I have often stated that front office AI (or AI receptionist, or voice AI, whatever you wanna call it) is the next hot Series A/B topic in healthcare. AI scribes were the first wave, then AI medical search, now AI receptionists.
Over the past 2 years, we’ve spoken with 20-30 of those companies in Europe and the US.
It’s time that I finally write down a my observations about that market. I’ll focus on Europe, since that’s our fund’s focus. I’ll throw in a few insider facts about traction and pricing, and will guess where the market is going.
Disclaimer: I’m slightly conflicted in this market, since Heal Capital is invested in two companies in the space. We just haven’t really publicly announced it. I’ll put that aside for now - the goal is to sketch out the market and not deep dive into individual companies.
The state of the market
In short: This is one of the rare healthcare markets that combines both a) an insane number of new companies and b) impressive traction.
Many of these companies have crossed €1m in ARR (some significantly more) in a year or less. Still rare for healthcare. As a result, a few frontrunners have raised large rounds early on, namely Vocca (FR) and Praxipal (DE), plus a few smaller €1-3m rounds across Europe. Most rounds that I know of came with steep valuations.
So far, most players remain local. Companies have typically picked one specialty (e.g. dentists, radiologists or aesthetic clinics) and stayed in their home country.
Nonetheless, there are a few strategic differences. For example:
Do you sell software or do you play a managed service approach? One might give you better margins. The other certainly gets you higher ACVs and a smoother entry into larger accounts. I personally like the managed service strategy
Once you’ve taken over the clinic’s phone with AI, where do you go next? Most companies go multi-channel immediately, so including text messages and emails. Then it gets interesting: Do you grow towards the back office, including documentation and/or billing, or do you grow towards the patient more in a Doctolib direction? Both valid strategies
Different customer sizes being targeted. Most go for very small clinics, some strictly do larger chains, and a few crazier founders even target hospitals
But overall I must say it’s hard to see true differentiation in the product itself.
Everybody has agreed by now that it’s a land grab and distribution game. More about that later. We’re now at a stage where we’re seeing first attempts to expand cross-border, and companies are raising new rounds. The next 1-2 years are going to manifest the market leaders.
Who’s joining the race?
My unpaid intern went on vacation to the south of France, so I didn’t have capacity to draw a full market map... sorry. But I’ll mention a few notable companies in the bigger European markets:
UK: Linda AI, InTouchNow
Germany: Praxipal, Praxisflow, 321med
Italy: CiaoDott, HelloSmile (thanks Antonio for inspiring this post btw!)
Spain: Ovianta
Netherlands: Inquira Health
Besides all the startups that spawned over recent years, there is additional pressure from multiple other directions:
Scale-ups like Doctolib, Nelly or Eterno. Doctolib was a surprisingly early mover with its acquisition of Aaron. I must say, I keep hearing that Aaron is rather limited, and not comparable to the latest LLM stuff. And Doctolib had the right instinct here, props to them. If anyone from Doctolib reads this: I still like your company overall! It’s just this product people seem split about
Other AI application companies, front and foremost of course AI scribes, will see this as a natural expansion opportunity. They have pressure to grow and this is one obvious escape route
Generalist AI voice companies. Fonio being a prominent example in Europe, or perhaps one of these many customer-support-voice-agents... They could very much focus on specific industries at some point. At this moment I tend to think healthcare is too specialized and tedious for them. Plus they definitely lack the necessary trust from physicians
Perhaps we’ll even see more DIY products built with ElevenLabs as another alternative. ElevenLabs’ marketing team would certainly like that! I don’t think it will play a huge role though, seeing that even in the scribe space true DIY products are rather rare, despite that being a much simpler product category. There are just too many edge cases and physicians don’t have the time to deal with it.
Bottom line? From outside the early-stage world, scale-ups can probably grab most land and I wouldn’t underestimate them. Their agility and distribution will do its part. I’m less concerned with the generalist companies and DIY tools.
What makes or breaks the market?
It is a tricky question what actually drives this market. Before you ask: Technology doesn’t. The cost and complexity of building an AI receptionist tool have dropped dramatically over the years. All the truly AI-native tools that I’ve seen and tested are super similar. Of course there’s a difference between AI-native vs older players - those that claim to be “AI” but are more or less pre-recorded voice snippets in a decision tree. I’m not even counting those in.
Then what else can drive success? Most of what defines the market, in my view, are commercial topics. To get great VC returns I believe it comes down to:
Sales & fundraising. My core assumptions is that sales and being the first to grab land is key to becoming a winner. If that is true, one of the underlying factors is how much you can raise. More money = more sales power. In such a market, that favors young, charismatic and irrationally driven founders... VCs love them. Look it up, the founders whose companies have raised the most are well below 30
Cross-border expansion. Local trust is an invisible force in healthcare. It protects local companies and blocks “intruders”. Very few B2B healthtech cases have successfully expanded into more than one large European market. To be fair, those who did are now worth billions!
Integrations & partnerships. As in most healthcare software areas, integrations into existing software, namely EHRs and calendar systems, are crucial. They’re the guardians of a clinic’s software stack. If you close large distribution partners or find a way to hack around lengthy and complicated integrations, that stands out
Pricing is another topic to watch. It might not decide the individual winner, but will definitely shape the overall market size. The question being: What are these companies’ ACVs and can they maintain that pricing as AI evolves?
The huge advantage of AI receptionists over AI scribes has always been their higher price points. A clinic with 3-5 doctors might pay €3-6k per year for AI scribes, but €15-20k for an AI receptionist. A promising starting point if you’re looking for VC returns! Whether you can realize €5k or €15k ACVs triples your market size.
Many VCs (including me) take this price level as a sign that the narrative of “we replace human work with AI” actually works, and translates into value captured by startups. That hasn’t been true for other AI applications, which are stuck in lower, seat-based pricing.
What the European market could look like
In all competitive markets an important VC question is: What will a winning scenario here look like? Is this a winner-takes-all market, a very fragmented one, or something in between?
I’d like to paint a scenario:
I believe in 2-3 larger winners in Europe that can bridge multiple countries. One or two might even manage a successful US entry. That brings extra upside. Next to that, every country will have at least one more national champion below VC-fundable size but still cranking out 8-digit annual revenues. Behind that, we will see a handful of specialized solutions for dedicated clinician groups. Dentists or therapists, for example, might have separate champions. Consolidation will clean up the market in a couple of years.
There’s one interesting variable: It’s unclear whether we’re going to see true network effects. Doctolib for example materially profited from those effects, accumulating patients on one and doctors on the other side. So far, I can’t see network effects in AI receptionists, but smart founders might be able to unlock them going forward.
All points combined, I’d argue it’s not a winner-takes-all market but still rewards scale. So looking at the boiling state of the market, founders currently face two options:
Go all in and raise what you can! Beat the competition by funding and enormous speed, no prisoners taken. At the risk of being completely over-valued and losing it all of course. Unfortunately, if you’re just starting to raise now, you might be late to the party
Establish a strong presence in a certain clinician group in a certain region. So dense, that even if a larger competitor enters your market, they’ll be forced to acquire you. At the same time, reduce your cost base early on. That way you’ll either have a nice profitable business or you’re making a good exit
I know that I’m captain obvious here - that advice is true to an extent for all VC markets. But I swear the dynamic is more intense here than usual!
Should VCs invest? And what’s next
Despite all open questions and variables I’m overall bullish on this product category. It combines many of the best VC traits: Incredible growth rates, great value capture and deep customer relationships.
Let’s see how large the VC returns are in the end. I’d bet on 1 or 2 VC-sized companies in Europe alone, the market size allows it. And Assort Health in the US has shown where this can lead with their latest $1.2b valuation.
What’s next? There are a few interesting use cases for AI voice technology in healthcare outside of the front office. For example, using AI phone calls to follow up after surgery (Tucuvi) or for intake in emergency rooms (Mindoo). Both awesome businesses!
I truly like those non-receptionist cases. Ping me if you’re working on one!
Speak soon,
Lucas
I am so curious about your thoughts on the US market. Planning to do a dive here.
Great take. As voice AI becomes easier to build, I agree that integrations, reliability, trust, and distribution will likely become the real differentiators—not just the technology itself.