Good afternoon, and welcome to Amwell's third quarter 2021 conference call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question-and-answer session. Please be advised that today's conference is being recorded. Leading today's call are Dr. Ido Schoenberg, Chairman and Chief Executive Officer, and Keith Anderson, Chief Financial Officer. Ido and Keith will offer their prepared remarks, and then they will take your questions. The Amwell press release and webcast links are available on the investor relations section of Amwell's website. Please note that we will be discussing certain non-GAAP financial measures that we believe are important in evaluating Amwell's performance. Details on the relationship between these non-GAAP measures to the most comparable GAAP measures and reconciliation thereof can be found in the press release that is posted on our website.
Also, please note that certain statements made during this call will be forward-looking statements as defined by the Private Securities Litigation Reform Act of 1995. Such forward-looking statements are subject to risks, uncertainties, and other factors that could cause the results for Amwell to differ materially from those expressed or implied in this call. Now, I would like to turn the call over to Dr. Ido Schoenberg, CEO of Amwell.
Good evening, and welcome to our third quarter earnings call. In the past quarter, we registered $62 million in revenues, reflecting stable subscriptions to our legacy platform and steady paid visit demand. The reception to our next generation platform, Converge, is significantly better than we hoped, as we are seeing clear signals of strong demand to our new platform from existing and new clients. Overall, we continue to believe Converge will growingly fuel our results in 2022 and beyond as more revenue is generated from existing client expansion and as new clients complete their deployments and begin using Converge. We are also pleased to note that larger than expected number of clients have already committed to upgrading to Converge, but some are waiting to expand their purchases and usage of additional care points, modules, and programs until their transition to Converge is complete.
As for our clinical services, we experienced some increase in visits quarter-over-quarter as the Delta variant led more people to use our urgent care services. However, at this point, we are maintaining a conservative view on Q4 visits, assuming relatively soft cold and flu demand due to masks and social distancing. While we are confirming our overall range of visit volume expectations, we are seeing disproportionate contribution coming from urgent care, lower fee visits due to the Delta variant. Because of these dynamics, we are setting our guidance to a range of $246 million-$253 million. In Q3, our gross margins were 43%, reflecting the continued costs of supporting our legacy platform and also commencing the conversion of our clients into Converge and benefiting from initial efficiencies associated with this transition.
Provider adoption growth on our platform is stronger than ever as total active providers grew from 71,000 in Q2 to 80,000, with AMG active providers steady at 4,000. We are also seeing a great expansion of the use cases on our platform across the continuum, way beyond urgent care. We expect our rapidly growing provider base to become a major catalyst of scope and frequency expansion as Converge deployments continue. We also saw a significant increase in the number of visits per active patient compared to Q3 of last year, which we believe indicates the growing demand for longitudinal care and virtual primary care programs. Most of our visits in Q3 were scheduled versus on demand, demonstrating sustained adoption by providers who are making telehealth a key part of their practices. Converge deployment is moving rapidly.
Together with HIMSS Analytics, we completed a survey of key decision-makers in healthcare organizations and found that 56% of health systems plan to invest more in virtual care over the next two years. Health plans also plan to increase investment in virtual care, with 88% of health plans intending to add more programs focusing on a range of needs from virtual primary care to behavioral health and chronic condition management. As we continue to roll out our first products on the Converge platform, we are noting that healthcare organizations are increasingly looking to consolidate their virtual care investments onto one platform. As demonstrated in the survey, health systems focus most on patient satisfaction and provider adoption. We are seeing positive feedback on both metrics with Converge. Over 4,000 providers have already used the Converge platform during its initial rollout.
With 43 enterprise clients now successfully using Converge, we have collected excellent feedback on our new offering and its significant value. For example, clients appreciate that our new platform video connection time is twice as fast on Converge versus our legacy platform. Most of these clients are using Converge through their use of Amwell Now, while growing number are using Converge EHR. At this point, we converted all Amwell Now enterprise users to Converge and have begun deploying Converge EHR with health systems focusing on Cerner and Epic clients. As a reminder, Amwell Now is our entry-level simplest product that has minor contribution to our revenues as a standalone offering. It does, however, share the common Converge platform, and we are encouraged to see it functioning so well in fully operational live environments. These successful deployments demonstrate the new platform's robustness.
They also serve as an important foundation for future same-store expansion. Converge enables innovation, agility, and faster speed to market with greater efficiency. For example, the level of effort in time and resources to release new software versions of Converge is less than 5% of the effort required to release new versions of our legacy platform. The integration of Conversa and SilverCloud onto Converge will enhance our already powerful and comprehensive digital care platform and enable hybrid delivery of physical, virtual, and automated care. These two companies bring important assets in data science, AI tools, and best practices in automation to drive proven clinical and financial outcomes optimization. Both feature important capabilities to enable our clients and partners to create and share chronic care, behavioral health, and other longitudinal programs, while significantly expanding the reach of providers and greatly improving their efficiency and impact.
The integration of both companies with Converge is progressing well and according to plan. Already in Q3 alone, we've seen cross-selling within both our customer bases. In summary, the receptivity to our new offering is excellent. Clients with relatively modest needs appreciate Converge's simplicity, reliability, and most of all, extensibility. Large enterprises value the comprehensiveness of our platform and its ability to integrate well with their existing assets. We believe that all entities will value the open nature of our platform and its ability to facilitate a diverse array of apps, modules, and programs from Amwell, our clients, partners, and independent third parties. Investments in digital health solutions are at historical high, driving a tidal wave of innovation.
The need for a common integrated platform to host, showcase, and enable broad spectrum of digital products is bigger than ever before, especially as it will streamline exchange of data, services, and capabilities across the ecosystem. Many customers and prospects commented that Converge flexibility and openness make it a smart future-ready investment. While other telehealth vendors focus on virtual care service provision and compete with payers and providers, we focus on enabling and powering their offering while promoting inter-segment connectivity. Together, we offer consumers a single, simple, delightful, and cohesive healthcare experience. With that, I would like to introduce you to our new CFO, my friend and partner, Bob Shepardson, who many of you already know from his previous ventures. Bob joined Amwell on October 31st.
I would also like to use this opportunity to thank my dearest friend, Keith Anderson, who will be departing at the end of the year, for his exceptional service and important contribution to Amwell over the past incredible three years. Keith?
Thanks, Ido, and thank you everyone for joining us on our third quarter call. For Q3, we reported total revenue of $62.2 million, an increase of 3% over last quarter, driven mainly by increased visit revenue. Total subscription revenue in the quarter was $26.7 million, flat over last quarter, and close to about 10% increase compared to the third quarter of last year, if normalized for the two customers lost due to M&A that we discussed on previous calls. Average contract values continue to be significantly above 2020 levels, with health plans at the $700,000 level and systems above $300,000. Total visits conducted this quarter on the Amwell platform was 1.4 million, with AMG performing 360,000 of these visits.
The increase versus last quarter was mainly due to an increase in COVID Delta-related urgent care visits while maintaining similar levels of specialty visits. Our clients continue to deliver significant care on the Amwell platform, as 75% of visits were delivered by our customers' own providers, a metric that has remained consistent over the last 6 quarters since the COVID spike in Q2 of last year. Total visit revenue was $30 million this quarter, a 9% increase over last quarter, but a noteworthy 6% increase over the third quarter last year. Unpacking the mix, AMG specialty volume has continued at 50% higher levels over last year, but sequentially was outpaced this quarter by urgent care visits due to the Delta variant.
This visit type mix shift toward urgent care has resulted in the revenue per visit metric migrating closer to $80 per visit versus $85 per visit in Q1 and Q2 of this year. This dynamic has also continued so far in the fourth quarter. Services and Carepoints revenue decreased slightly to $5.4 million from $5.9 million last quarter as several health systems work through their government grants that this time around are not time-bound. In other words, they don't have to be spent before a certain date. Additionally, similar to the dynamic Ido earlier discussed, a few health systems have pushed out new hardware deployments until they have been fully migrated onto the new Converge platform.
While these minor delays have so far continued in the early part of Q4, we are seeing the hardware pipeline continue to build strongly due to this dynamic and are confident in the ultimate conversion into revenue of these Carepoints once the health systems work through their grants and separately are up and running on Converge. Gross margin remained relatively flat at 43% of revenue this quarter versus 44% last quarter and 33% last year. Both of these comparisons are important and noteworthy since as compared to last year, we actually have slightly increased visits as a percentage of revenue while significantly increasing margins by over 1,000 basis points. This is counterintuitive, given visits are a much lower margin business.
Said differently, visits made up 48% of revenue this quarter versus 46% in Q3 2020, yet our gross margins were 43% this quarter versus 33% last year. Now add on the burden of increasing migrations of customers onto Converge, all while maintaining this improved 43% gross margin. This has been achieved through continued technology and operational efficiencies unlocked in the AMG business, as well as early aspects of the efficiencies achieved on the Converge platform. As a reminder from our Q2 earnings call, the migration of some of our more complex customers onto Converge would begin in Q3 and Q4 this year and would take additional resources to ensure the migration is seamless.
What we are seeing in the business and what we have done operationally and with technology confirms our ability to meet our IPO target margins in the mid-50% range and with continued mix shift over time. Eventually, even higher margins as our subscription business, our core business, is currently in the 60% gross margin range. Again, this will take time as we must first migrate our clients onto the Converge platform, execute on the pent-up demand for our programs and module subscription business, and the related Carepoints. Moving on to OpEx, we have also achieved similar efficiencies within our company operations. R&D expense in the third quarter was $27.4 million, representing 44% of total revenue.
While this is above spend levels in Q1 and Q2 of this year, as we continue to develop and roll out Converge, we believe the overall development cost of Converge will be less than originally estimated based on three quarters of development. We do expect Q4 R&D spend to increase versus this quarter, Q3, due to the additions of SilverCloud and Conversa, but we continue to achieve better R&D efficiencies through continuing to leverage our strategic partners in the development of Converge and also offshoring more aspects of engineering that we originally thought to keep in-house. Sales and marketing spend increased to $16.4 million versus $14.8 million last quarter and was mainly due to an increase in in-person conferences and clients ramping up marketing programs for their virtual care offering.
G&A expense was $33.7 million for the quarter, an increase of $9.5 million over last quarter, mainly due to the fees associated with our two M&A transactions, the hiring of our new chief growth officer, and some executive stock vestings. These one-time M&A expenses and stock-based comp have been excluded from our Adjusted EBITDA calculation. We are reporting an Adjusted EBITDA loss of $31.5 million compared to a loss of $23.7 million last quarter, with the majority of the increased loss coming from increases in our R&D and sales and marketing spend and the executive hire that we previously discussed. We ended the quarter with $790 million of cash that includes payments made in Q3 for the acquisitions of SilverCloud and Conversa.
In terms of active providers, we closed Q3 with 80,000 active providers delivering care on the platform, which is a dramatic increase of over 10% in just one sequential quarter. AMG providers remained constant at 4,000, so all of the 9,000 provider additions came from our customers adding their providers to the Amwell platform. As Edo explained, this is a clear and important indicator of strong provider adoption among our clients' own provider groups. Regarding our guidance, overall revenue growth in Q4 is expected to come primarily from increased subscription revenue, but also services and Carepoints. This could have been even higher without the Converge transition dynamic, but as Ido discussed earlier, we are seeing a deferral of new customer deployments and in same-store growth as a few customers are waiting to fully transition onto Converge to expand their module subscriptions and related purchases of Carepoints.
Similarly, some new customers are simply waiting for the completion of certain Converge components so as to go live completely and wholly on the new platform. Despite this familiar and normal dynamic, we are expecting subscription revenue to grow about 15% sequentially quarter-over-quarter, and services and CarePoints to grow about 50% sequentially quarter-over-quarter from Q3. We are forecasting visit revenue in the fourth quarter to be flat versus the third quarter due to the shifts we are seeing within the visit type mix resulting from the Delta variant. Throughout 2020 and the beginning of 2021, we saw behavioral visits correlated with COVID. Toward the end of Q3 and continuing so far in the fourth quarter, we unexpectedly observed a decoupling of this correlation, with urgent care visits continuing to expand while growth in specialty and behavioral visits less so.
This unexpected decorrelation was also observed by the CDC, who recently published these observations and findings in their October 8th MMWR report. Accounting for this unexpected and significant decorrelation, we are confirming the overall annual visit volume to be within the updated range we provided last quarter of 1.4-1.5 million visits, but at the lower quartile of that range and at an average annual revenue per visit now in the low $80 range versus $85 range we were previously expecting. Recall that except for last year's COVID spike in the second quarter, our fourth quarter is typically the largest all year for visit volumes, so this change has a magnified impact on overall revenue.
Taking all this into account, in terms of overall full-year revenue guidance, we are reducing the midpoint of the range by $7 million while at the same time narrowing the range as we approach year-end from $252 million-$262 million to now $246 million-$253 million to account mainly for this unexpected mix shift within our visits. In terms of Adjusted EBITDA, we are favorably increasing and narrowing our range by $10 million to an EBITDA loss of $143 million-$136 million from previous guidance of $154 million-$146 million due to this visit mix shift to higher-margin urgent care visits as well as further efficiencies unlocked within gross margins, continued favorability within R&D spend and sales and marketing.
In conclusion, we are pleased with the continued progress of the development of Converge and the migration rollout to our health system clients. We're excited for our health plan customers to also experience the benefits of the new platform. It is especially encouraging to experience the competitive and operational advantages we are seeing in RFP situations where the advantages of Converge are one of the primary reasons for our win. Further, our customers adding 9,000 providers to the platform in just this past quarter alone shows we are on the right track with provider adoption as providers are increasingly making telehealth a key part of their practice. Operationally, we are encouraged by the positive results we are seeing in our gross margins and operational spend. On the M&A front, we continue to evaluate opportunities to enhance the overall functionality at Converge to deliver full longitudinal and coordinated care.
With almost $800 million in cash on our balance sheet, we have ample funds to execute both on our inorganic strategy as well as to fund our path to profitability. As this will be my last earnings call with Amwell, I wanna publicly thank Ido and Roy for the opportunity to partner with them and to finance the realization of their vision. I also wanna thank the board, the other executives, and the entire Amwell family, but especially my heads of financial planning, accounting, and M&A, and their respective teams who worked tirelessly to prepare Amwell for the IPO process and then ultimately had to execute the IPO remotely as our physical office was shut down due to the pandemic.
It was never lost on us that in the end, we are a healthcare company that empowers healthcare providers to deliver their care remotely to their patients, and during COVID, this remained our primary focus. I'm now handing the reins over to a tenured executive who knows Amwell extremely well, as Bob was my banker counterpart on our IPO. He's a good cultural fit with Amwell and a well-known figure on Wall Street who I'm now trusting to lead my teams. I have complete confidence in Bob and wish him all the best as this is truly a unique and fantastic opportunity. I'll now turn the call back over to Ido for his closing remarks. Ido?
Thank you, Keith. By now, all of you realize the enormous transformation in Amwell and with our clients and partners. We have never been more bullish or excited about our future. Our incredible effort to streamline our platform transition to Converge is going very well. We are seeing clear signs supporting our strategic path and look forward to sharing more proof points with you when our clients and partners permit. Now, I would ask the operator to open the call for questions. Operator?
Thank you. Everyone, to ask a question, that is star one on your telephone keypad. Again, that is star one on your telephone keypad. We'll go first to Eric Percher of-
Your line is open.
Thank you. Excellent to hear the client feedback on Converge. Is there any perspective you can provide on the headwinds, tailwinds, and acquisitions as we exit 2021 and think about beginning to model 2022? I'm thinking specifically the demand for Converge and a reasonable view on impact to subscription as well as the impact you're seeing on flu and urgent care on volumes.
Thank you, Eric. It's good to hear your voice. As usual, your questions are so great that I wish we had an hour to answer them. But I'll do my best to give the group the headlines. I think in summary we see some dramatic changes in the market that luckily for us fit very well with the mission and purpose of Converge. I'd like maybe to give you a number of examples, and then if time permits, I'm happy to explain in more detail how those tectonic market changes are being addressed in the new platform. Essentially the name Converge is very appropriate because we all witnessed some serious market convergence in many ways.
The first area is the area of segments, starting with payers and providers. We all know that very large payer organizations are definitely going and owning more and more provider assets. United is a great example with their 60,000 or more providers that are part of their group. Providers are taking more and more risk and offering risk products. Intermountain, UPMC, and many others are a great example to that. In any event, even if those entities do not share both tasks, the future of any entity is dependent on its ability to become collaborative. There are very few platforms, if any, that have very strong footprint both in the provider and the payer space, and very importantly, enabling connectivity. Converge does that.
The third segment that is converging into this mix are innovators. We're seeing historical phenomena, and I think you alluded to that in your question, with digital health innovation. This year alone showed more than $20 billion in new digital care investments up from $14 billion in the previous year. That was about double the year before. You can only imagine the confusion and complexity of decision makers, CIOs, chief accounting officers, actuary officers, I'm sorry, that need now to try to prioritize and embed all those solutions into their ecosystem.
The role of Converge is to really allow for a consistent middle tier that brings together all those three segments and really align those organizations to growingly behave like population health services organizations, which is really a combination of care services and risk management. To try to reimagine the future, not as a replacement of physical care into equivalent digital future, but rather total reimagination of a new single better healthcare experience. I would also suggest that the Converge includes a new focus and clarity as to what's important by many ecosystem players. There is great focus on improving clinical and financial outcomes. There is clarity that the market is shifting towards risk and fee-for-value with great importance to digital tools that are now moving from the sidelines into center stage.
Everybody's comfortable in understanding the importance of omni-channel, and no one suggests that there is any other solution but a hybrid solution that brings together physical, virtual, and automated care, including a liberal view use of AI of natural language processing, and many other technology. Transactional is working together with longitudinal, not only for acute care or urgent care, but really across the full care continuum, addressing the full person. These type of tools are very, very effective to help contribute to network optimization, referrals, steerage, things of that nature, and also allow for extensibility, bringing insight and services into the last mile. Lastly, there is growing recognition and acceptance that of the great importance of the consumer experience and the provider experience.
Getting a great experience is practically impossible without unified identity, without creating a unified profile for those, for both those entities which Converge allows and allowing for consistent engagement of these, important, players. Probably a mouthful, a pretty long explanation, but we were lucky to think about those requirements a few years ago when we started to bring Converge onto the drawing board. That really explains the extremely favorable, almost surprising, very vibrant market positive reaction to the platform that we are deploying from small players and large players, existing clients, and very important sophisticated new ones.
It's a vivid description. If I can pinpoint the question, are we to take from your commentary that the adoption and response is running better than you anticipated, I think was your comment, and more than offsetting any delay as there's any pause on purchases until Converge is rolled out?
Yes and no. Let me explain. There's no question that what we see in the market is much better than we expected or even dared to hope. Unfortunately, as you know, we don't share those things. We let our clients use our platform as empowering their business agenda, and they will share that when they're ready. As far as the impact of a platform transition, I'd like to really explain it a little bit to the team. This is something that is not surprising, that we expected and is baked into our guidance, but it's something that basically happens each time you do such a major move.
When you announce a platform, and we wanted to announce it early to really make sure that our existing clients and new ones really understand where we're going and to allow them to participate and plan on what's coming. As you remember, we shared Converge in April of this year. When that happens, usually, as Keith said, and I said as well, usually clients, more and more clients, especially when the component delivery nears, prefer to delay a little bit expansions and extensions to their upgrade to the platform. The same is true with new clients in many ways that really are excited about what's coming. The nature of Converge favors large complex clients.
Add to that the fact that those wins usually come with a very, very large deployment cycles that take a little longer to begin to play out. There is an impact on revenue growth when you transition one platform to another. It's not negative, it's normal, it's expected, and it's something that is completely not seen in the numbers that we shared today.
Thank you.
Our next question will come from Ricky Goldwasser of Morgan Stanley.
Yeah, hi. Good evening, and Keith, it was good working with you and good luck in the future. Bob, looking forward to working with you, more. Always enjoyed that. Two questions here. One, clearly we're only in November here, so still early, but maybe if you can sort of give us a sense of the headwinds and tailwinds for 2022. Ido, just to clarify, as we think about clients sort of delaying their expansion of the platform or those new large complex clients sort of delaying decisions, do you expect to see those clients committing in 2022, or should we think about the inflection point in 2023 once the Converge platform is fully rolled out and deployed?
Sure. Ricky, it's good to hear your voice. I want to point out very clearly and early on that we don't see the dynamics of any delays in decisions. On the contrary, I just mentioned that clients are ready to make decisions. They are making decisions, existing clients and new ones, and these are very favorably positive decisions. The time it will take for these decisions to play out into our revenue really depends on things like the delivery of Converge component and the sequence that they are released. When we look into 2022 and beyond, we really see greater conviction and greater clarity the dynamics I was talking about in previous earnings calls, which is that we plan to basically convert and upgrade the lion's share of our clients through that year.
As we convert those clients, we also expect to see sensible growth, and we also expect to begin to see new clients kick in. All these things are very positive and obviously we at the same time, we're going to gradually obviously retire our legacy platform, which would leave us with something that is very scalable, very efficient, and we believe we're going to see a very strong growth drivers as this trend develops. When I look at tailwinds, I think I mentioned them in my reply to Eric.
The main tailwind is the transformation in the market, the acceptance of digital formulary, digital capabilities into the main care pathways, into the mainstream of healthcare, and we have the technology that fits that very strong brand. In way of headwinds, I think the market is overwhelmed. I think the people are confused. There is so much innovation, there is so much investment in digital health that people really need to take their time and understand what's working and what's not working.
One of the benefits that we bring to the table is to create a platform that allows you to experiment very, very quickly with those solutions without giving up on common reporting, on common engagement capabilities, on identity management, the lower, more fundamental elements that are required to really streamline the new model of care while relying on innovation that is probably very necessary in order to truly realize optimal outcome improvements.
Ido, just to make sure that I understand correctly, because to your point, you said it's not that the clients are delaying their decisions, they're making decisions. We understand that the contracts are being signed, they're just a greater lag between signed contracts and implementation. If that's the case, does it mean that you now might have better visibility on sort of kind of like the revenues that you are expected to come in within the next 12 months-18 months?
I would say that we have much greater conviction in the trend that we were talking about because we see the very important component, end-user reactions, sponsors' reactions, clients and partners voting for Converge. This is a serious decision. When you look at especially very large organizations, this is not a one-year decision. This is a decision for a platform that may be their platform for the next five or 10 years. My conclusion is driven not by a quick event that happened recently, but in many cases, we were looking at RFP processes and diligence cycles that took a year or a year and a half and are now behind us. That's a very thoughtful, meaningful decision by those very sophisticated buyers to side with us.
Obviously, that trend does increase our confidence in our ability to grow revenue in 2022 and very much so beyond as what I was talking about is going to play through.
Thank you.
Next, we will take a question from Charles Rhyee of Cowen.
Yeah, thanks for taking the questions. Keith, good luck with everything, and it was great to work with you. I guess my question, you know, is kinda following up a little bit. You know, obviously good to hear some positive feedback on Converge. Can you talk about, I might have missed it, but what percent of your clients have now made the switch to Converge? Secondly, you know, can you talk about what other telehealth platforms, you know, some of these clients may have been using? Have you seen a concerted effort by these clients to consolidate all of their virtual capabilities solely onto Converge?
Sure. Hi, Charles, good to hear your voice. As I mentioned earlier, we have 43 enterprise users of mostly Amwell Now that are now on Converge. As I mentioned in my prepared remarks, this really demonstrated the heart of Converge is working very well. But these are, to be fair, use cases that contribute very little to our revenue and are the simplest type of use of the Amwell platform. Converge EHR is obviously more complicated, and in the next phase, you're going to see also payers and really go for our entire client base. Counting the number of accounts may not give you the right answer.
The number of accounts, if you're talking about a couple of hundred, you get a proportional client, but you need to take into account also a complexity. What we do see, which is not reflected in this number, is the reaction of clients and the readiness to embrace Converge. Many more clients, obviously, raise their hands and made a decision to upgrade to Converge or buy Converge as a new client. That's a very meaningful decision. Of course, as Ricky mentioned, there could be a gap between this decision and the actual deployment and its impact on revenue that you need to take into account.
The excitement of my team and our clients and our partners really relates to that type of decision. Now, when you look at those clients, obviously, almost all of them have different solutions running in parallel. What we see quite frequently are different video conferencing solutions, from Teams to Zoom and other tools. None of those companies necessarily see us as competition. We happen to work with Twilio, but that type of activity really resembles a little bit what we have in Amwell Now. I don't think they even aspire to do the many other things that I explained earlier on this call.
What we do is pretty unique, it's extremely complicated, and thankfully now, very highly appreciated. I think that the difference is that we are very likely going to become part of the care pathways. It's no longer going to be a situation where telehealth is promoted as a side event, but rather a digital interaction, not only video visits, but everything else that we do, including automated programs and interconnectivity of information and insights that is designed to improve outcome, is really going to become part of the healthcare. That's a very big lift that took us 15 years to develop. We invested more than $1 billion, and we're still investing heavily. It's enormously difficult to recreate, and now it's a very, very relevant.
We don't feel threatened by these types of companies. At the same time, I would suggest that people like Epic are definitely trying to raise their way into this space by working on episodic care and things of that nature and programs. We have a different view into how to reach that. We see enormous progress of Microsoft trying to enter into healthcare from a different angle, but again, very dissimilar to what we are trying to do. Based on the market reaction and based on where we are, we feel pretty confident about our ability to maintain and even significantly grow our market share in the not so distant future.
I appreciate that. If I could just one follow-up to Ricky's question. You know, I think when you announced Bob joining Amwell, you made a comment around expectations of well over $300 million for next year. Can you comment on that? Is that still sort of the target for next year? You know, any way to maybe frame out what, you know, well over could entail? You know, we're talking, you know, significantly over or is it just, you know, a little bit over? Just any ways to maybe frame what, you know, maybe well over means. Thanks.
Well, we are going to continue and grow next year and the years after. We are going to dedicate our next earnings call to talk about our focus for next year, and you'll have your answer, I promise. I don't want to give unclear hints at this point. Such a dynamic market, and I want us to be very thoughtful and take our time, but directionally, we see very positive and encouraging signs.
Great. Thank you.
Our next question will come from Sean Wieland of Piper Sandler.
Hi, thank you very much. I'm just generally confused on what's going on with visits. Maybe if you could just take another crack at it, starting with last quarter when urgent care visits were weak, now they're strong, the relationship with margins and how they factor in a paid versus unpaid. Like, where are these, where is this mix shift showing up in the model?
I'll take that one.
Go ahead, Ido.
Thank you. Bill needs to complete my answer. We are all confused. I mean, COVID and Delta are very confusing. Let me maybe walk us through the visit dynamics and try to bring some order to understanding it. In general, clinical services and AMG visits are very important in supporting our technology impact. They're important in at least two ways. The first one is obvious. We are checking the box on urgent care, convenient care availability. Everybody needs that in the market, and we offer short wait times and scalable solutions today in a very efficient way. We focus on quality and safety and regulatory compliance and all those things that we and many other telehealth vendors are doing.
The more important part of our network is designed to really help improve outcomes. What I mean by that is that these services are now growingly embedded in programs and clinical pathways. For example, our ability to use AMG in virtual primary care in longitudinal programs like the one that Conversa had in chronic care programs is here to stay. Unlike the previous era that eventually we believe will fade out as many of those transient or transactional providers are going to really be replaced by our doctors. The ability to use our network for specialty visits and longitudinal visits is important. It's also helpful because the way that you create demand for the second category is much more efficient than the first one.
As you probably know, customer acquisition costs for various reasons is much, much higher today. When companies try to sell urgent care visits, the margins are fairly depressed and the customer acquisition cost is very high, and it's becoming really not as optimal a business as it was before. The demand for embedded clinical services is really coming from your doctor and from your hospital and from your care pathways. That is something that is much more natural, and it's going to grow. The customer acquisition cost is much lower, but it also bring much higher value because it's connected to much more important outcomes. Now for what happened this year in way of the visits.
Of course, the year before we had COVID, everything showed up quite dramatically. In general, demand for urgent care came down slightly for us and many others versus the peak of last year. We saw a much softer, or we predict much softer, cold and flu season because of masking, because of social distances, and because of hygiene. That was some of the reasons why we adjusted our guidance for visits in our previous call. What we saw now in Q3 was surprising and interesting. If you look at the dynamic, and the CDC gave very good data that you're welcome to really look into.
At the beginning of the year through Q3, there was a very clear correlation between COVID and behavioral health visits. As COVID showed up, the demand for behavioral health, anxiety, depression, things of that nature, showed up as well, and that also contribute to the mix of specialty visits that we already reported. In Q3, that broke. We saw the Delta variant go up and demand for behavioral health actually break or separate from that correlation. While it's anybody's guess to really understand that, hopefully it's a good thing. People get more comfortable and live with COVID in a different way, and that creates less of reaction or emotional reaction.
We had to adjust our forecast for Q4, which is a very important quarter for us in well visits, because of the seasonality in urgent care visits during this quarter. Keith, I don't know if you have anything to add.
I mean, Sean, I think your question was more, you know, dynamics. I'm happy to give some math, I guess, to just confirm, you know, for the models that the adjustment and guidance is just focused on the visit line. You know, we're gonna hit where in the beginning of the year we targeted for, you know, the subscription line as well as in the services and CarePoints. You know, for visits, given all the reasons that Ido said, that we were observing, you know, toward the end of Q3, and we're still continuing to see this play out, and the CDC predicts this will continue through the end of the year. You know, we are adjusting our visit forecast.
I guess before I end, I just, you know, Ricky and Charles, thanks for the kind words and look forward to working with you again in the future.
All right. I'll leave my question there. Keith, best of luck to you in the future, and thanks for taking my question.
Thanks, Sean.
Next, we will take Ryan MacDonald of Needham.
Thanks for taking my question. Ido, maybe for you, on the first one. Of the 9,000 providers that you added with clients, obviously some great progression there. Just curious what you're seeing, in terms of activity from those providers, that have joined Converge and, perhaps what you're seeing from a mix of specialty versus urgent care, use cases so far.
Hi, Ryan. That's a really good question. Materially, all of them are not AMG providers. They are all our client providers. As you know, our clients typically are academic medical centers, large IDNs and so on and so forth. Most of these visits actually are all over the place in the way of a specialty. This change really is demonstrating a secular clear trend that we see of growing adoption of digital virtual connectivity between our clients and their patients that are not in the room from their EMR, from the EHR. This is really covering the full spectrum of the care continuum, and very rarely, if at all, is primary care or urgent care.
That's helpful. Maybe one for Keith on gross margins. Obviously, you talked about some of the great progression and efficiency in the model despite sort of this mix shift back towards visits. You know, as we think about next year and the continued progression there with Converge and some of the benefits, you know, how should we think about that progression? Are we at a point now where we can get back to, say, pre-pandemic levels when thinking about sort of the gross margin range? Thanks.
I mean, given what we've seen, you know, we're now supporting two platforms. You know, as we migrate customers over onto, you know, the new platform, the new Converge platform, we have to maintain the old platform until all the customers are over on the new platform. You know, what we saw this quarter, and you should read into it, you know, is really exciting because, you know, we started in earnest in Q3, and it's continuing in Q4, you know, starting to migrate some of the more complex, larger, you know, Amwell customers onto the platform. We were expecting, you know, gross margins to, you know, hover around 40% in Q3 or in Q4, maybe even go below 40%. You know, we ended up at 43%. Part of that is the mix shift.
You know, urgent care visits, we do generate higher margins, you know, for those than specialty. You know, but also just the continual unlocking of efficiencies within the AMG group that we've been talking about all year, as well as, you know, technology margins continuing to expand, you know, by further leveraging our partners and, you know, partnering with them. I think we're gonna end the year, you know, right around, you know, the 40-41% mark, which is above where, you know, I thought we were gonna be, you know, in the beginning of the year when we announced Converge. We're really pleased with, you know, with what we're saying, or what we're seeing. You know, we now have
You know, a significant degree of confidence getting back to, you know, what we said at the IPO of the mid-50% range. Our subscription business, you know, generates margins in, and, you know, the 60s and not at the lower end. Just a matter of finally getting all of the customers migrated over onto the platform. You know, continuing to unlock the efficiencies within, you know, our services side of the business, the AMG bit, and then further partnering and capitalizing on the partnerships that, you know, we've established with some other technology players.
Thanks. Best of luck in the future.
Great. Thank you.
Now our next question will come from Jailendra Singh of Credit Suisse.
Thanks, and thanks for taking questions. Keith, good luck for your future. I joined a little late, so apologies if you already addressed this, but can you update us on any traction you're getting with virtual primary care? How should we think about that opportunity within the context of Converge? I know you guys have very small presence in direct to employer market, but as you go to market for virtual primary care, will that strategy evolve in any way where we might see you guys potentially market more direct to employers or not really?
Hi, Jailendra, and congratulations for your recent wins. The virtual primary care or digital first is going to be a common way that people interact with healthcare in the future. I think it's very clear right now. Of course, it is a very important capability of Converge. We don't necessarily see it as an independent program. It's one of the modules, one of the capabilities. We believe, based on what we see, that it's going to continue to grow and become a very popular capability.
As you know, we are very loyal to our payer partners and believe that they are the right channel to manage services with the employers, and we don't expect any change in this strategy in the near future. What we do right now is from them, obviously, we serve very large number of employers that seem to be very comfortable with the growing quality and capabilities that are offered to them by their payers. The sophistication of payers could be very helpful to employers by basically now allowing to connect, get some care and other analysis with the providers that we enable connectivity to.
That is the type of service that an individual employer may be hard-pressed to obtain alone. It's not their day job in way of really stratifying the risk. That is part of the reason for our strategy and maintaining it.
Thanks. Thanks, everyone, and thanks for your comment. My follow-up, I was wondering if you could provide updates on your Google partnership. In addition to the capabilities captured in the Converge platform, what are some other areas where partnership has helped the two companies? Any other areas where you think that opportunities are still untapped and could be future opportunities for you guys?
Yes, absolutely. If you remember, there are a number of areas of collaboration. There are a growing number of components from Google that are integrated into Converge. Many of them are an option for our clients to choose from, and they're very powerful and capable. Some were announced, and some are in play and in process. We are also enabling more capabilities on the Google Cloud, which is working well both financially and operationally for us. The last element is relates to our collaboration around distribution. As you know, we really focus on the U.S. market today. Google is a global company.
Growing, we built Converge, so it could really be relevant in the full global term where Google has some really interesting assets. And that's pretty much what I can say at this point. I would summarize that we are thrilled by this partnership, and it's just beginning to really develop its very strong full potential.
Great. Thanks for taking questions.
Next, we have Stan Berenshteyn of Wells Fargo Securities.
Hi. Thanks for taking my questions. First, I guess I'd like to echo my sentiments towards Keith and Bob. I guess a question on the willingness for clients to be first movers onto Converge. Is the expectation for larger clients to be first movers or for smaller clients to be first movers as you're thinking about the upgrade cycle?
Stan, that's a great question. You know, if you've asked me this question in April, I would definitely say that probably smaller clients would be ready to jump in more easily. It's less complicated. It could be a less perceived risk. Interestingly enough, what happens right now is that we see just very large clients jumping in early. They probably realized the enormous value they could extract from the platform. A lot of it is fairly special and unique and could provide them a special advantage. I'm really looking forward to the day that I could be more specific when these entities will be comfortable sharing their business strategy that really relies on the Converge.
To be fair, we don't have any difficulty not with small or big clients. There are different advantages and disadvantages for each. Even small clients love the idea that there is a very, very big engine under the hood, even if you only use Amwell Now, and you can really grow it to your heart's content in a very diverse way when you're ready, coupled with the fact the experience is pretty awesome. I mentioned earlier that there's this super-fast video times. There are many other attributes that really delight both providers and the patients, even in the simpler usage.
Got it. That makes sense. I guess there's obviously a lot of moving pieces here. You have the upgrades, you have cross sells, you launched a new product, Converge. Just thinking about it from a sales standpoint, what is the focus for your sales force over the next 12-18 months, given all these moving pieces?
Sure. Our number one focus is our existing clients. We have a very large market share today, and we are laser-focused on streamlining the upgrade process as much as possible. White gloving it, we're being very attentive, getting feedback as quickly as we can and implementing it as fast as we can, to really make sure that our very vast installed base is happy and ready to take advantage of numerous new options that the new platform is bringing. At the same time, we are also looking to expand our market beyond our current market share. Naturally, we're looking at the big clients first, because they represent an enormous disproportionate part of the ecosystem.
As you know, the number one priority for Amwell is actually providers. We really want to make sure that through those very large entities, we get as quickly as we can to as many existing providers, so they become part of our virtual network because we believe that's the most sticky, most relevant way to really create enduring impact on consumers and on clinical and financial outcomes.
Got it. That's helpful. I guess maybe just one quick housekeeping question. I was surprised to see the non-AMG providers were up this quarter. I think the expectations were probably to see some more slippage. Anything to call out where those providers came from, why the sudden uplift?
Sorry, do you mean?
Do you mean active providers or do you mean non-AMG visits?
Providers.
Yeah.
Non-AMG providers.
Yeah, the non-AMG providers went up 9,000 in just one quarter. I mean, that's a significant uptick, you know, sequentially in just, you know, 90 days.
Yeah. My question is, I believe the expectation was for that to be a negative growth this quarter or maybe still kinda light, because I think there was some COVID-related onboarding last year that you saw that would possibly flow off the platform this quarter. I'm just curious what actually ended up driving the uplift in providers in non-AMG providers this quarter.
You know, we tend to talk about our new platform, Converge, a lot, and we tend to discount the fact that our current platform, our legacy platform, is working very, very well in the market, and so many people are using it and things of that nature. That growth is really secular growth. It's the adoption of providers and embedding digital connectivity as part of their workflow. You're right, technically, that we saw an enormous surge before, and I'm sure we lost a lot of those providers, but we won net many more. Actually, the 9,000 is, you're right to point out, is really reflecting even bigger change if you take into account those one-and-done COVID-related providers that are now clearly out of the mix.
I would simply suggest what we see is historical. What we see is very dramatic in the sense that digital tools are now becoming part of the main pathway of healthcare, and a lot of it will be on our platform.
Got it. Thanks so much.
Next, we will go to Ravi Misra of Berenberg Capital Markets.
Hi. Good evening. Thank you for taking the question. Just on the first one, on the subscriber business, I think you mentioned average contract value is over $700K in the health plan and maybe over $300K in health systems. Is that kind of on an organic basis? Because that seems to be trending in the right direction or is there anything from the acquisitions that's pulling that up? Maybe secondly, and I'll just ask my second question up front. More of a maybe just like an operational question. You know, the visit business has been, I don't know how to put it mildly, but you know, it's caused a little bit of a variance in the kind of forecasting here.
I appreciate there's a lot of mix shift going on. There's a lot of new providers being added. But what gives you the confidence maybe, you know, looking into 2022, without giving us numbers, that you have, you know, the understanding that you need to kind of limit the volatility, I guess, in that space? Thanks.
It's all organic on the subscription side. I mean, we just closed those two deals at the end of the quarter. In terms of the subscription and the growth there, that is building nicely by our current and new customers. In terms of the visits, we went public in the middle of the pandemic, and we knew that the visit component of our revenue was gonna be volatile. You know, I mean, it was a pandemic that nobody had a crystal ball, and we said those exact words when we went public.
That's why we compartmentalized, you know, the visit revenue and the visit part of our business, you know, from the rest of the model. When we went public, we gave a number, you know, and in the beginning of this year, you know, we confirmed the same number. You know, in Q2, the Delta variant hit and, you know, many different sources as well as us were expecting a lower flu season, so we had to adjust when we started seeing that. The significant unexpected decoupling of behavioral to the spike in COVID visits, you know, we started seeing that after the call, you know, in August, which was, you know, halfway through, you know, the Q3. You know, that's why we're adjusting it as well.
You know, again, we're not adjusting any of the subscription expectations we had in the beginning of the year, nor the services and care points. It's just this visit component that, you know, COVID is still happening, and there's still dynamics that, you know, people are discovering how it spreads, you know, through society. We'll continue to isolate the visit part of our business and, you know, continue to talk about, you know, the subscription technology aspect of our business, you know, the subscription line, as well as the services and care points that also feed the subscription line.
Great. Maybe just one quick follow-up. I appreciate the color there. You know, as we hopefully soon get out of COVID, whenever that may be, do you think the kind of variance in that visit fee business should kinda normalize to, you know, your what your expectations were prior to COVID? Maybe help us think about that. Thanks.
Sure. I mean, it's
So, so, uh, go-
Yeah, Ido, you can take that.
No, no, please go.
Absolutely. I mean, this is, we have a deep analytics team that is all that they do. You know, they look at the different sources. We, you know, are able to look to the southern hemisphere as a leading indicator for flu season. You know, we also monitor what's happening with the CDC as well as, you know, other telemedicine companies. We also get feedback from our customers, you know, the health systems, and we can see the volume, we can see the dynamics, you know, of how their providers are delivering care, you know, to help forecast with a degree of accuracy what's gonna, you know, what's gonna happen with our AMG business.
You know, what's interesting is, you know, for the first two quarters and all, you know, toward the end of last year, you know, the specialty business was continuing to grow nicely as a greater portion of our customers' doctors were delivering more and more urgent care. Remember, that urgent care aspect is the core business of some of the other telemedicine companies out there, and we do see that continuing shift to, you know, doctors delivering that type of care to their patients. It's nice to see the continued expansion of specialty. It is still growing in Q3, and it's still growing, you know, slightly in Q4, but there's a decoupling. It's not following the dramatic steepening of the slope of urgent care happening, you know, toward the end of Q3 and continuing so far, you know, halfway through Q4.
Ravi, maybe just to add color on what Keith just said. The answer to your question is yes, we're going to have much more reliability or predictability going forward for the following reasons. The most important reason is that growing part of our business will be sticky ARR. Very large clients that use our technology in a very integrated way for their day job business, for their main business. This is something that very much like a big Epic implementation. Once you do it, you stay with it for a really long time, although we are very, very different from Epic, so that's the only analogy here.
The second point that increased reliability and predictability of services, of the part of clinical services, relates to the fact that we believe that the urgent care component, the one-and-done business, is going to basically be flat or even be much smaller than the growth of prescribed services that are part of programs. Think about the SilverCloud Health, their program that includes a lot of automation that prescribes now a conversation with a therapist, with a psychiatrist or with a psychologist. The likelihood of that visit of happening is much, much higher than a campaign of an employer or a payer suggesting that you should try to talk to a psychologist, which we know is very expensive and very inefficient.
As digital will become part of the mainstay of healthcare, the ability to predict volumes is going to become much easier because unfortunately or not, healthcare works all the time. It's pretty consistent to see the demand for those type of services. That's what AMG and APC, our networks are going to do growingly next year and in the years to come.
Next, we will go to Donald Hooker of KeyBanc.
Great. Thanks for taking my question here. I'll just stick to one here. Just so obviously, we're all seeing and hearing about labor shortages and wage inflations across the healthcare system. I'm just wondering if that could, you know, do you see that potentially creating a scenario where sort of AMG sort of demand for outsourced visits might increase around some of the lower acuity might expand that business in a way that you hadn't contemplated before? Or is that not the case?
Well, it's hard to tell. It does make sense, obviously. To be fair, I'm not sure what proportion of the healthcare demand is this type of relationship can solve. Your point is very valid, though, and I would say that the answer to this is make sure that there is appropriate care, so navigation to the most appropriate intervention through virtual primary care and things of that nature is a very good step in the right direction, and the use of automation. The use of programs like the SilverCloud programs that is deployed with 80% of the members of the NHS, so it's a very big deployment there, and was proven over many years to really dramatically increase the reach of therapists.
Saying it another way, you have very few therapists that treat many more patients much more effectively by the appropriate use of tools like AI and conversational navigation and things of its nature that are all automated. I believe that the technology impact would far outpace the availability of urgent care network in way of helping this problem.
Don, where we are seeing it continue to exacerbate is on the psych side, you know, psychiatrists. There-
Mm-hmm.
The shortage in the U.S. continues and, you know, there are increasing visits happening, you know, in the emergency room. You know, that was the whole thesis why we bought Aligned Telehealth a couple of years ago, and that thesis has been, you know, very accretive to our business. You know, there's so many of our health system customers that just don't have the staffing to meet the demand that they're experiencing of people showing up in the emergency room. You know, so that business is performing nicely.
Okay, super. Thanks so much.
Now we will go to Allen Lutz, Bank of America.
Thanks for taking the questions and for squeezing me in. I guess, going back to AMG and more of a conceptual question, you know, I guess first, are the amount of customers using AMG increasing? And then is there any way to kind of frame, what visit growth would look like in a normalized environment? You know, you kinda mentioned that urgent care expect that over time to be kinda flat to down. Maybe you were speaking more about the industry as opposed to the AMG business. But just trying to get a sense of how to think about both urgent care and behavioral, you know, going forward. Thanks.
Sure. That's a great question again. Deserves a really long answer, which I won't, you know, bother everyone with. But in a nutshell, there are two types of visits, right? There is the visit where I couldn't find my doctor, and it was very convenient and impossible to find a really simple provider online to prescribe whatever I need. I believe that there is value in that type of service, and it's very, very convenient. It's not going to die out. Obviously, people, they need that. But as more of my doctors, my trusted environments are going to be more and more available digitally, I believe the public will always prefer those trusted relationship with full access to your record that you can see in person and so on and so forth.
On the flip side, what with the growing use of longitudinal programs that heavily use automation and things of that nature that are based on very sophisticated collection of insights, that can engage the consumer and providers continuously, there is a need for, if you will, cloud availability of clinical services, both urgent care and specialty services. There is a place for these type of services. Imagine a group of experts around COVID or another special topic that are available virtually, nationally, and are making their services available to everyone, including primary care providers in their offices and their patients, when needed. That's not going to go away. The redistribution of services, including cloud availability of clinical services, in my opinion, is here to stay.
That business is very, very important because it can really contribute to closing gaps in care that cannot be closed by automation alone or by existing relationships alone. That's a very, very different type of business that is just beginning to grow right now. To sum it up, I believe that overall, the AMG services are not going to die out. They're going to replace their purpose and they're going to grow, but they're going to be growing less fast and in a less impressive way versus the technology, which is really the main contribution that we bring to the new model of care.
That's why we're so focused on the specialty visits, because specialty, you know, again, our mission is to supplement, you know, rather than, you know, replace. Supplement, you know, the capacity of our customers' own doctors to deliver care to their patients or their members. Specialty in so many situations is more supplemental than, you know, being the primary, you know, deliverer of that care.
Great. Thank you.
Thanks.
Your final question will come from David Larsen of BTIG.
Hi. I just have a quick one. Can you talk a little bit about your relationship with Anthem? Anthem obviously has a digital-first plan design. Is American Well the telehealth solution that will basically power a big piece of that virtual first plan design? Can you also talk a little bit about your relationship with Walgreens? They're obviously investing heavily into VillageMD. How many Walgreens stores are you in, and what could that increase up to? Just any color around those two relationships would be very helpful. Thank you.
David, you know us really well, and you know that we never, ever talk about clients' projects, and we're not going to change our way today. I would only suggest that our relationship with Anthem is a very important relationship for Amwell. It's been going for almost a decade, and it's very, very strong. We're doing different things with them. They obviously do other things as well. The digital activity in Anthem is vast and impressive, and we're very proud to contribute to it. I will not say anything beyond that.
Okay. Appreciate it. Thanks very much.
Sure.
Thanks, Dave.
With that, ladies and gentlemen, that does conclude today's conference call. We'd like to thank you again for your participation. You may now disconnect.