The world has become digital first - what will it take for pharma to catch up?

Panelists:

  • Dr. John Reeves- CEO, conversationHEALTH
  • Dr. Andrew Rut - CEO, MyMeds&Me

Moderator:

  • Dr. Andree Bates - Eularis

Speaker  

Welcome to the webinar. The topic today is the world has become digital first, what will it take for pharma to catch up? And I have two panelists today, representing different sides of the healthcare spectrum. Dr. Rut will be focusing on the backend safety, the business side, and the regulatory insights, whereas Dr. Reeves will be focused on the front end, the customer and the physician and the patient. I'll introduce my panelists, I've got Dr Andrew Rut, who is the founder and CEO of MyMeds&Me, he is a trained physician, and he specialized in endocrinology and genetics, he has held multiple roles in drug discovery and development in GSK in Europe, North America and APAC. He was previously Head of Safety at GSK where he drove significant innovation and streamlining into the organization. Welcome, Dr Rut. And I've also got Dr John Reeves, who is the CEO and founder of conversationHEALTH. He is also a physician and a serial entrepreneur, Dr. John has been disrupting healthcare since he went into clinical practice in 1994. He is constantly looking for new technology applications to resolve key clinical challenges, and he firmly believes that physicians should lead the disruption of healthcare. He has over 20 years in digital health leading innovation at large organizations as well as startups including conversationHEALTH. Welcome Dr Reeves. Pleasure. So, with what happened last year, it's been said that the digital transformation has really leapt forward by 10 years in one year, with consumers first forced and then choosing to go digital first with both their personal and their commercial activities. So consumers, patients and physicians, expecting to source medical information but also to report side effects, product issues and other things through digital means. So we're going to start this panel by looking at how can technology help address the needs of pharma, and I thought I might just start with Dr Rut because we've done a webinar on what happened with the vaccinations and the adverse event reporting so why don't we just start with that perspective first and then we'll move on to Dr John's perspective.


Speaker  

Yeah, sure. Andree. I mean, I think, as we said in the previous webinar for me it's all about data, and you know I think Dr. John and I will probably resonate on that point is, whether you're in trials or whether in the post approval phase you need data. And what we've seen during the pandemic is getting robust data, whether it's on the COVID products or other products has been, has been challenging, and where companies have had the data. And I would say I would point to those companies who've been really transparent about their data, they've stood above the crowd and their products of their vaccines have actually been call it trusted and well received, where there's been any sniff, whether it's real or unreal of data not being completely open and not being collected in the way that the protocol might have defined people have, people have have questioned that. So I think for me, you know, we've moved from a world maybe five years ago, Where trials were perhaps much more paper based we're moving into a world of much more accelerated trials, but then that pressure and maybe this is what we'll cover that puts huge pressure on the market authorization holder at the time of authorization, because those commitments then have to fill the blanks in terms of the efficacy and particularly the safety profile at the time of launch are huge. So we've got actually a very narrow boundary of knowledge at the time of launch and that boundary of knowledge has to be filled very quickly in the post approval phase and we can call that real world data or whatever else. But if it's not filled appropriately if the data is not collected appropriately, actually we've got a real problem. And I think that's where the digital first part will come to the fore, because we can access robust data across large numbers of people very quickly, and actually make assessments of how those products work across a much more general population that we add in trials. Maybe I was thinking about this webinar and I'll hand over sort of shown after this is thinking about the digital first part so yes I think, most patients, and most consumers now expect their healthcare to be digital first, whether that's in trials or in the post approval setting, and I think we all believe that they want to access that their healthcare or information, you know 24 /7/365 In an accurate and trusted way. The second part I would say is physical fast and as a physician is very difficult for a patient to have a rectal examination virtually. So I think the first part is, let's play digital where it should play, which is providing access to healthcare information to ways in for patients live, you know, essentially, in the digital space, while enabling the physical care, whether as part of a trial or real world study or real health care to occur as it as it should do and make that, you know, focus physicians and healthcare professionals. On the pieces they should be doing. And I think that's where the digital world can help them stop.


Speaker  

Well, just to follow up on that. So, with technology for adverse event reporting, what kinds of things can technology do in that space.


Speaker  

Well firstly, it brings the reporting to the patient, I think that's the first part, so I think then, you know, if we're taking a clinical trials scenario if we take COVID or even accelerated programs for for oncology products or companies want is accurate data, A total picture of their product as accurately and as fast as possible from from the patients, there's ultimately a patient's experiencing the adverse event. And at the moment, even today, typically, even, even in our call it is a decentralized trial model. Patients typically provide their adverse event data at the time of a visit, whether that's a virtual visit, or a real physical visit, and that's just nuts. And that data then has to be sifted through by a healthcare professional reported to a pharma company within 24 hours, and the patient may have experienced those issues a month before. So actually, there's two pieces one, the data is inaccurate because recall is inaccurate. The second part is there's a lost opportunity to collect data that's time limited, as most of us know that it's pretty difficult to recreate a set of blood results. If the events occurred a month ago. So, you know, we've actually lost an opportunity to ask relevant questions to collect relevant data around the time of the events occurring, and therefore, I think this is where, you know conversationHEALTH come in, and therefore to actually communicate with the patient in a way that's meaningful to maximize their adherence to minimize their, their adverse events to minimize their risks, and potentially intervene in terms of reducing dose or taking them off the drug. So I think those are pieces within the clinical trial setting that can be done in the real world setting exactly the same applies where, actually, you know, currently we know that, probably. I mean if we're on a good day with a fair wind behind us about one and 100 of the events that occur to patients are actually reported on most products, and it's probably one in 1000 for all the products. So we're losing data, big time so we have a tiny amount of the real data coming through. And I think digital tools can make it much easier and much more intuitive and much more rewarding for patients to provide that information in real time and get feedback about what's happening and an information that helps them manage their disease and manage the products that they're on.


Speaker  

Interesting. and Dr. John, could you talk about how technology can address the needs of pharma from a customer centric point of view.


Speaker  

I'll also just connect back to that concept of digital first, so there is you know and this transformation that's taking place because of COVID. I was just thinking a little bit earlier that you know, we talked about it, compressing 10 years of digital kind of evolution into a year, I don't actually know if without COVID, if we'd actually evolved to where we are now because some of the transformations are just a fundamental transformation that we don't actually have to see physicians, you know, face to face for so much of what we do. I don't know if we actually would have overcome the friction of the way the system was set up had the pandemic not taken place. I mean, obviously that's driven massive thinking about, you know, the digital first experience our customers, consumers, patients, physicians have always, not always, for the past, you know 10or 15 years have been Digital First, we know that there's a billion searches on Google every day for healthcare information, some of those searches are by physicians right they're looking for information. The question now was was there going to be a transformation in pharma from moving away from the concept of the push model the heavily vested into the human touch point in more than digital touch point and I think what we've seen is that there is a new capability, a new capacity through the virtualization of healthcare to achieve many of those business objectives, but to do it in a way that our customers actually want, I think, you know, you know, I think, probably, Andrew, you know would would agree that, you know the magical moments in healthcare typically in the past had been, you know, that kind of face to face experience something magical happens there. And I think that the opportunity we have now is to make sure that those kinds of experiences these human like experience and sometimes they will be aI powered actually don't take place at that episodic moment, but are more at that moment of need. And so it's really about that fundamental transformation about going to a pull model than the push model right so digital is inherently an on demand pull strategy, right, you need to be there, and you don't, you know, customers don't come to our digital touchpoints, to be marketed to. It's just not the way it works right it's more of a PR channel, you need to be there to answer their questions first. And so what we're seeing is that the experience is digital first 24 Seven, on demand, all of these things. But I think what Pharma are realizing is that if you can be there at that moment of need for your customer and answer their question that they will give you them the opportunity to communicate more commercial messages that are contextual and the, the, the, this comes to reality when you start looking at COVID What are the demands on one of the new, you know, let's call them vaccine manufacturers to respond to inbound conversations or information, it's overwhelming. Right. And so one of the, you know the key products we've been working on in q1 is actually deploying a conversational experience that allows you to handle you know millions of inbound queries or questions from consumers, patients and physicians about products. Okay so, so literally I mean we've gone through a transformation now where every brand every pharma company has to realize that that is the starting point. And if digital is the starting point. It's a new way of interacting.


Speaker  

Yeah, well is the industry ready is the next question.


Speaker  

So, let me build on a little bit more the, the pharma industry is sort of trying to get ready and I think we've got the privilege of working together with across our two companies and one of the, one of the outreach programs we're running is with with organizations that are all pharma companies that have strong consumer healthcare portfolios so direct very direct to patients where people self pay. So where people self pay they tend to be rather interested in what happens to them, and particularly where they self pay for their medicines without a medical professional intervening. And you know, if I look at something like the Abbvie website which I mean, call it in the Zoom world now focuses on a product you all know well is Botox. They are really reaching out to consumers directly and using every marketing tool possible to, to market a product that's you know, a medicinal product in a way that's very different to what we've seen in in the traditional pharma world, but yet that product is a pharmaceutical product. And if you look at how it's administered. Obviously people now looking carefully at who they should choose to administer that product is no longer, You know, just go down to any old clinic and find somebody to inject and, and I think that's a key thing and then you've got counterfeits you've got maladministration so there are all sorts of issues, but people want information about that product, want to know who's the trusted provider, or to know what doses, they can get how compared can be used and obviously it's used for other, you know, more, more, medical, rather than the aesthetic purposes as well. And I think that epitomizes where what John was saying is that 24 Seven, access the pool where people are wanting that information all the time simultaneously obviously Pharma then can provide wealth, less suffering in that consumer world, messaging, but in the medicines world probably slightly more subtle marketing messages or call it adherence messaging that enables that patient to trust the brand and stay on that brand.


Speaker  

Interesting. And then, when we're talking about digital first with Pharma. What are you seeing what you know what do you think, Pharma, are doing in the digital, you know, how, how widespread is digital first across Pharma.


Speaker  

Probably I'll speak from the conversational AI perspective so far so first of all pharma, it's a great industry now that's accelerating very quickly into digital use obviously when you lose key channels to your customers you need to quickly adapt and they've done a fabulous job at that I would I'm incredibly impressed by the work that's been done in the last year, in terms of moving to digital, in our space you know I think one of the questions and I was around for the first couple of horizons when it was web and when we went through the kind of the social nap horizon.


Speaker  

purchase these new capabilities. I think one of the good things is that because we've been through a few digital horizons already. I think the way they approached new, new capabilities like conversational AI are formed far more strategic and measured, and that they're doing the appropriate kind of POC, they're doing all the right things, and I am really encouraged by how quickly they're recognizing this as a new capability. The other thing I would say is that, you know, historically in the past as we move into a new capability we tend not to understand that capability and leverage it the way it should be used right we use more historical models and how to use it as a channel, I think one of the things that we're seeing very quickly is that pharma understands that a conversation is different than content right they're rooted in conversations they have conversations that are taking place 24 Seven, through their human touch points, and then what they're realizing is that it's not just about a technical capability, it really is about a communication capacity and understanding how to best leverage this new channel to drive all of the things that Andrew was saying it's for everything from awareness right to to product start to compliance with all of these things are critical components of a communication strategy. And so it's really not that different. When one thinks about it from having a conversation with another human right.


Speaker  

It's all about understanding basic UX principles, and I think that just as in general, to sum it up, the industry is has great digital capability and I think that they're bringing that into this new horizon of conversational AI, and actually using it in a very appropriate way.


Speaker  

Interesting. I did read something in CS Pharma recently that, and it also with Accenture, there's this very interesting divide at the moment where physicians are very much into fully digital or fully face to face but not into the hybrid with the records on the Zoom call with them. So that seems to be the seems to be a lot of that but the efficacy of the rep on Zoom call has just gone very low, according to faze Pharma, and most or Accenture was saying I think something like 87% of doctors are saying that the first fully digital, which was interesting, there was another study saying either fully digital or fully human, but not this human on Zoom which is interesting, maybe it's zoom fatigue recall that I guess, but probably on that note, I think that you know not all doctors at the same not all consumers are the same. We have to in many ways, let our customers, self select the channels that they want. I'm actually a firm believer in more of a hybrid model, that there are two kinds of all of our customers want to engage with a human right. But there are also times when a digital touch point can be far faster and provide certain benefits. So, for example, as a physician, the odds that I'm going to call in to a med med info line, to ask a simple question about dosing etc. things that, you know I should know but may not know, I'm not gonna do that right, but I will use a virtual agent to respond to those queries, right. And so let's call these questions or interactions that sometimes make the user feel embarrassed or stigmatized or judged, right, and so it's really more about understanding what the customer use case so for very simple things. Really easy to grab. You know and resolve these through conversational agents and by the way, as these solutions mature, the complexity of what they can respond to is always increasing, but it's really important to understand then when do you hand these off, you know, to the human and allow the human to step in and provide that extra level of of information and service.


Speaker  

So I suppose they get in some compensation for your compensation I can basically handle the whole thing smoothly from end to finish, and in other cases, maybe there's something that's really just needs a human input on it. If it's yes or there are times based on so when you really understand what the question is from a customer. There are times and adverse event would be a great example. When you are going to move it beyond the the AI capability immediately, so there'll be certain, certain intense, that as soon as you understand what the customer is looking for you may immediately move that to a human interaction. So it's up to the business side, to determine what they want to do what's critical of course is that the AI has to understand that entire portfolio of conversations that are coming into the system. Right. And just as an example, we kind of let our clients know that there could be up to 300 kind of core conversations that you could have with your customers. It's now up to you to triage and understand, create a policy model by which ones are answered by a rich conversation, which ones may be escalated to a human right away which ones may go into an a detection response solution. And sometimes, you know, you're simply just giving simple answers right not every AI needs to be a rich human like experience sometimes you just want to confirm a number so that you can move on in your clinical day with patients in front of you.


Speaker  

I mean I think that, I think we're focusing on the two areas that John and I work on which is I think the front end communication with the patient or the healthcare professional that John's team does really well on the capture of call it negative data in real time that we do.


Speaker  

I think the other part is, we're all part of now a suite of technologies that operate in, call it the telehealth or decentralized trial landscapes and all the rest of it. I think the only the only thing I would add here is that, you know, it's great to see that because I think it means that the pool of patients and the access patients have to meaningful healthcare or meaningful answers to their questions is much better than never was. The downside is that trial integrity is at risk and I think everyone who's running trials now, whether they're pre or post approval is, is trying hard to, to assure themselves and probably the regulators that those trials have integrity. And then the second part is obviously that people are playing with every single organization we talked to you now says well you know can you connect with wearables, you know, yeah, okay, well, what's the what's the problem the wearable is trying to solve, you know, is it making the person's health better is it worrying them is it giving them diagnoses which are meaningless, you know, It's like the Apple Watch and the cardiac rhythm disturbances, I mean it's another piece of technology which yes it may evolve but right now it's not a point where it can be mainstream. So I think you know there's a, there's a piece around how do we augment what humans do well, which is really important, which is where the conversational AI and the capture of data in real time is really important.


Speaker  

But that doesn't, that doesn't negate the human interaction that's needed I think John is highlighting that there. And I would say it's like a physical examination, the laying on of hands, is still a fundamental whether you're in trials or post approval, whether that's because the person has, has a, you know, is having a problem with their medicines or that they've got a fundamental problem with their disease that laying on of hands or the human contact is really important that contact can be initially done can be triage to a virtual human concert as John says you can do it through a virtual assistant and ultimately flag it to human at some point, there is a physical interaction that's needed and I think we just all of us who work in that digital space needs to be mindful of that, that at some point, we hand off to a human, that is physically going to see a patient. Now that may or may not always happen in trials but I think at some point, people want to know is there a real patient at the end of all this.


Speaker  

Yeah, that's a great point. I just have to give double click on that, I think it's about using humans out to their highest value, right. Yeah, to your point, I mean, if we can, you know, remove some of the, the busy work that a physician is doing during history capture etc. you know, can we get high fidelity information that allows people to focus in very quickly, even when you're moving things within the call center system right. The information that's being provided in advance will allow humans to get much faster to where they need to get at the endpoint. So it's really about efficiency. I think that that all, you know, participants in the system are realizing that now.


Speaker  

Before with Android, didn't worry about the adverse event, call centers being overwhelmed. Yeah, so, so exactly that and I think there's two parts one is the call center is overwhelmed with busy work but I think there's another part which is the accuracy of data so the more Trent, the more transcriptions the more handoffs there are, in terms of patient being handed off to various people whether it's a call center. It is somebody working in a data processing hub, they're often they're not with the patient, I mean what always distresses me as a physician is when, you know, I mean, let's get back to you know I can talk about an area so drug safety or we call it medical coding of of anything that occurs in trials or post approval where analysis needs to be done on that companies tend to employ people to code that data and structure the data because it's been, it's coming in in this, so they either use humans they use AI, whatever they, whatever else, to structure downstream so they're not focusing on the upfront, sensible capture of data in a structured way, they're focusing downstream to clean that up. Those people are completely dissociated from the patient. And I always say, so how can a person who's a data scientist sitting in, let's call it Philadelphia or, you know, in Eastern Europe or wherever, make an assessment of what that patient really was trying to say, they come and they say well we we employ medical experts, or you may employ medical experts, but then with the patient so that they're making an interpretation. So I think one of the things I think both, both are going to both our organizations feel strongly about is, why not front end a lot of what we do, and have that interaction as meaningfully as possible in as real time as possible with a patient or a consumer that enables them to give you that information or have that communication, and then ultimately, people downstream can have got opportunities whether it's because things are flagged to them or they've got data which is actually a lot better than they would have received otherwise.


Speaker  

And so, the way a lot of things are done in pharma out in the, the way it's the incumbent processes are much more human centered. What do you both think it will take to change that to becoming a more digital first process. Yeah, I mean I'll speak to this from a conversational side which is I think one of the reasons why there's less friction is that when you build conversational solutions to engage your consumers, patients, physicians, whoever that may be. It's not that you're creating net new content right Creating content is a massive barrier. In this case, you know the content exists, whether it be on websites or other digital assets that you've created, you, you have content or responses that have come through call center, etc, etc. So really the challenge that we look at is how do we conversational allies, your existing assets, and literally transform them into something that your customers want to engage with right. So that's, that's kind of the key thing is that the barrier isn't about crafting and creating net new assets in most most cases it's really about just changing it into a different interactive model, right. And so that's really really important that we've seen and so it's not like we're starting from scratch, where, you know, we've got the football on the 75 yard line and now we're just making it far more meaningful and impactful for your customers, which is what they want, I mean, no one goes to their doctor's office, to just receive a brochure right they go there to have a conversation. And so it's a similar model here is you know, and of course, you know when you can start to move your customers away from just reading content and into a conversation then you open up all the capabilities of what a conversation allows, and in many cases, and I'm trying to do is okay, you know, as experienced as many many times when a patient asks a question that's just a lead into where you want to take them right.


Speaker  

In many cases our customers don't know what they don't know. And so conversational experiences now allow us to look at the next best action or the next best conversation so when they ask a on the poll. What do we now wishes the next 234 conversations. And by the way because it is a conversational experience maybe we can reach out and initiate a conversation with them on day two or day four, day five right so it's this new capacity, that really though is leveraging existing assets and content and insights.


Speaker  

I mean one of the, one of the people I follow, was, was looking at the whole digital landscape I mean beyond what we're describing and I think he was looking at his career and it sort of mirrored some of my experiences in the past is so, so, back, back in the day when, when genomics, became the latest hot thing like digital is today. Everyone believed that you could identify the right patient you know basically the right drug for the right patient at the right time as dementia, you'd know their genetic profile, you know which drugs going to work in them, you'd know their side effect which who was going to get side effects. Well, it surprisingly we're not, we're maybe just about scraping the surface in some areas of moment and genomics is just about, it's just coming into more mainstream given the cost, the cost of the technology has come down dramatically so you can sequence, a person's genome for about $1,000 whereas in the past it was probably a million.


Speaker  

So it's taken a long time and I think the, the, I think we also need to be mindful yes digital tools will we can deploy many digital tools today we're showing as John has described is, how do we make that interaction much more meaningful. Today, there are other digital tools which I think require a bit more thought. And I think lumping everything digital into the same box is not right because I think there are tools that need to be validated. It's like John was saying you know we've got medical content here which is readily readily accessible. So let's use that the tools we have for conversation, are there some of the other tools on the periphery, I don't think are ready for mainstream and we need to be upfront on that and say okay, when we can try them. Some of that is trial and error and we need to work through that. So, so I'm you know I'm a passionate believer in, in making is, is creating opportunities to connect with patients, so that the data that they can or their, they can share experiences and data in a very simple meaningful way, that can then be used for robust analysis downstream. There's only then, once we've got that data downstream can we continually assess how those products are working across the populations they won't work the same way in all of us so we can begin to tease apart what are the intrinsic and extrinsic factors influencing responses to the medicines that they're taking, and in influenced that in a positive way, that the alternative which is I think we're Pharma focuses at the moment is, well I'm not going to bother trying to get really great data at the front end and really connect with the patients or doctors who are managing those patients the front end. What I'm going to do is put a massive effort into structuring the junk that's coming in, in the middle, and then put middleware in that's going to structure that junk, And then hopefully my analytic teams will be able to make sense of the junk that I've structured.


Speaker  

I mean I think however many times, science has tried that approach, it tends not to work. So, so I'm a firm believer in yes you know you, let's do the best we can, of creating the best possible and most and highest value data flow from from patients, recognizing, not every patient can communicate everything digitally, but we can, we can provide them assistance, whether that's through an AI system or a human, human, Assistant, that they have the same digital tools at their disposal that enable that data to flow seamlessly to people at the back end who are then going to make meaningful and assessments of what that data, what that data is telling us about the way the products are working or causing adverse issues with, with, with those patients. And I totally agree and I've seen pharma companies spend so much money just getting all their data into data lakes and cleaning and scrubbing but a lot of that data is old and out of date and actually completely useless and pointless doing anything to. So, particularly now with the way the world has changed in the last year, you know, a lot of different you want up to date data that's, it's right there. So, how do you both see how pharma can balance, driving this change to become digital first but setting, practically achievable goals, How can they balance those two things.


Speaker  

Well, I think, you know, one of the things I've definitely seen over the last year, two years with conversational AI is taking a much more strategic approach as I mentioned earlier, right so it's new capability and so what we see is that different, or each business unit can use conversational AI in a different way. Right, it's just, it's a new capability that is relevant across the enterprise. And so what I think we're seeing is just a very structured approach and typically you'll find that the look, initially at areas where they there's great cost savings right considering as a critical determine in terms of investment and so you see a lot of this automation of call center, where you have expensive humans coming in. And then, you know, over the last, you know, year, six months, we're seeing a big movement towards more the commercial side of the business right and how do we, you know create virtual Salesforce solutions, how do we take our E details, all of the kind of commercial things, the brain, websites that they've been building, how do they translate those from being kind of the push model to much more of a pull experience as conversational. And then of course now you see this moving downstream into clinical trials, etc. So, one of the things that we've seen is this kind of ability for them very quickly to get in and do PLCs, what do they want a PLC with a new capability. Well the first thing they test out is can it actually work, Right, an AI actually understand what someone is writing to you or speaking to you, and then understand, you know how to, how do you deliver a conversation to them, that's step one right to see AI NLU challenge that you face which is fairly significant in healthcare where lexicon that we use is different than other verticals right, step one is done then to you kind of figure out whether or not your customers will actually find this experience to be valuable and positive right and so what happens and again informer, they did a great job is, get over those very quickly, and you know the world is to a point now where you do a great job but those two are something we've been able to achieve that and then it's really about understanding but how do you apply it then across the enterprise, where does this new capability work across, you know men Affairs has worked for us commercial, where does it work in, in all these areas, and then start to test them out for each one right not every brand. We use conversationally II in the same way. Right. Obviously mature brands and totally different use case, they can all use it but the use case in the investment is completely different, right, so we just been, you know, really positive in the way that's being approached, not from an ad hoc, you know siloed approach and one of the key things I think is because conversationally, it is more of a platform play. It's recognition, quickly, quickly that this is going to be something that will can be applied across the entire enterprise. Right, and so it's not just being done quickly by silo myself, that they realize that data is ultimately what powers this taxonomies power this and every form of company needs to go in and establish that is a foundation, and then figure out exactly how to invest across each one of the business units.


Speaker  

Yeah, I agree. No, I would concur with what John's saying I mean what we found this successful is where companies tried to do meaningful assessments of a specific piece of work, so those that are you know, the offer word of boiling the ocean if somebody wants a fully fledged solution on day one that does everything. It's just unrealistic I think most companies that succeed take a sensible stepwise approach and say well we thought this through. We know what our end goal is, and we don't quite know how we're going to get there, but you look like the company or the partners with whom we can work together. So let's use an iterative approach, let's start with something we can see you can actually do, that's number one is that the amount of smoke and mirrors that occurs in client demos is quite extraordinary. So you can say, right, you've actually demonstrated something that works in his life and he's actually in use great, right, why don't we try and test that out within a specific part of a specific region specific language. Once we've got confidence that that's moving ahead and actually the team then have a better understanding of what the business requirements are, we can move to the next level. And that usually works really well.


Speaker  

Have you seen examples of companies that try and do much more than they can chew and then. Yeah, we all have, I mean, companies do, and sometimes it's not. It's not done for bad for bad intent, it's really done often because there's an unrealistic view of what is possible, and often that's driven by organizations that they've seen or talked to who've come in and said, You know what you can have the earth, moon and stars in the car, not only can the car drive along and it can go through the city and go goes up hills can fly, and, you know, you go, Well yeah, and yes it can now be shown occurs you know it's possible, and I think the, the, the, I think where we've seen companies trip up is where they come in with a set of requirements which go well, how fast the car can the cargo. Does it have air conditioning, does it have 20 inch wheels. It does have a via engine or doesn't have an electric engine does it have to have flip up does it have also this and you go what are you trying to use the car for, you know, are you, is your problem to go around the city by yourself, because you want to commute into the city, or are you cashing in the family on a camping trip while you're a Pharma who has to carry large amounts of hay to to their cattle at the weekend, you know, during the week, and the Oh, right. We haven't thought through what our business use case we don't want to know what the car can do. Well, yeah, well then you're going to have, you're not going to get product that actually meets your requirements. That's the biggest issue I see all the time with Pharma is no I will go to a company and they'll say we want AI and I'll say okay what do you want it for. We don't know. Can you find the use case. And I see that a lot. I also see it being led by Tech where they buy the technology, but they don't. I've had that time and time again where I've gone to a plant.


Speaker  

Okay, great. Why didn't you find out what we should do with it. I always keep saying exactly what you just said, you know, start with your business problem or objective, and then figure out where you know what's going to be the right solution for that. But yeah, I said, the cart before the horse so often.


Speaker  

Double click on that I would, I would simply say that you know some of the challenges, Sometimes he is that some of our clients want to move too quickly into AI and so moving very quickly to voice or digital human solutions that would take an initial step which is foundational kind of text experience. There's so much that gets learned in that first foundational build not only in setting up taxonomies and data structure, etc. But it's super sexy to go to voice, it's super sexy to go to digital and those are great. Absolutely amazing opportunities for Pharmas specifically right because, you know pharma lives in a world where human facial expression and and voice experiences are very very important, but they just need to come at the right sequence I would say that that's what I would also say that you know, as a new capability conversational AI is something that every part of the enterprise if you show it to the next place and will say Oh, I have a use case for that. Why because every brand every part of the business wants to have on demand, you know, human like interactions with their customers, right. The question of course is which ones do you do first, right, and there because there are so many you know, challenges in the business that can be overcome, but it's not, you know, it's really about defining what is the first use case the first experience. And as I said that one shouldn't be predicated on up your business ROI, because sometimes a use case. Isn't that powerful, But what you're putting out is the ability to set the foundation to set up your taxonomies your NLU right your machine learning, all the things that are critical to the ultimate success as you deploy across the enterprise. So that's super important. By the way, just double click on that you know we had one of our clients in the last six months, you know, instead of actually building something upfront, they actually said let's go and look at this from a more strategic approach. Let's look across all the different use cases. And as an enterprise let's put together a roadmap for what we're going to actually do in horizon one in horizon two in horizon three across all business units that we're actually making all the investments in a very smart way that we're not racing ahead to something, a channel like voice like additional giving too early, that we're applying the right solution so like business units, and you know making the right investments at the right point and of course making the investments that are actually appropriate so as I said, conversationally is founded on having a great, you know enterprise SAS platform that once you design and build it can roll out for every brand, every customer group in every market in every language right that's the massive power around conversational AI right conversations are consistent in generality or, you know, in 90% coverage, or in every market. And so it's really important that you set that foundation, early on that this is where we're going to go as a business right, we'll want to be able to create conversations that we can use, you know, in all markets we want to have the capability of doing multi language, all these things are just foundational.


Speaker  

We see language is one of the key barriers so companies can easily jump into we can do it language, reading, language, I think once we get into AI multi languages, it's a real challenge, And we've seen this over and over again as most AI engines don't do well outside the English language, and Japanese obviously a key language performer. And that's one of the more challenging for for AI components I think a phased approach generally works well because you can then address some of the glitches, early on, and read through that, and then go to your next set of priorities that will keep you at night and you can't do anything. But why do you need to imagine that there's an order of diminishing returns in terms of some of the some of the markets at the moment still have relatively here still there's still relatively low penetration so actually having every language is not relevant.