a list that is not working, tell me on LinkedIn where it is wrong. Wrong site, wrong system, wrong person. I will tell you why.

Supreet, Co-founder and CPO, Dmand AI