Fixing the $300 Billion Medical Billing Problem

1. 0:00 – 0:26

   What Candid Health Does

   1. GAGustaf Alströmer0:00

      [upbeat music] Today I'm here with Nick Perry. He's the co-founder and CEO of Candid Health. Um, Nick, welcome to YC.

   2. NPNick Perry0:10

      Thanks for having me.

   3. GAGustaf Alströmer0:10

      What does Candid Health do?

   4. NPNick Perry0:12

      We automate RCM for healthcare providers. So we've rebuilt the entire medical billing stack that healthcare providers use to bill insurance from the ground up to be automation and AI first.

   5. GAGustaf Alströmer0:23

      What's the problem that Candid is trying to solve?

2. 0:26 – 2:13

   Why Medical Billing Is So Complicated

   1. NPNick Perry0:26

      It is extremely difficult if you're a healthcare provider to get paid for providing the services that you provide, whether that's from patients or insurance. And the, the core challenge is if you're one provider, you may be seeing patients from hundreds or up to a thousand insurance companies, and all of those insurance companies have different rules for how to submit claims.

   2. GAGustaf Alströmer0:43

      Mm-hmm.

   3. NPNick Perry0:44

      So you as a provider need to constantly be on top of the latest rules for how to submit claims to each payer. So if you, you are a therapist and you're seeing, just doing the same therapy visit, uh, throughout the day, if all those patients have different insurance, the actual literal claim you submit for that same service to each payer will look different because it has to comply with all of their rules. And staying on top of that in an automated way is extremely difficult to do.

   4. GAGustaf Alströmer1:05

      When I go to the doctor, um, and I get a bill, what's on that bill? Who actually pays? Like I, I know this is something that sounds sim- simple, but it's actually a lot behind that.

   5. NPNick Perry1:15

      There are a bunch of ways that that might get paid for, but the simplest way is you as the patient are gonna owe some amount of that. So it's the patient's responsibility.

   6. GAGustaf Alströmer1:23

      Mm-hmm.

   7. NPNick Perry1:23

      That might get... If you, if you literally go into the doctor, they are probably gonna collect a copay from you up front. That's part of what you might owe. Then you're gonna go do the visit. You're gonna leave. Um, all the visit will get typed up into a claim. The claim will be submitted to the insurance company, and then the insurance company decides, applies your benefits against that claim.

   8. GAGustaf Alströmer1:39

      Mm-hmm.

   9. NPNick Perry1:39

      And then they decide do you need to pay more. Maybe you have coinsurance or maybe you have a deductible. Those are kind of the other two main ways. Deductible is some amount you will pay before your benefits kick in, and then a coinsurance is you as a patient may just owe some percentage of a, of a visit. So if it's $100 insurance and you have 20% copay, insurance will pay 80, you'll pay 20. That would all happen after the fact, and the insurance would pay whatever they're going to pay there. So you'll, you'll go in, you'll pay your copay, and you'll find out two to two weeks to six months later if you owe some other component and, uh, your insurance in the meantime would pay whatever they're gonna pay.

   10. GAGustaf Alströmer2:12

       And,

3. 2:13 – 7:27

   How Medical Claims Actually Work

   1. GAGustaf Alströmer2:13

      and technically how does that work? There's something called billing codes. Like we as patients aren't familiar with this, but, but the doctors is. What's, what is that?

   2. NPNick Perry2:20

      Yeah. So there is actually... It's probably easiest to think about it on a piece of paper. There is a physical claim form that most insurance companies will still accept today. The vast majority of claims, 95% are submitted electronically.

   3. GAGustaf Alströmer2:31

      Mm-hmm.

   4. NPNick Perry2:31

      But you can kind of think of it as this form.

   5. GAGustaf Alströmer2:33

      Mm-hmm.

   6. NPNick Perry2:33

      There's a bunch of stuff that goes on that claim. Um, a lot of it has to do with the patient, the provider, where the service was, and the insurance information, kind of-

   7. GAGustaf Alströmer2:40

      Mm-hmm

   8. NPNick Perry2:40

      ... metadata about the visit, um, unrelated to what actually happened in the visit. And then there are some bits about what actually happened in that visit. So what diseases, ICD-10 codes did the, uh, patient present with-

   9. GAGustaf Alströmer2:52

      Mm-hmm

   10. NPNick Perry2:52

       ... and then what services were provided against those-

   11. GAGustaf Alströmer2:54

       Yeah

   12. NPNick Perry2:54

       ... diseases and ICD-10 codes. And so the, you will come in, the insurance will... Uh, when you go to the front desk they'll collect your insurance card.

   13. GAGustaf Alströmer3:00

       Mm-hmm.

   14. NPNick Perry3:01

       They're capturing, and your name, all that. They're capturing a bunch of that other data that's more who's your insurance, who's the patient, date of birth, all that stuff. They n- that all needs to go on the claim. And then when you go into the visit, what comes out of there are the ICD-10 codes, CPT codes in the United States, and those go on the claim, and then that whole claim gets shipped off to insurance.

   15. GAGustaf Alströmer3:16

       Got it. And then the insurance company match that towards like the specific plan that a patient had? Or how does it work?

   16. NPNick Perry3:22

       So yeah. So step one is insurance decides if this is even their patient.

   17. GAGustaf Alströmer3:27

       Hmm.

   18. NPNick Perry3:27

       If the insurance at that point doesn't wanna adjudicate the c- they'll look at it and they'll... That's called a rejection.

   19. GAGustaf Alströmer3:32

       Mm-hmm.

   20. NPNick Perry3:32

       So like we, this isn't our patient is effectively what they're saying. Assuming you got past that, then insurance says, they don't say this, but what they mean is like, "Okay, this is our patient. We're gonna compare, we're gonna apply the benefit, and if there are benefits we'll apply those and pay it out accordingly." The next step is they decide if this is in or out of network. That also goes through some of that stuff. If you have a comma in the wrong place in the address in some cases, or have other typos, the claim will get adjudicated out of network, and that has big consequences for the patient.

   21. GAGustaf Alströmer3:57

       Of course.

   22. NPNick Perry3:57

       The benefit structure for in and out of pa- out of network is very different. Out of network, we're in an era of high deductible out of network plans, so the patient ends up paying a lot more cash out of pocket. So if it's in network they will pay it out according to those benefits. If there are benefits, they may decide this service has no benefit. We don't pay for this thing. And it gets really complicated because there are, this is actually where the complexity is on, on billing as a provider or even-

   23. GAGustaf Alströmer4:19

       Yeah

   24. NPNick Perry4:19

       ... as a patient dealing with it. Every insurance company has different rules and different plans. And so-

   25. GAGustaf Alströmer4:25

       Of course

   26. NPNick Perry4:25

       ... they all have, they cover different things. There's no standardization of what they're required to cover or there's lim- there is some but not, it's not universal, and it doesn't cover-

   27. GAGustaf Alströmer4:32

       Mm-hmm

   28. NPNick Perry4:32

       ... it's not comprehensive. It doesn't spill out for everything. On the employer side, aside from Medicare and Medicaid and a couple of others, you get your insurance through your employer. Um, there's a thing called a self-funded versus a fully funded plan. Most big companies have self-funded, meaning they decide their insurance benefits. It's their insurance, all that stuff. But that means they, they decide what gets covered and the benefits too. So now the insurance company has this complicated job of deciding who's the benefit structure. They're managing all these benefits of across all these different plans of is it a fully funded, is it Aetna's insurance plan, and like they decide what it is? Or is it this employer's one, and what does that employer have for their benefit? And so you get, they're trying to adjudicate it across all these different benefits, and then that all spits out from this giant gonculator-

   29. GAGustaf Alströmer5:12

       Yeah

   30. NPNick Perry5:12

       ... like what is your copay, coinsurance, and deductible, and what does the insurance company pay the doctor?
4. 7:27 – 9:14

   The $300B Medical Billing Market

   1. GAGustaf Alströmer7:27

      The US healthcare system is a huge portion of GDP, and I think we've, uh, did some research in the past to look for, like, the percent of that that goes to admin is a really big piece of it, and, like, it's the big- the biggest spend actually in this, this country. How do Candid Health attack this, and, and where are you entering this huge, this big percentage of GDP spend basically?

   2. NPNick Perry7:46

      Yeah. It depends on who's doing the math, but the market's somewhere between $200 and $300 billion. Almost all the estimates go somewhere in there for medical billing.

   3. GAGustaf Alströmer7:54

      Mm-hmm.

   4. NPNick Perry7:54

      We can go into later why it is so big and complicated and, and stuff like that. But our approach was, so if you, if you look at that market size, you... About 50% of it is inpatient, so health systems, hospitals, stuff like that, and ha- 50, the other 50% is outpatient, ambulatory, primary care, mental, stuff like that, not where you're not going to a hospital basically or a health system.

   5. GAGustaf Alströmer8:14

      Right.

   6. NPNick Perry8:14

      The health system side is dominated by two or three players. On the billing system side, it's Epic, Cerner, and Meditech, and maybe some others. On the outpatient side, there are hundreds of EMRs and billing companies. It's way more fragmented.

   7. GAGustaf Alströmer8:25

      Mm-hmm.

   8. NPNick Perry8:25

      Largely because there are so many different medical specialties on that side, and health systems are, kind of do many things all at once, and they kind of think of it that way. On the ambulatory side, there's EMRs, there's some generalists, and they work across many specialties, and then there are specialists that kind of work in their specialty, and they kind of dominate that specialty.

   9. GAGustaf Alströmer8:41

      Right.

   10. NPNick Perry8:42

       The way we thought about it was, I mean, eventually we want to work with everyone, but we have a having problem. They all have these crazy moats and feature breadth.

   11. GAGustaf Alströmer8:49

       Mm-hmm.

   12. NPNick Perry8:50

       And this is w- before agents, when we started the company.

   13. GAGustaf Alströmer8:52

       Yeah, yeah.

   14. NPNick Perry8:52

       So it's like software, like features were a moat in that if you had, if you support enough check-the-box type things, it's hard to go replace an incumbent.

   15. GAGustaf Alströmer8:59

       Right, right.

   16. NPNick Perry9:00

       Which is why there hasn't really been a big new billing company in 30 years since all these companies were built. The vector- the go-to-market we used, which I think was the only one you could have used in the last 25 years-

   17. GAGustaf Alströmer9:08

       Mm-hmm

   18. NPNick Perry9:08

       ... was the digital health wave.

   19. GAGustaf Alströmer9:09

       Mm-hmm.

   20. NPNick Perry9:10

       So we started, we were in the last batch-

   21. GAGustaf Alströmer9:11

       Right, right, right

   22. NPNick Perry9:11

       ... right before COVID.

   23. GAGustaf Alströmer9:12

       Yeah.

   24. NPNick Perry9:12

       COVID was happening right in there.

5. 9:14 – 11:00

   Finding Candid’s Wedge

   1. GAGustaf Alströmer9:14

      It was the inflection point of basically a ton of like online-

   2. NPNick Perry9:16

      That's when digital health and telemedicine-

   3. GAGustaf Alströmer9:18

      Yeah, yeah

   4. NPNick Perry9:18

      ... all these law, all... Before, telemedicine was very rare. No one really wanted to cover it on the insurance end. All of a sudden, government, there were laws that you have to cover-

   5. GAGustaf Alströmer9:24

      Right

   6. NPNick Perry9:24

      ... telemedicine and all that stuff.

   7. GAGustaf Alströmer9:25

      Right, right.

   8. NPNick Perry9:25

      So we had this birth of this very fast-growing digital health wave that we basically rode.

   9. GAGustaf Alströmer9:29

      Yeah, yeah.

   10. NPNick Perry9:29

       Um, because those digital health companies were looking for a Stripe for medical billing. They were looking for-

   11. GAGustaf Alströmer9:34

       And they existed when you got started, but they weren't... Were they already by then, when you guys started, your target market?

   12. NPNick Perry9:40

       Yes.

   13. GAGustaf Alströmer9:40

       Or did they become that as a result of COVID?

   14. NPNick Perry9:41

       They were the target market.

   15. GAGustaf Alströmer9:43

       Yeah.

   16. NPNick Perry9:43

       We kind of joked about it being like kind of paid R&D. It wasn't a huge market at the time.

   17. GAGustaf Alströmer9:47

       Mm-hmm.

   18. NPNick Perry9:47

       Um, they're awesome customers. They're doing amazing work, but it's not, it can't be the only market you end up in. So we're kind of like, okay, we're gonna start here-

   19. GAGustaf Alströmer9:53

       Mm

   20. NPNick Perry9:53

       ... build the core of what we need from the product, and then we're gonna expand from there as we build more and more features and get into the a- outpatient, ambulatory, the rest of the market. But we needed time to catch up on our features.

   21. GAGustaf Alströmer10:03

       Yeah.

   22. NPNick Perry10:03

       Then digital health just exploded-

   23. GAGustaf Alströmer10:05

       Yeah

   24. NPNick Perry10:05

       ... super quickly over, with all the telehealth stuff going on. And that let us really... We, we obviously benefited a lot from that, but that really let us focus on the core of billing, which is that first interaction with billing the insurance company, because everything else stems from there. You might argue that, like, the collection of all the data up front is like the start of it, and that is certainly true, but the core is how do you submit, and this has been Candid ethos to begin with, is how do you submit claims correctly the first time? How do you reverse engineer the rules of all these payers, pu- pull them into the platform so that you know how to submit claims correctly the first time?

   25. GAGustaf Alströmer10:38

       Right.

   26. NPNick Perry10:39

       The, the reason we were able to focus on that was we didn't need to build a bunch of the other stuff that people were using. Like, we didn't need to build scheduling. We didn't need to build patient billing-

   27. GAGustaf Alströmer10:46

       Yeah

   28. NPNick Perry10:46

       ... 'cause a lot of people were using Stripe already or something else that we could just plug in, we could-

   29. GAGustaf Alströmer10:50

       Right

   30. NPNick Perry10:50

       ... plug into a lot of these other things that let us just focus and pare down on the product side to, to the part we really wanted to focus on.
6. 11:00 – 13:00

   Winning the First Customer

   1. GAGustaf Alströmer11:00

      Why did they trust you? Or why'd the first customer trust you? 'Cause I'm assuming this is something that, I mean, all of the revenue basically.

   2. NPNick Perry11:05

      Yeah. It is a very important system to them.

   3. GAGustaf Alströmer11:07

      Yeah.

   4. NPNick Perry11:07

      I think they trusted us, one, I think, um, I had a lot of credibility coming out of Palantir.

   5. GAGustaf Alströmer11:12

      Mm-hmm.

   6. NPNick Perry11:13

      Two, I think this is such a good example of the... I actually didn't know when I first... I remember having this learning moment when we were getting started, the, the rest of the hair on fire problem. You're looking for a hair on fire problem.

   7. GAGustaf Alströmer11:22

      Yeah.

   8. NPNick Perry11:22

      I didn't realize the rest of it was they will take a brick to put the fire out. It, RCM very much felt like that kind of problem, where it was so hard. Our first customer, we were... I was meeting with one of the founders. We were in a cafe, and we were kind of talking about, I, I was telling her about, "Hey, I want to work on RCM. I think it's really hard for these reasons. How's it going for you?" And, and the, the line I remember that was really memorable, and this happened eight years ago-

   9. GAGustaf Alströmer11:43

      Right

   10. NPNick Perry11:43

       ... and I, I remember it very vividly, was she said, "I will let you work on any part of RCM because I hate this so much, and it is so hard."

   11. GAGustaf Alströmer11:50

       Is it fair to say that, um, there's a misalignment between the insurance companies and the, and, and the, um, providers, and they're taking b- like, the insurance companies are taking, taking advantage of that, of that situation?

   12. NPNick Perry12:00

       I do think that is true.

   13. GAGustaf Alströmer12:01

       Yeah.

   14. NPNick Perry12:02

       I would not... It has not been my experience where insurance is actively trying to like do deci- dec- deceptive things. Like I, I heard all these anecdotes when we first got started of people saying they worked at insurance companies, like, yeah, Tuesday afternoon, the boss would like deny all the claims.

   15. GAGustaf Alströmer12:19

       Yeah.

   16. NPNick Perry12:19

       Like I've never actually seen that, having done this for seven years.

   17. GAGustaf Alströmer12:21

       Yeah.

   18. NPNick Perry12:21

       So I, I don't think that is going on. But what I do think is going on is they do have very different incentives.

   19. GAGustaf Alströmer12:25

       Mm-hmm.

   20. NPNick Perry12:25

       Insurance is incentivized probably to be careful about what they're paying out-

   21. GAGustaf Alströmer12:29

       Yep

   22. NPNick Perry12:30

       ... and how much they're paying out, and providers are incentivized to, they want to get paid for what they're doing.

   23. GAGustaf Alströmer12:36

       Mm-hmm.

   24. NPNick Perry12:37

       You as a provider have to bill a thousand insurance companies, and all those insurance companies can change their rules whenever they want.

   25. GAGustaf Alströmer12:42

       Right.

   26. NPNick Perry12:42

       And so they're changing things, and so the, the hard part is it's not reverse engineering how to do it once. It's not a set it and forget it problem. It's a how do you actually keep up with all of this-

   27. GAGustaf Alströmer12:51

       Right

   28. NPNick Perry12:51

       ... change and evolution of these rules, and as chain payers change things and what they want and introduce new steps into the process, how do you constantly build a system that can keep up with that? And it's that

7. 13:00 – 15:33

   Doing Things That Don’t Scale

   1. NPNick Perry13:00

      feedback loop.

   2. GAGustaf Alströmer13:00

      Got it. Let's go back to that first customer, the, the, um, online health ca- um, provider. What, what was that like? Like, how did you onboard them? What was sort of the first product like? I know you were doing a lot of things manually in the beginning.

   3. NPNick Perry13:13

      Yeah. At that point, there was still a lot to build.

   4. GAGustaf Alströmer13:16

      Yeah.

   5. NPNick Perry13:16

      Our very first product that we shipped to them was actually literally just an API. It was-

   6. GAGustaf Alströmer13:20

      That did what?

   7. NPNick Perry13:21

      So it was an API that they would submit the data we needed to submit claims.

   8. GAGustaf Alströmer13:24

      Mm-hmm.

   9. NPNick Perry13:24

      And then honestly, at that point, it stopped on the very first day. It was like we, we, we would then manually submit the claims 'cause we were building out the rest of it, and we would manually look over and make sure everything looked good. We would actually, like, literally look at every electronic claim before we submit it. I literally did this for days, and I would just like, make sure they're all correct, then we submit them, and then we didn't have reporting built, and so, uh, I would build out all the reporting in Google, Google Sheets. In our weekly meeting, we'd-

   10. GAGustaf Alströmer13:48

       Mm-hmm

   11. NPNick Perry13:48

       ... review all the reporting of, like, what's working, what's not. And over time, we were building up more and more features to do more things.

   12. GAGustaf Alströmer13:54

       Mm-hmm.

   13. NPNick Perry13:54

       But on day one, it was literally just an API for data capture, and then we were making it work and building more automation, making... On, on day one, I was probably reviewing every claim manually, and then we went to submitting 10 claims at a time and then kind of going to-

   14. GAGustaf Alströmer14:07

       Yeah

   15. NPNick Perry14:07

       ... maybe 100 claims at a time, like, going up like that.

   16. GAGustaf Alströmer14:09

       And then did that get out of control some? Uh, this happens to YC companies all the time when they do things manually. Did it grow so fast that you had to automate things, or where you had to hire a lot of people to manually support them? Or how did you solve-

   17. NPNick Perry14:19

       I think we got crafty with our internal tooling and-

   18. GAGustaf Alströmer14:22

       Yeah

   19. NPNick Perry14:22

       ... figured out how to do a lot of this stuff at scale. And I, I think I did all this. So I was, I was Candid's biller for a year, and a year and a year and a half, and I was figuring out how to do this. And I have some friends that teased me, uh, on, on, uh, how much I was doing this and, and maybe, uh, we did it too long that way. But I learned a lot, and it helped, I think, inform our product on all these different edge cases of what can go wrong.

   20. GAGustaf Alströmer14:43

       Mm-hmm.

   21. NPNick Perry14:43

       R-RCM is not this, like, power law where a small number of things are driving most of the pain. It is a extremely long tail-

   22. GAGustaf Alströmer14:51

       Mm-hmm

   23. NPNick Perry14:51

       ... of, it's a death by a thousand paper cuts problem, and so you have to really get this exposure and, like, call the insurance companies to figure out what's actually going on. I was doing that, and at some point, as we were starting to scale, as we were bringing other customers, we had to figure out how to do this at more scale, so we brought on some more people to do it that, like that. We, uh, and then our first push, I can't, this might've been in the first year, where we actually built our system to actually automatically submit the claims as they come in.

   24. GAGustaf Alströmer15:14

       Mm-hmm.

   25. NPNick Perry15:14

       It was no longer, like, us carefully curating and reviewing everything and making sure they all looked good.

   26. GAGustaf Alströmer15:19

       Yeah.

   27. NPNick Perry15:19

       We cared a lot about being correct and delighting customers.

   28. GAGustaf Alströmer15:22

       Yeah.

   29. NPNick Perry15:22

       We still do, but at that point, we didn't have enough software that we built to, to handle that, so we were just very carefully going through that ourselves manually to make sure everything was really, really, really good and correct.

8. 15:33 – 17:55

   Automating RCM With Software and AI

   1. GAGustaf Alströmer15:33

      Is it fair to say that there was, like, a pre-AI version of, of Candid and then, like, something happened? Uh, w- actually, the, assuming a bunch of things here, can the models h-handle this stuff?

   2. NPNick Perry15:42

      I would say our evolution has been, first, we were a company where it was, we just sold the outcome.

   3. GAGustaf Alströmer15:49

      Yeah.

   4. NPNick Perry15:50

      We called it medical billing, but it's like, "Send the claims, and we will take care of everything else. We handle all the services," everything like that. That was all you could buy. You couldn't buy just the software-

   5. GAGustaf Alströmer15:56

      Right. Right, right

   6. NPNick Perry15:56

      ... 'cause there was no software to buy.

   7. GAGustaf Alströmer15:57

      Yeah.

   8. NPNick Perry15:57

      That was the lily pad number one. Then the next bigger lily pad was, uh, with another customer later. We, we worked with them to co-build our actual software platform that you could just-

   9. GAGustaf Alströmer16:07

      Yeah

   10. NPNick Perry16:07

       ... buy the software.

   11. GAGustaf Alströmer16:08

       Right. Right.

   12. NPNick Perry16:08

       Uh, and that was mostly contained to just the billing set. And then the last one, we did one more lily pad where we kinda finished the extension on both sides, built a patient registration, co-pay collection up, really a front thing for around a in-office visit. Those were three major evolutions of the, uh, the pro-product, and those all happened before agents and AI, LLMs got big.

   13. GAGustaf Alströmer16:27

       Yeah.

   14. NPNick Perry16:28

       The last one has been the introduction of AI and LLMs. It is more of a addition to the platform as opposed to, like, it is, like, totally different, fundamentally different platform.

   15. GAGustaf Alströmer16:37

       Because the platform was already automated pretty much.

   16. NPNick Perry16:39

       Yeah. Our-

   17. GAGustaf Alströmer16:40

       Yeah

   18. NPNick Perry16:40

       ... our vision is to automate RCM.

   19. GAGustaf Alströmer16:42

       Yeah.

   20. NPNick Perry16:42

       And we're not dogmatic about what. We want the best tool for the job, and agents are really good at some things.

   21. GAGustaf Alströmer16:47

       Yeah.

   22. NPNick Perry16:47

       If I have denied claims, and I'm going to appeal these claims by calling payers or going to-

   23. GAGustaf Alströmer16:51

       Mm-hmm

   24. NPNick Perry16:51

       ... paper or whatever, several companies are having agents solve that problem.

   25. GAGustaf Alströmer16:54

       Yeah.

   26. NPNick Perry16:55

       And that is probably better than a person.

   27. GAGustaf Alströmer16:56

       Right.

   28. NPNick Perry16:56

       We have a different philosophy. We wanna go upstream of that and just submit the claim correctly the first time.

   29. GAGustaf Alströmer17:00

       Yeah.

   30. NPNick Perry17:01

       And avoid that denial altogether.
9. 17:55 – 19:25

   From Palantir to Candid Health

   1. NPNick Perry17:55

      them.

   2. GAGustaf Alströmer17:55

      Got it. Tell me about your background. You were at Palantir prior to this, and also Doug's background. Like, how do you guys end up doing this, this idea?

   3. NPNick Perry18:02

      Yeah. Ended up at Palantir, where I was for five years. The team I joined was actually the same team my co-founder was on-

   4. GAGustaf Alströmer18:08

      Mm-hmm

   5. NPNick Perry18:09

      ... and where we met. And the, the, the reason that team was stood up was the COO of Palantir at the time, Shyam, wanted to build a team around solving the, whatever the most pressing growth problem was.

   6. GAGustaf Alströmer18:19

      Right.

   7. NPNick Perry18:19

      And it was a really great learning environment for how to be a really amazing operator and how to do really amazing work. Palantir obviously has a great reputation for-

   8. GAGustaf Alströmer18:25

      Yeah

   9. NPNick Perry18:25

      ... amazing people to learn from and apprentice from, which I, I benefited a lot from. But we were on that team. Uh, and then- Several of us went on to go work in external stuff with customers and lead different parts of customer work. I went onto the commercial side and did mostly healthcare things. I was still mostly interested in healthcare at that point.

   10. GAGustaf Alströmer18:39

       Yeah.

   11. NPNick Perry18:40

       I did pharmaceutical work, and then I did some work with payers. And that was my first exposure to claim adjudication in RCM. I saw it from their side, and we weren't really working on it there.

   12. GAGustaf Alströmer18:49

       Right.

   13. NPNick Perry18:49

       We were working on other stuff, but you, you learn about it just being there. And it was... It's very ... The first thing that's very apparent to me on both sides, it is extremely manual process on both sides.

   14. GAGustaf Alströmer19:00

       Mm-hmm.

   15. NPNick Perry19:00

       And nevertheless, this is a very large and important problem. It's consequential. It was important for so many reasons. And so when we were leaving Palantir and looking for ... Well, I wanted three things. I wanted a big problem, I wanted an important problem. I guess maybe just those two things. I just like, I wanted a big problem, like in a big market, and something I really cared about, was fired up to go work on. This thing matters. RCM checked both those boxes for me, and that-

   16. GAGustaf Alströmer19:23

       Mm.

   17. NPNick Perry19:23

       That's how we got excited about it.

10. 19:25 – 20:44

    The Moment They Hit Product-Market Fit

    1. GAGustaf Alströmer19:25

       I remember a call. You called me at some point and said, "Hey, I'm pretty sure this is now working." Like, like we've seen-

    2. NPNick Perry19:29

       Yeah.

    3. GAGustaf Alströmer19:29

       Like we are now bidding on way bigger contracts than we used to do. And I was like, "Oh, I love, I love hearing when like the founder is like convinced that we have PMF." Like, can you just describe the journey to after like maybe after that first year and second year?

    4. NPNick Perry19:40

       Yeah. We always had really strong traction in terms of growth.

    5. GAGustaf Alströmer19:43

       Mm-hmm.

    6. NPNick Perry19:44

       Um, we grew really fast the first several years.

    7. GAGustaf Alströmer19:46

       Yeah.

    8. NPNick Perry19:47

       I think... But you never really-

    9. GAGustaf Alströmer19:50

       No. Always that gap

    10. NPNick Perry19:50

        ... I don't know if you can ever really know. It's like, okay-

    11. GAGustaf Alströmer19:52

        Yeah

    12. NPNick Perry19:52

        ... well, we did that. Well, I can, I can come up with some reason why it's not gonna work the next few years.

    13. GAGustaf Alströmer19:55

        Yeah.

    14. NPNick Perry19:55

        And so you don't really. But there w- there was a moment where we ran an RFP with a really big private equity-backed MSO, where we beat, uh, one of the legacy incumbent competitors, and they've been ar- there for 25 years. And we won that. Where I was ... I think that was my moment of like, oh wow, this thing is like, uh, we've like crossed that threshold, and I have a lot of... I have conviction at this point that we hit product market fit, of this thing is in demand, and people want it because we just... We're a four-year-old company that just beat a 25-year-old company-

    15. GAGustaf Alströmer20:25

        Right

    16. NPNick Perry20:25

        ... in a really competitive process.

    17. GAGustaf Alströmer20:27

        Got it.

    18. NPNick Perry20:27

        And they were pumped about the thing that we were shooting for and was our North Star, which is automated RCM. Like, that was the thing they want. And we weren't even feature complete for what they wanted. They knew we would have to build some more stuff around them.

    19. GAGustaf Alströmer20:37

        Yeah. Yeah.

    20. NPNick Perry20:37

        And we did. We executed really well on that. And, but the thing that they were really excited about was the actual, like actually automating RCM

11. 20:44 – 23:39

    Building the Team and Culture

    1. NPNick Perry20:44

       for them.

    2. GAGustaf Alströmer20:44

       Okay. Tell me about the team that you built. Like, who joined Candid early on? Who are joining C- Candid right now? Like, what's people's background, and how does it look like? You're 250?

    3. NPNick Perry20:53

       Ish.

    4. GAGustaf Alströmer20:54

       Ish. Yeah.

    5. NPNick Perry20:55

       Yeah. We've hired pretty similar profile people the entire time. I, I guess the way I think about it is two things. Um, the first, I think by far and... This is something we learned from Palantir. I think by far and away, the most important thing about the company is your culture.

    6. GAGustaf Alströmer21:07

       Mm-hmm. Yeah.

    7. NPNick Perry21:08

       If you get the culture right, I think a lot of the rest will just figure itself out.

    8. GAGustaf Alströmer21:11

       Yeah.

    9. NPNick Perry21:11

       And culture is a byproduct of partially people too. So you got to get the right people and culture. So my co-founder and I are deeply involved in hiring basically everyone at Candid still.

    10. GAGustaf Alströmer21:19

        Mm-hmm.

    11. NPNick Perry21:20

        We, everyone at Candid still gets interviewed at some point by me or Doug, and several... A couple of other people that we've like, um, were really amazing interviewers and have that-

    12. GAGustaf Alströmer21:27

        Yeah

    13. NPNick Perry21:27

        ... have that skill and can, can see talent. So we, we care a lot about that and invested a lot of time in doing that kind of thing. We look for several things, um, in people, but we actually don't do functional interviews generally. We, we'll do like some for some roles of like, are you good at Excel, and are you like analytical?

    14. GAGustaf Alströmer21:43

        Mm-hmm.

    15. NPNick Perry21:44

        But we actually, we look at much more of the fundamental stuff. Do you have high horsepower? Are you gritty? Are you introspective? There's a, there's a bunch of things that we, that we look for. Those have been important to us for from day one, and I actually don't think ... We have like one kind of life story interview that's at the end of our process, where it's not even a trick question. We're just asking people about their life.

    16. GAGustaf Alströmer22:01

        Yeah.

    17. NPNick Perry22:02

        Um, and figuring out the, what, what went really well and what didn't go well. That scorecard for that interview has not changed since the first day we started the company. It's like the same attributes that we look for in like everybody.

    18. GAGustaf Alströmer22:12

        What's like on your agenda for the next like two years or three years? Like, like you specifically. Like, you're now a company that's, it's working, it's large, it's an enormous market, and you, um, are probably the best company of, of the newer ones that exist in this market, and you wanna win. Like, what's, what's on your agenda? What are you gonna do?

    19. NPNick Perry22:33

        We are big and fast-growing.

    20. GAGustaf Alströmer22:34

        Yeah.

    21. NPNick Perry22:34

        We are only scratching the surface of this market. This market is so big.

    22. GAGustaf Alströmer22:37

        Yeah.

    23. NPNick Perry22:37

        We could have 1/280 of it and still only have a billion in revenue. Uh, it, it, it's like, it's a really massive market.

    24. GAGustaf Alströmer22:43

        Yeah.

    25. NPNick Perry22:43

        It is extremely fragmented. One priority is just continue to grow and work with more, more players.

    26. GAGustaf Alströmer22:49

        Yeah.

    27. NPNick Perry22:49

        I actually think we're having a really-

    28. GAGustaf Alströmer22:50

        What's the limiting factor?

    29. NPNick Perry22:52

        Probably the size of our sales. It depends. A couple of years ago, I would've said product features.

    30. GAGustaf Alströmer22:58

        Yeah.
12. 23:39 – 27:24

    From the First 10 Customers to Scale

    1. GAGustaf Alströmer23:39

       I'm, I'm curious how the first like 10 customers happened. What, like how did you get them? Like, you mentioned that the first one where you went-

    2. NPNick Perry23:45

       They-

    3. GAGustaf Alströmer23:45

       ... to their office in Hayes Valley.

    4. NPNick Perry23:47

       Yep. They were all, I just went through my network and went through LinkedIn. I came up with like lists of companies, and I went on LinkedIn, and I found who their founders were, and then looked to find connections, and I emailed my friends for connections-

    5. GAGustaf Alströmer23:59

       Yeah

    6. NPNick Perry23:59

       ... my network for connections to these people, and it kind of, it kind of grew from there. At some point, we were doing, as, as I had mentioned, we cared a lot about creating delight and doing a really good job.

    7. GAGustaf Alströmer24:08

       Mm-hmm.

    8. NPNick Perry24:08

       And so the investors of our customers-

    9. GAGustaf Alströmer24:10

       Mm

    10. NPNick Perry24:10

        ... started hearing about us in board meetings and other things, and then they started reaching out and putting us in touch with other, uh, of their portfolio companies 'cause we were doing such a good job-

    11. GAGustaf Alströmer24:18

        Yeah

    12. NPNick Perry24:18

        ... and had such a good brand around this. We still, that still happens all the time today.

    13. GAGustaf Alströmer24:21

        Are, are, is inbound growth still the majority, or are you, are you at the point where like that doesn't scale anymore?

    14. NPNick Perry24:26

        That... Yeah. It's, it's not the only lever.

    15. GAGustaf Alströmer24:28

        Yeah. Yeah.

    16. NPNick Perry24:28

        That's still a healthy amount of-

    17. GAGustaf Alströmer24:29

        Yeah

    18. NPNick Perry24:30

        ... different kinds of referrals that we get.

    19. GAGustaf Alströmer24:31

        Yeah.

    20. NPNick Perry24:32

        But we do a lot more outbound too.

    21. GAGustaf Alströmer24:33

        What percent roughly of healthcare spend is, is, is the thing that you go after?

    22. NPNick Perry24:37

        It is about 5%, more or less. So of all every, of all the money spent on providing healthcare-

    23. GAGustaf Alströmer24:45

        Mm-hmm

    24. NPNick Perry24:45

        ... about 5% of it, if you're looking at the whole zoomed out thing, is, uh, spent on a-

    25. GAGustaf Alströmer24:50

        It sounds high to me. Is that-

    26. NPNick Perry24:52

        It is high

    27. GAGustaf Alströmer24:53

        ... is that a wrong assumption to think it was high?

    28. NPNick Perry24:55

        No. I've heard... In the most extreme cases, I've heard 15 to 20%, depending on what you're getting and what you want.

    29. GAGustaf Alströmer25:00

        Yeah.

    30. NPNick Perry25:00

        But I, I think most people are... There, there's a concept in medical billing called cost to collect.
13. 27:24 – 30:14

    Where AI Agents Matter Most

    1. GAGustaf Alströmer27:24

       Got it. We talked about a- agents a, uh, a second ago. The introductions of models and agents in, in this world, like, where do you think it'll have the biggest impact?

    2. NPNick Perry27:32

       There isn't a... In RCM, there has not been a Harvey or a Lawvora or a Sierra-

    3. GAGustaf Alströmer27:37

       Right

    4. NPNick Perry27:37

       ... or something that grew extremely fast. There were the Abridges of the world.

    5. GAGustaf Alströmer27:41

       Mm-hmm.

    6. NPNick Perry27:41

       Um, the Scribes.

    7. GAGustaf Alströmer27:42

       Mm-hmm.

    8. NPNick Perry27:42

       Um, those grew very quickly.

    9. GAGustaf Alströmer27:44

       Yeah.

    10. NPNick Perry27:44

        In RCM, on claim adjudication, there's not been one of those. And one of the main reasons is it goes to the data that they all need to thrive and do their thing. In law, stuff changes, but you're, these documents are broadly... It's document based.

    11. GAGustaf Alströmer27:58

        Yeah.

    12. NPNick Perry27:59

        It's broadly available. If something changes, it's public. You can ingest it.

    13. GAGustaf Alströmer28:02

        Yeah. Yeah.

    14. NPNick Perry28:02

        It's not like there's secret stuff going on. In RCM, the rules for how to submit claims, going back to the core problem, there's 1,000 insurance companies, and I'm one provider, and I need to keep up to date with all of their changing rules that they change-

    15. GAGustaf Alströmer28:12

        Yeah

    16. NPNick Perry28:13

        ... every quarter annually. Not all of those are published. And so some of them are published. They'll be published maybe publicly in a PDF, and you can go get those to figure out-

    17. GAGustaf Alströmer28:22

        Mm-hmm

    18. NPNick Perry28:23

        ... "Okay. Well, for this payer, I need to put this."

    19. GAGustaf Alströmer28:25

        Mm-hmm.

    20. NPNick Perry28:25

        "They'll pay for these things and do it this way. They want this modifier, not this other one." Most of it is not.

    21. GAGustaf Alströmer28:30

        Mm-hmm.

    22. NPNick Perry28:31

        And so you have to... The only way you can learn is you're actually submitting claims, and you're calling payers, whether it's a person or an agent, and you're reverse engineering, you're building these rules, uh, and you're, like, learning how to submit claims-

    23. GAGustaf Alströmer28:40

        Right

    24. NPNick Perry28:40

        ... in that, that data set. We've had to... We've been building that for seven years, and that is a really important part of why we're so good at what we do.

    25. GAGustaf Alströmer28:47

        Mm-hmm.

    26. NPNick Perry28:47

        But I think agents are a really important part of that because they can help us even more quickly and more expansively cover all the data exhaust around what we are doing. We have every action in... We have customers that have hundreds of billers doing stuff in Candid.

    27. GAGustaf Alströmer29:01

        Yeah.

    28. NPNick Perry29:01

        Every action is tracked. Most people write a history of, like, what happened on this claim.

    29. GAGustaf Alströmer29:05

        Yeah.

    30. NPNick Perry29:05

        But we have machine learning or agents looking at the data of what's happening. They'll, they'll suggest automation. It'll be like, "Hey-
14. 30:14 – 37:33

    Raising a Series D and What Comes Next

    1. NPNick Perry30:14

       becomes very challenging.

    2. GAGustaf Alströmer30:15

       You just raised a Series D.

    3. NPNick Perry30:16

       Yep.

    4. GAGustaf Alströmer30:17

       Tell us about that.

    5. NPNick Perry30:17

       Yeah. We just raised a Series D. Super excited about it. $120 million from Sixth Street.

    6. GAGustaf Alströmer30:22

       Congrats.

    7. NPNick Perry30:23

       Thank you. Um, they're awesome partners. They know healthcare extremely well.

    8. GAGustaf Alströmer30:27

       Mm-hmm.

    9. NPNick Perry30:27

       And to... In our level, what we're doing, we work with a lot of private equity groups. We work with a lot of private equity roll-ups, a lot of big medical groups. They're just deeply connected to everybody and can help introduce us as we're getting into this more traditional market. Getting our name out there is a really important part of that, and just helping get introductions and connected to people.

    10. GAGustaf Alströmer30:42

        Mm-hmm.

    11. NPNick Perry30:42

        And so they've been, they've been an awesome partner on, on that front so far.

    12. GAGustaf Alströmer30:45

        It's clear to me that you have a sp- Very clear mission on making healthcare better. Can you just describe a little bit, little bit more, like, for the audience? Like, how, how does healthcare become better as a result of all this?

    13. NPNick Perry30:56

        There's one really tangible one-

    14. GAGustaf Alströmer30:57

        Mm-hmm

    15. NPNick Perry30:57

        ... which is basically everyone you and I know has paid for a medical bill they probably didn't need to pay for.

    16. GAGustaf Alströmer31:02

        Mm-hmm.

    17. NPNick Perry31:02

        And it's not that there's some bad guy out there, like some bad provider trying to, like, pull one over you or some insurance company trying to get you. It is so easy to make mistakes. Lit- literally-

    18. GAGustaf Alströmer31:11

        Yeah

    19. NPNick Perry31:11

        ... an insurance company, there is, there is one insurance company I know of that if you put a comma in the wrong place, they will adjudicate your claim out of network. So if you put, uh, 301 Howard, or if you put, um, some address comma suite whatever-

    20. GAGustaf Alströmer31:24

        Mm-hmm

    21. NPNick Perry31:25

        ... as opposed... And their system doesn't have a comma, your c- claim will get adjudicated out of network.

    22. GAGustaf Alströmer31:29

        Yeah.

    23. NPNick Perry31:29

        And so that means it hits your out-of-network deductible, which you're probably gonna pay a lot more. So now a, what might've been a $25 copay is now a $1,000 charge.

    24. GAGustaf Alströmer31:37

        Mm-hmm.

    25. NPNick Perry31:37

        That is really-

    26. GAGustaf Alströmer31:39

        And do patients know that?

    27. NPNick Perry31:40

        They don't know that that's not supposed to happen. They don't... They, they may not even... One, they might even not notice it.

    28. GAGustaf Alströmer31:45

        Mm-hmm.

    29. NPNick Perry31:45

        Two, they... 'Cause you get those explanations of benefits that are actually kind of hard to parse. The thing says-

    30. GAGustaf Alströmer31:49

        Yeah
15. 37:33 – 38:59

    The Startup Lesson Nick Wishes He Knew

    1. NPNick Perry37:33

       on right now.

    2. GAGustaf Alströmer37:33

       Is there any specific lesson that you, um, have learned that you wished you kinda knew when you started the company?

    3. NPNick Perry37:38

       I think a lot of people, when they talk about seed-stage company or early, even pre-seed, they, uh... There's a lot of talk about product-market fit and iterating, talking about product. I wish somebody had pushed me to build a financial model sooner. Um, I don't think-

    4. GAGustaf Alströmer37:51

       Wow.

    5. NPNick Perry37:51

       Uh, at the end of the day, if you squint at each company, it's a spreadsheet. Investors invest in spreadsheets.

    6. GAGustaf Alströmer37:56

       Yeah.

    7. NPNick Perry37:56

       And spreadsheets, they don't, they don't really care what the company does to make their spreadsheet, but, uh, it is a set of s- sort of, they're comparing spreadsheets of which ones they wanted to invest in. And I think having the level of clarity of your business, every business is that spreadsheet, really. Having the clarity around, uh, what you want to do by when and ba- It forces you to back into the planning of like, "Okay, to have this by this point, to have my spreadsheet competitive with the other spreadsheets, I need to have this in place." And I think it's that process of forcing you to draw lines in the sand of your goals and help you work backwards that I think is really valuable. And for whatever reason, I just, none of my friends, other starting companies otherwise were like, "We need to talk about our spreadsheets or build our, our models for this." It was almost looked down on, like, "Why are you-

    8. GAGustaf Alströmer38:39

       Yeah

    9. NPNick Perry38:39

       ... spending time doing that?"

    10. GAGustaf Alströmer38:40

        Yeah. It's-

    11. NPNick Perry38:41

        I found it to be the most clarifying planning tool.

    12. GAGustaf Alströmer38:43

        Nick, thanks for coming back to YC. It's really great.

    13. NPNick Perry38:45

        Thanks for having me.

    14. GAGustaf Alströmer38:46

        Yeah, this was a very interesting conversation.

    15. NPNick Perry38:48

        This was great.

    16. GAGustaf Alströmer38:49

        Yeah.

    17. NPNick Perry38:49

        Thanks.

    18. GAGustaf Alströmer38:49

        Thank you. [outro music]

Episode duration: 38:59