Part of my "job" at Seedtable is surfacing interesting companies you might've not heard of. One of these companies is accuRx, a London-based company working on a platform that brings patients and their healthcare teams together.
Their story is fascinating. After struggling with their initial idea and pivoting, they captured over 90% of the primary care market in the UK. So you probably can't tell from the transcript, but I'm VERY excited about this one. We discussed:
- Healthcare as a communication industry vs. knowledge industry
- The non-obvious impact of COVID-19 in the healthcare industry
- The underlying incentives driving the healthcare industry
- The early years of accuRX, going through Entrepreneur First and the pivot into messaging
- accuRx's unique culture and product development process
Now, on to today's interview...
Tell me a bit about accuRx and what you are building.
Jacob Haddad: We’re a communication platform for healthcare.
We think that healthcare is a communication industry. It wasn’t that way 50 years ago, it was very much a knowledge industry led by individuals, and now it’s led by teams and those teams need to communicate.
So we’re building a platform in which everyone involved in a patient’s care can communicate with each other.
This "healthcare is a communication industry" it’s sort of one of those things that are hidden in plain sight but no one is doing much about it. For instance, the NBA recently realized that, since 3s are 50% worth more than 2s, it’s worth shooting way more 3s. How did you discover that insight?
JH: When we started as a company four years ago we were entirely focused on antibiotics prescribing. There’s a huge problem that we haven’t developed any new antibiotics in the last forty years, we rely on them for all of our medicine and if we overuse them they become less effective. So we started as a company working on that.
About 18 months in, we learned that that just wasn’t going to be a business so we went and just spent 3 months living inside one GP practice, one primary care clinic and built all sorts of prototypes. You know, how they understand their demand, how they do knowledge management, how they collect data from patients, how they message those patients, all of that, and then we did a ton of user research. Shadowing GP appointments, shadowing doctors appointments, shadowing the reception teams, speaking to patients, looking at the practice’s accounts, at how they spent money, all these different things. And we made lots of improvements there. 90% were completely unscalable, but the underlying theme of what we were trying to solve was communication. Healthcare, actually, is really really messy–you know someone has a referral, they need to go meet this other doctor, they have a prescription but they can’t get it until they get this test done, but they already have an appointment for next week so they can get the test then… so there’s all this mess. And humans are really good at handling that mess, they’re very good at problem solving, they just need really lightweight and generic means of communication--and that’s proven true in every other sector, email and Whatsapp and phones.
"The underlying theme of what we were trying to solve was communication."
So we spent a lot of time in the front lines--even after those three months. Even still, everyone in our team has to spend time in GP practices. We put a lot of user researchers in our time, we’ve got clinicians in our team, we have one of our users join our whole team call every single morning to talk about the front lines and how they’re using our software. It’s a really chore part in our DNA, being close to the front lines.
That’s very cool -- you mentioned this started 4 years ago right? And you started on Entrepreneur First. Let’s talk a little bit about the program, how you met your co-founder; was there something you really didn’t expect of those few months you spent there?
JH: I joined EF with a healthcare background, so I knew I wanted to build a business in healthcare. I knew that antibiotics resistance really interested me just as a huge problem facing health systems.
Laurence, my cofounder, comes from the software engineering industry and really wanted to work to improve society, the environment, these things. So we met there. The beginning of Entrepreneur First is sort of like awkward teen dating––people are working with one person, then they’re working with someone else. The EF team is really good at, if you’re not working well together they get you working with someone else. Luckily, Laurence was actually the first person I worked with and we really really got on. And those are the things that people ask you at EF. Are you happy? Are you productive? And since then, we’ve always been very happy and incredibly productive working together so we kept going.
The program was very good at providing weekly accountability. At the very early stages we didn’t have any customers or users we were accountable to, so essentially as our only investors, having them put a small amount of money and having them ask, you know, what are you going to do this week? And then a week later, ask, these are the things you said you were gonna do, did you really do them? That helps drive the motion during the program.
You went to EF, you started working on this antibiotic resistance problem. You figured out that the incentive didn’t match the externality… so you pivoted into what you guys are doing now, this platform for healthcare messaging. Is there one counterintuitive or unexpected insights that makes the whole accuRx thing work? Something that you guys realised that made the whole thing come together?
JH: Oh, there were certainly things we didn’t expect. Our core product today is embarrassingly simple. Doctors pull the patient’s record open, they type a message, they hit send and it goes to the patient’s phone via sms. We discovered that because when we built our decision support system for antibiotic prescribing, that was the very favorite moment for our users. It was like witchcraft. They typed something and it went into their patients record system and to their phones at the same time. So there was definitely something there. And I think what that also really taught us, what the pivot taught us was that we have to create immediate delight and really really minimise the time to value. So what you want to do to take it from 20 Gp practices using us to try and make it a self-service product, what you really have to optimise for is: can someone who has never had access before hit our website and within five minutes message their patients. So minimising that time to value was… it might not sound like that much of an insight, but I think that in healthcare in particular, pretty much all procurement is very very chopped down, so that’s why you end up with these legacy systems that tick all the procurement boxes but aren’t actually useful or useable for patients. So that’s why we had this enormous effort of user research and building something that users can use.
You’re very public about this whole "time-to-value" concept, and even compared accuRx to Microsoft Excel, where you said you build products where you have "fixed use cases" and that the users can just innovate on how they use them.
This reminds me of this product building heuristic of "High ceilings and Low floors". Low floor means it's very easy to start and high ceilings that you can achieve a high level of mastery. Do you think about that framework when you talk about product development? What drives those product decisions we've been talking about?
JH: I never heard it described as high floors and low ceilings. I like that analogy. I think of high-ceilings more in a sense of opportunity. We don’t know what our users can do. We hear anecdotes. The way we think about it day-to-day when we build products is not trying to build rigid workflows, we don’t build a work-flow for making a referral or for reminding a patient about an appointment. We’ve got a workflow for sending a message. We’ll make it easier to do those things… We'll have templates and features that let you do those use cases but we don’t want to ever say this is for one thing and one thing only because that’s when people stop innovating. Some of the most rewarding experiences is when you hear users coming up with super innovative ways to do things.
We had the other day, in the middle of the pandemic, practices who are telling some patients to come in, they speak to patients on the phone and they say come in we want to examine you and put out signs in their car parks that said wait in your car and we’ll send you a text message when we’re ready. And the text message will say come straight into relay. Now, we’d have never a million years thought of that. But by having a really generic product, people have been able to do that. People have been able to do video-calling, send the link for the call to a family member that can interpret for the patient if they don’t speak English, all these sorts of things we hadn't originally accounted for--they are incredibly rewarding.
This is, I guess, incredibly rewarding because you work ridiculously close to your customers. I heard this anecdote that you have this physician’s desk set up in your office with a slow internet connection and a small computer to do user testing, just like your customers would your product. Tell me a bit more why you think that being obsessed with customers is important and how you’re implementing that across the company.
JH: So, in healthcare, there’s a few challenges. One, you’ve got to serve the 100%. If you’re building an e-commerce platform or whatever, you can say well, some users have got some weird setting on their computer so they can’t use our product, but it doesn’t matter because they’re not our target users or whatever. For us, we get delight from obsessing over edge cases like that one user on Internet Explorer 11 who can’t get it to work. It’s a quite different mentality from traditional productivization where we obsess over that long tail of users.
One of the other big challenges in healthcare, you’ve got the clinicians and all the staff who are very very aware of the problems but don’t necessarily know how to go about a solution, and you’ve got technologists who know how to build solutions but don’t see the problem. And so, where that goes really wrong is two ways; one you don’t build any solutions or you have a technologist try to shove solutions down the system with buzzwords like AI and blockchain and all sorts of smoke and mirrors, bells and whistles, and it doesn’t actually help anyone. You only have to go to a doctor's practice and sit in the waiting room and ask people what they’re coming in for and you quite quickly realise that a chapel isn’t really going to help someone if they’re very isolated and lonely and they just really need to talk to someone or things like that. I think it really helps us bridge that divide between what’s going on behind the reality in the front lines, which is like people are still using pagers and sending faxes and sending most documents by post… the reality between that and what we’re doing today… we can never forget that.
One of the other big challenges in healthcare, you’ve got the clinicians and all the staff who are very very aware of the problems but don’t necessarily know how to go about a solution, and you’ve got technologists who know how to build solutions but don’t see the problem.
So we really do have an uncomfortable, clunky old desk in the corner of the office where we like to test software. Even really small things you get out when you test it in that way. You learn things like if somebody sitting there is the patient what are they seeing on the screen at the same time that the doctor is there with the software. Where if you’re trying to have a conversation with the patient and are seeing the software at the same time, how much cognitive load is there to be able to do the action whilst talking to the patient? And that’s quite different to how a lot of UX testing normally happens. The best way, though, is to really just go to GP practices. Obviously, this is not happening at the moment because of COVID but just going out to GP practices and seeing them in their own environments, we do that a lot.
I guess the desk also works as a tangible indicator of what’s important for the team, right? It’s right there in the office.
You mentioned COVID-19, which is in everyone’s mind right now. Let’s dive a bit into that. We discussed briefly healthcare as knowledge industry vs communication industry, how that sort of changed over the last 50 years… but for the patient, their journey has been mostly the same for the past few decades, and technology and innovation had little impact, or I think so at least. The patient they develops a symptom, the patient visits a doctor’s office, the doctor diagnoses largely from outward symptoms and the doctor sends the patient home under a "watch and wait" protocol, and the patient either recovers or that gets escalated. How do you think COVID-19 is changing that patient journey?
JH: I think that the need for more streamlined ways to communicate, everything from asynchronous ways to communicate, not having to be face to face, being able to easily get a specialist's advice, that’s been there for years, even for decades. But the healthcare system is very driven by risk rather than opportunity. So it’s very rare for the system to say, oh, what’s possible here? It’s much more likely to say, what are the ways this could go wrong?
What’s happened with COVID is that that opportunity has become a risk. If we don’t have a way to see patients remotely, or for them not to have to come in to get a document or whatever, then we’re putting ourselves at risk, we’re putting our patients at risk of infection… so we need to breach technological thresholds rapidly.
What has actually been great for us is that we haven’t had to go and build many things specifically for COVID, we just take the things that we were already working on or planning to work on and done them a lot faster. And they’ve been received in an order of magnitude faster than if we would have done them outside of COVID. I think that's sort of one big impact.
The other thing is sort of a nuance. I definitely agree that that model of consultations and things has definitely been around for decades, that’s particularly in keep with presentations of people who have new symptoms. The majority of what primary clinics are dealing with is people with ongoing medical conditions , requiring check ups or where their issues aren’t quite getting resolved, or patients with multiple health problems. That’s only increasing. People live longer and we’re getting better at detecting disease and things like that. The healthcare system hasn’t adapted to find ways of managing that--why technology is particularly important there is… in the UK, GP appointments are ten minutes, in other countries they’re fifteen minutes… but generally you’re talking about ten to fifteen minutes a number of times a year and there is only so much healthcare that can take place in that time.
So I think that one of the paramount shifts that we’re trying to drive is rather than relying on those times, how can we use those times to sort of like, steer and navigate and leverage the healthcare that is going on because healthcare happens every day. The decisions someone makes around their lifetime, the risks they take, how they take their medications, whether they worry about things or not. And so, having real easy ways for patients to quickly ask a question over an asynchronous message, without increasing a lot more the burden on the system, I think those are some of the approaches where we can shift the current model.
The big challenge in shifting that model is that in pretty much every healthcare system around the world, healthcare is at capacity and so there isn’t much slack to try new things or sort of open up new ways of people accessing care. People are just so worried that that will increase workload and it will collapse. But actually when you talk to some innovators who have changed the ways the patients can access them, you found that the network load has decreased, but a lot of people don’t think network, they think about “oh there’s this new thing I’ve got to do now”, but they don’t necessarily see “oh, that saves me from making all these appointments in six months time” or something like that. I think that will need to change.
The other big thing that will need to change is incentives in healthcare. So, here’s somewhere where the UK is actually more advanced. In primary care, family doctors aren’t paid for every appointment, they’re just paid for every patient they look after for the whole year. But if you’ve got a model where you’re paid for every appointment, you’re not incentivised to have more efficient ways of communicating rather than needing an appointment because it will reduce your income. I think that’s the other thing that needs to change. The current model is very tied to that.
Are you thinking about any ways of changing those incentives from inside? Because it’s fairly hard to change entrenched incentives.
JH: Yeah… So, in every country there is a realisation that healthcare systems need to move towards a value-based care or outcome-based care, where rather on being paid by activity, you’re paid by on healthcare outcomes. I’d say that in primary care here we’re in a good position in that in the UK there are good incentives around providing good care rather than activity. In hospitals in the UK there is a big push to reduce hospital outpatient activity, so we’re trying to ride and even lead that transformation. In other countries, it really differs between countries. So what we’re hoping is that we can show a really stark improvement in the healthcare system here and then be able to demonstrate that to other health systems. But systems are starting to modernise. I’ve seen, you know, looking at the German healthcare system is in a place where you can do 20% of your consultations remotely and still get reimbursed in the same way and things like that. So they are starting to update.
You are very intentional about culture and team building: how you hire, how you run the operations, how you buld a company. Let’s talk a bit about that. One of the things that I find very interesting is that you’re fairly against consensus decision-making. Why is that and how do you think about decision-making then?
JH: I’d say we’re more against consensus decision making in the context of, it becomes a massive barrier, asides from the team and the health systems, things just can’t get done until things get agreed and that’s fairly echoed at the team level. So I think that as a team it’s very important to have a very high diversity of opinions but, if you're trying to optimise for using all of them you’ll nowhere; you’ll disappoint everyone and waste a lot of time. You’ve got to reach a decision and then commit to it as a team, whether people are bitter or they run with it, we’re very clear with the team when we try something new or somebody else in the team implements something new that, like, this is an experiment. The only thing we know with complete confidence is that it probably won’t work, but we’re gonna try it out anyway to learn. So, that gives people a lot more confidence to try new things and it also means that if you try something new, we don’t have to keep it working just because we tried to move to this new area of running teams or something like that. We were very clear from the start that it is an experiment. That’s part of how we like to approach things as a team. I’m happy to talk more about team stuff, if you have any particular questions about it…
Well, you just mentioned that having diversity of opinion was crucial. Why do you think building a diverse team is important?
JH: So… there’s like… every company is saying diversity is good and it’s 2020 and we need diversity--
Yeah, that's definitely not where I'm going. I’m going after the actual reason why you value diverse opinions and I’m sure it’s not a D&I thing for you guys.
JH: Yeah, so… In healthcare we’re building for the definition of diversity. We can't get more diverse than the patient population and the fact that is massively skewed toward more vulnerable groups… so that’s both true of the patient population and also the staff. Healthcare workers are incredibly diverse, even in terms of gender. 70% of the workforce are female, 75% of decisions in healthcare are made by women, so if we’re not representing that in our team, the long shot of it is that we’re just gonna make terrible products that aren’t going to solve real problems. There’s another side of it as well, which is , I want to come to work every day and have lunch with interesting people and chat with interesting people before a meeting starts. You’ve got different perspectives and people aren’t the same. I’ve worked in workplaces before where people are the complete opposite. Where most people are close to a carbon copy and you don’t get that same level of ideas but it also means that if you aren’t a part of that carbon copy you feel very out-placed and really good ideas get shut down. We want to have a place where actually we’re bringing the best and most creative working ideas and perspectives onto the product but also to how we build the team.
What processes have you put in place to make sure that you optimise for diversity without compromising on other things?
JH: We had in 2019 a big push in gender diversity in our team. We started the year with about 20% women and ended the year with about 50% women. And we did a whole range of stuff. We made sure that our interviews, A) candidates never had just one face-to-face interview, so if they were particularly nervous on one day they had another chance, B) that they were never interviewed entirely by one gender, that they had a mixed panel across the interviews. We published our employee handbook and tried to be really really transparent about our culture and how we support personal development and just like the day to day, week to week, month to month running of the team. Some of our team started teaching at CodeFirstGirls, a great coding bootcamp for women. We started just trying to make our culture a lot more authentic, publish more photos and blogs and things like that, we put all of our job adverts through a gender decoder to make sure that they didn’t have language or wording that was particularly appealing or not to certain genders. We lowered the job requirements on a lot of the job adverts, but we kept the same bar when reviewing applications. There’s a load of evidence saying that men will apply to something if they meet 3 out of 10 criteria and women if they meet like 9 out of 10 criteria. So by lowering the bar we tried to just get in, basically, to make sure our whole funnel was diverse rather than doing anything different about how we select people. It was all about making sure all the applications we were getting were really really diverse. These are some of the things we did, but it was a real team effort to do a lot of these things.
Where do you see the company in 12 to 18 months, and where do you see the healthcare industry in 12 to 18 months? And… no, just go with that.
JH: So, our big push at the moment is going beyond GP practices and primary care and getting adoption in hospitals and community providers like district nursing and therapy services and pharmacies and things like that. That’s our big focus at the moment.
In twelve to eighteen months we would like to show the same bottom-up adoption that we did in primary care, where in two years we’ve gone from zero to over 90% of the market. We would like to show that same trajectory in hospitals and these other providers. And from a core functionality point of view, we want all of those different people involved in the patient’s care to be able to log-in, look up a patient and then message other people involved in that care.
The other thing that we’re going for is that we want to be launching in another country, and that might be somewhere like Canada, which is very similarly structured to the UK system, or it might be going to a less developed country where they actually don’t have any existing system and we leap frog that, and the communication records become medical records. There are some places that really excite us, where we think that we potentially can have a massive impact, but we need to learn so much about those healthcare systems and their user needs and the like. That’s some of what we wanna do as a company. I think a lot of things won’t change, though. Like our commitment to user research, having a really strong culture, having a real drive to ship products and not get sucked into press and buzzwords or any of that stuff. That’s not going to work.
In terms of the healthcare systems, I think we’re at a really defining time now where a decade of digital transformation is going to happen over the following year. So we want to not just ride that out, but actually enable and lead that change and bring a lot of solutions, new solutions to the table and implement them. I really hope that, certainly in the UK system, in a year’s time, if you go to the hospital that that doctor there will be able to look at your primary care, GP records through our system. That means we will be able to get much better care for you, because we’ll know what your test results are, what operations you had before, what medication you're on. And then if they need to talk to your primary care doctor, they’ll be able to do that through accuRx. I think some of those transformations in terms of how information flows around the system are really exciting to start riding out in this country.
Are there any other non-obvious dynamics that COVID-19 will change that the rest of us aren’t talking about? Everybody is changing about how telemedicine will rise but is there anything that we, who don’t work in the industry, are missing but you think will actually happen?
JH: That’s a very good question. I actually think that the video consultation part of what we’re doing is actually a really small part of it and that once we’ve had really amazing growth in that, I think practices will find that being able to send documents o patients, being able to collect documents and get patients to message back are actually a lot more transformative. I think that some of the non-obvious impacts are that there’s a big impact at the moment on all of the routine things that the healthcare system was doing six months ago that have completely stopped. Managing people with diabetes, managing people’s asthma, parents worried about their children and bringing them in… there’s a big problem at the moment that in lots of cases that capacity is there but patients aren’t seeking it or taking advantage of it because they want to not constraint the system or not expose themselves to risks. I think there’s going to be a lot of catching up, particularly around the managing of chronic diseases, coming. What are the other unexpected changes? Let me see… None that jump to mind, I’ll have more of a think… I think really it’s just rapidly, rapidly accelerating how people are being focused on the opportunity rather than the risk.
What I think COVID-10 has really proven out is how, and I don't think this is unique to healthcare, is that in times of a crisis, very decentralised decision making can be far more effective and faster than very command and control, top down decision making. We had, when we released our services for consulting and triage to healthcare places for COVID-19, we had 3,500 practices. Four weeks later we had double the number of practices. Nobody told them to, you know, use us, but people were like we need a solution, this is available, we can set it up ourselves, let’s go. So I really hope that level of decentralised decision making and, in particular, innovation, continues to be on this, rather than things being centrally mandated.
Cool. I think that’s a perfect note to end on. Where can people find you?
JH: If you go to accuRx.com, I’m still embarrassed by the website. It's not great, but you can see a bit about what we do. If you’re in London and interesting in working for us there is a Careers page, and there’s lots of love from our users on Twitter, so if you want to search that and scroll through it you can get a real feel of how they’re using us.