Hello all. I'm going to talk about how do we leverage sales data analytics to optimize the pharmaceutical drug launches. We'll do a deep dive on the technical analysis of using predictive modeling and real time market insights. I think we'll cover a few topics. I think we'll talk about, Standard, challenges in the, in, in the drug launches. Also, how the data analytics, can be used as a solution. what are some of the important KPIs do we use? also talk about predictive modeling and s impact. on, on those launches, we look at one of the case study, also we'll talk about, the real time analytics, and also some of the best practices. And the recommendation and then we'll look at the some of the key results and impact, starting with just giving us more background and challenges like why is such an important, issue to talk about, right? So to develop a drug, in the, in the highly regulated environment, it takes. On an average 2. 6 billion to develop one, drug and then 60 70 percent of the revenue Produced by this drug is achieved in first six months. So Think about the first few months of the launches are very critical a lot of products a lot of drug at their initial phase, they fail to meet Their sales goals and protection, and then typically around like you would see 30 40 percent, Below on the expected revenue. Also, you can see our development costs are also going, high every year so with that I think it's an important problem to solve and then We are talking about like how the data analytics how the sales data can be used as a solution So I think the very first case is, predictive models, right? So right now You When you think about pharma industry, we have very good models with more than 85 percent accuracy in forecasting. Also we do have a lot of real time data, millions and millions of data points for the patients, for the doctors, for the hospitals, which can be used for the decision making. Also when we are deploying the field force, a lot of AI model can be used to increase the efficiency. And then I think we have seen, when there are the launches which are more analytics driven, they have a higher penetration as compared to your traditional, methods. Also, we see, an improvement on the launch trajectory in the first 12 to 18 months of the launches. I think once you apply these predictive models, and real time insights, the KPIs play a crucial role, right? So that you know that, are you on track on the launch, right? Are you on track to get to your goals? As we talk about, I think the first six months are very critical, like 60 percent of the lifetime revenue is achieved in the first six months. so in that time, I think the prediction accuracy is definitely an important piece. We need to have 85 90 percent prediction accuracy, using all the advanced analytics methods. A market penetration or adoption by the early doctors. That's also an important KPIs to measure, by different teams. also if you want to make sure we're reducing the forecasting variance, based on the machine learning models. I think now maybe let's talk more about the, how the predictive models are using, what kind of models do we make and like what kind of framework do we utilize? So I think that the very important one is, the doctor adoption or the physician adoption. one of the very, popular models are doing segmentation based on these doctor behavior. And then there's a lot of data coming, right? There's like thousands and thousands of data based on these doctors, their prescription, their demography, their sales information. So these all can be used to segment these doctors. into different bucket, right? And then we can have a different tactics for each of the segment. Also, when we think about the real time engagement, that also are doing a very good accuracy. the use of these ML models, we are reducing a lot of variance when we think about the forecasting of this launches. there's a lot of different machine learning algorithm, used to analyze, and prescribe, patterns. you look at the engagement metrics. we look at a lot of patient level data. which is your transactional level, medical gains, pharmacy claims. we also use a lot of ensembled model, like random tree to make sure, there's a stability right in the prediction, and we are making sure that we are baking in all the market condition in these models. I think maybe, a case study, and this is not a real data, but, think about, there's a pharma company, they're trying to launch a product in type 2 diabetes, disease area, which is a very big area. Yeah. when you think about the market, like you're talking about, maybe 20 to 30 billion dollars of market value. And there are already a lot of competitors, there are already 8 to 10 market competitors, with very high market share. So when we apply these methods, when you apply the segmentation, you do a lot of real time insights, have the analytics KPIs in use, definitely you, you achieve a very high 72 percent prescriber reach in, in first three months, which is very important. Also the new patients who are starting on therapy that also exceed your target, right? So for example, in this scenario, we said, Oh, we want to reach to 12, 000 patients in first three months, but because of the use of all of these analytics tool, we are reaching to 14, 500. Okay. also, we are able to do a lot well in different payor channels, like if you think about commercial, Medicare, Medicaid, so these, and then I think that the other element is when you think about the territories, so you can do also a lot of AI based territory optimization, and then look at a lot of patient journey, like how patients first diagnose when they get treated, what is this whole adoption funnel looks like, and then because of the real time data processing, we are able to process a lot of thousands and thousands of data points across these doctors and the patients. Now, maybe let's pivot more on the real time, analytics infrastructure. as we talk about now, we have a lot of data elements across doctors. patients, payer, hospitals, right? And then to process these data elements, we also have, most of the pharma companies are investing a lot on having, big data warehouses where we have a lot of processing capacity which can, where you can have your analytics tools sit on top of that, right? We're talking about like processing capacity in, in, in petabyte, right? So a lot of real time data, very fast processing. and then with that, also all of these, I think the data nowadays have very good capture rate. So capture rate means how much of the actual data is captured, right? So when we get like a third party data, We can see that it covers 94 95 percent in the pharmacy side, and then these Claims or these transaction also have different layers of quality checks right to make sure That these are highly accurate when we are trying to use in the predictive models, right? It's garbage in garbage out So the data quality is very important So that's where there's a lot of data stewardship and data quality checks are in place When we are processing all of these big data sets And then with that, because of the real time market insights, it also allows us to respond in a very faster way, right? So we can have a market response as soon as like three days, right? When we see something and important insights out of data, we can, we can have a change in tactics in, in, in three days. Also the data validation, like the data you get is like 99. 8 percent accurate. So with the having a right infrastructure in place, you can process of tons of data also having a proper quality checks can give you a higher accuracy and that eventually allow us allow companies to react in a very fast time. Also, Thinking about the best practice and recommendation in the context of the product launches or drug launches in the pharma industry. We talked about that the data integration is important, right? Because you're getting data from different data sources, right? So you have like sales data, you have demographic, maybe you have the social feed like this, like tax mining. So all of these data has to be integrated together, right? So there's a lot of De identified or encryption decryption happening with which is done by a dedicated companies right to make sure there's this compliance and privacy when we are looking into the data for the doctors and the patients. there's definitely a very high compliance standards across pharmaceutical industry, right? it has to go through a standard, encryption decryption model where we cannot identify a certain doctor or certain, patient. And then also a lot of time when the data is not available, we also do a lot of data amputation, right? To make sure. That even for the missing data, can we predict what can be a possible data outcome? data integration is an important piece, when you're trying to combine all of this data together. Also, I think when you're trying to do, analytics deployment, either you talk about predictive models, or an AI model, right? I think a lot of times these models are expensive to implement, right? It requires a lot of infrastructure from the IT side, and a lot of investment from the business side. I think the best, I think a good practice when you're trying to do any new model, you can start with a pilot, right? You can start with, one smaller area. for example, you can start a pilot just for the California state and try to see, based on that, are you able to get, a higher response from the doctor or for the patient, right? And if you see, a good successful model and then the same thing can be applied for the whole nation, right? so this pilot implementation is very popular, in the industry. and then you can also always make finer tune based on what we see on the pilot side. I think the other important thing is we talk about the KPIs, but then, we need to have a lot of dashboards, for different teams, like sales may want to see a different KPIs, marketing may want to see a different KPI. So having a dashboard, which is updated more on the frequent manner, like weekly or daily, right? and then having an important KPIs, is an important thing, right? And then I think the. The other layer on top of that you can also we can also create a lot of alerts which are automated So we I don't have to look at the dashboards every day But it does create an alert right when it try when it identify an important pattern in the data also based on that, the kpi dashboard there should be a proper, resource allocation and optimization. So these are like some of the best practice and recommendations, especially from the analytics side and the infrastructure side when you're trying to launch a product. I think we talk about, and these are some very high benchmarking number, right? When you think about, When you apply these best practice, when you apply these analytics model and AI, you see almost 45 percent higher market penetration, right? Compared to the traditional model, your launch trajectory also improves a lot, based on the insight. And then real time data analytics allow you to response in a few days if there's any market event. Also, we have seen, an increase in the prescriber engagement, we increase in the patient engagement on these AI driven models. so I think, now I think I want to maybe conclude and I think what are the key takeaways from all of these things, right? a sales data analytics is definitely, and key piece, when you're launching a product in the pharma industry, it definitely have a measurable improvement. it gives you a higher market penetration. It gives you a higher Doctor engagement. It also give you an operational efficiency real time analytics Enable you to make like agile decision making and you can navigate through a complex Market tool like where you have so many competitors and so many market events happening at the same time Also, I think the advanced predictive models also provide like very high accuracy in the forecasting which Eventually help you to do a resource optimization Integration of these, the AI models, and I think, it definitely going to improve the model accuracy and the new AI tool will definitely help that, and real time monitoring, enhance the global survey across the market insights. Thank you.