Empower Your Life Sciences Business with Leads and Conversions for Life Sciences

How to create outlook signature

hello everyone this is Dylan Carter with rapid miner like to welcome you today's webcast maximizing lead conversion success using predictive marketing analytics we're joined by our speakers today from an Blix a rapid miner certified partner mu decir has son as well as nakida Naidu or data scientists and an Blix if we've got a good agenda in store so I'm going to start off with introduction to rapid miner for those of you who are new to what we do and then handed over to me - Sarah is going to talk about challenges that customers face in marketing well then talk about processes for how to use machine learning to overcome some of these challenges particular focusing on customer acquisition nurturing customers and then ultimately converting customers into client well then hand it over to Nikita who's going to walk you through some demos so you can see it live in action and a wrap away with some key points for consideration so rapid miner you know our core focus as a software company is making software that helps analytics teams be more productive right and we do this through an Opel an extensible platform really focusing on if you think about sort of that lifecycle from preparing your data and that includes you know blending data sets joining data sets cleansing data sets to actually applying the machine learning to ultimately deploying models in production rapid miner has a you know many of you on this phone are part of the user community so you'll find the worldwide users who are contributing to this open community to help really drive knowledge and then help you know help each other get more value out of data science and that innovation you know think think a lot of you who are on the phone today to help contribute to that the end of the day it's about driving you know driving business outcomes and so predictive analytics is and machine learning can be applied for revenue challenges so to help improve revenue help you know obviously save cost and more importantly mitigate risk these are some of the use cases you'll find rapid miners use as a data science platform yes again think use the user community very strong you'll find universities using rapid miner so many of you may have used rapid miner University as well as a good client base what I would encourage you if you those of you on the phone again who are new to rap and minor and want to know you know what do in users say or what does the analyst community say you'll find that Gartner who covers space has has recognized rapid miners a leader for the last five years and they're Magic Quadrant Forrester has also recognized rapid miner as a leader for the last two years in predictive analytics and machine learning so that you know reach out to the analyst and converse with them to find out more you also find a great resources KT nuggets for rapid miners recognize but also a great resource for just you know data science topics in general for those of you who are learning and then another source to really find out more is hey what are other you other end users like yourself saying and so GG to crowd you'll find a rapid miner really strong endorsement but she gave you some confidence that you know others have tried out the tool in can be used for your data science challenges so things you'll see today as the team goes through and the we you know the helps sort of differentiate rapid miner it's again it's that speed stuff that you do over and over from a data science standpoint that's really what we're focused on and so things like improving your performance you know making an easier to collaborate among analyst teams those are some core some core capabilities you also find the platform is very open so from a technology ecosystem standpoint easy to integrate with other technology that you may be invested in and it's extensible so if you want to plug in your own machine learning algorithms if you want integrate again with other technologies rapid miner provides the hooks to make that happen and again it's it's it's one platform right so not having to jump from tool to tool but really from a data science perspective you know being able to cover your product we need model validation needs and ultimately putting those in production so hopefully some of that will come across in today's webcast and with that I'm gonna turn it over to food is here thank you didn't hello everyone thank you for joining with us today my name is Melissa I work at endlich's a data scientist and in this webinar I'll be talking about the current challenges in the marketing industry how machine learning and avid miner are revolutionising this area and also take you through the different concepts of predictive marketing analytics I am joined by my colleague Nikita here who will further take you to a case study to help you understand how this is done using traffic miner so coming back there was a recent study done by deunan Bradstreet for the global marketing teams where they highlighted top three challenges in the current industry all the teams who participated 57% of them say that they are not able to understand their audience 46% are not able to crack the ry and off after all the marketing efforts they say just 17% or daily its convert right now there are so many popular marketing automation tools in the market but still they are not able to fully understand their customer our track ry even as a result a lot of leaders gather ground for which we need big data tools these challenges can be addressed by developing a relative marketing analytics solution using rapid miner as a defense platform which helps you analyze the data to understand your audience and tackle other challenges as well we'll see how this is possible in the next slide so this is how a customer journey looks like I've simplified them into three different stages to show how data scientist can new science can be used to improve the output in each stage in the acquisition phase data science can help you easily understand your customer and you demand channels that are working well in the nurturing phase it can Taizo needs an determind products that are more relatable to your customer and in the last conversion phase it can help you reach out your leads with a discount offer is highly likely to accept so we'll deep dive into these concepts and see what they have to offer the concept of segmentation a part of acquisition phase slices and dices your audience to simplify your understanding of the customer base for instance if we take the image on your screen as an example for a mobile phone company we can we can understand that there are three basic personas in their audience the first one is a business user and the second second persona is the casually the rule may be uses the phone for the sake of it and the third one if the power user who uses the entire feature is sort of form so this way the mobile phone company will know who their audiences and they can reach out to them in a very personalized wave and also chart it from marketing strategies channel attribution also a part of acquisition phase gives you an all run perspective of your customer journey it essentially tells you what channels are working well and give you an idea of where a buy decision being made in the customer journey in the nurturing phase lead scoring is a methodology that helps each other pick the top quality it's by assigning a school to each profile through its learning on the path behavior this way you will be investing your time efficiently with leads that are highly likely to convert moving on product affinity also a part of nurturing phase gives you an insight on what products are close available to your customers it gives you an outside perspective of what products can be bundled together for sale hence helping you chart new and intelligent product strategies this called recommendation the last the concept comes in the last phase of the customer journey it is d it is that last chip you need to push a customer to an end the fundamental idea here is to furnish a discount which is a monetary profit to both customer and the vendor so now my Nikita will take us through a new scale from one of our clients who are looking to improve their marketing efforts and also improve their customer journey she will take us through all of these concepts and also show how the insights from each of them help our client in every phase at the customer journey so how do you wanna get down thank you for the introduction Madison hi all this is Nikita ever presently data scientist with Alex and I will be taking it to one of the use case which is based on one of our clients who's an educational consultant based out of UK and give it a white comprehensive education services to international students to prepare them for graduate and undergraduate programs in a cake I mean Chris's objective was to like so nicely conversion rate and minimize sales and marketing cost and one of the biggest challenge for this client was in prioritizing their lead pipeline they had a very huge number of Nick community systems sometimes as high as thousands of broken teeth on part a basis in the peak season and it had become nearly impossible for recruiting agents to reach out to all the leads in the system they experienced very low lead conversion rate but they also had to understand what kind of products will this content center should be operated these leads so we followed a consultative approach we first understood their business model and business processes we broke the problem into multiple stages corresponding to the stages of the sales funnel he identified the problem areas which required enhancements or improvements and came up with his case which would have direct impact and the lead conversion and marketing in sales cost so let's deep dive into these in these cases this is the solution architecture and the client basically got leads to distinctive channels online mode and offline mode online mode is through the digital channels and offline mode is through third-party recruiting agents they use market before they use market marketing automation tool and all the marketing and nurturing campaigns were created in and through our kettle lead activities were captured in market or activities like emails and email opens resources downloaded web page visits etc cotton market oh this was a very useful information which we use for understanding the lead behavior for the day you Salesforce as the CRM tool to capture to run their sales and engage sales and lead engagement related business processes we use rapid minor studio and rapid - server to create and deploy machine learning models onto the client system market to read Latin minor read data from Marketo salesforce data warehouse and trans machine learning models some of the models were like heat Corinne who runs on a daily basis through cron schedulers but that we use tableau to summarize business insights and create infinite dashboards and reports which were used by recruiters stakeholders and management so let's take a look at the data flow so the gang gets a request from two different sources our leads come through digital channel and third-party recruiting agencies so theta related to demographic at max leader activities etc is captured in it captured in the CRM to a dump of that is taken into data warehouse this data is not clean because the forms are hosted on multiple platforms in some of the homes are not even in the English language so this data was not fit for modeling we had to do data preparation to make the reporting and reporting to give you a start on the size of the data we use two years of historic data for modeling and it had around 500 clave 500,000 lead profiles with 150 million line items of the lead activity we had to aggregate these lead activity data and create a summarized create a summary summary for each of the leads based on the activity of a half of whom throughout their journey and this data was spread across 60 or tables with thousands of attributes and it was not consistent so we had to clean that up and create a consistent model which was saved again in data warehouse with our customers kheema this whole data preparation step has taken 189 days we had three data scientists working full-time on this project at the pickle days to do the data preparation stuff so let's take a look at the use cases one by one we started with segmentation for us because it was important for us to understand the need personals or what kind of need profiles are present in system and if there is any pattern where most of the lead is there any profile where most of the leads are converting or any other way that is not converting so we used k-means for machine k-means or distance based clustering algorithm for creating the clusters and we used our rhythm sum of squares this metric to create find out optimal size of the cluster and we found that five is the optimal number of clusters and we have used to lead activity data in the demographics data segmentation so let's take a look at the rapid miner process so this is the data used for segmentation and each row represents a record for each lead activity and you can see that we have summarized the activities performed by the leads across different activity types so their journey some of the attributes are coming from lead activity data and some of that tributo denies bias that is like some attributes like agent the final recency meets frequency needs backup touchpoint etc mint of touch points here represents how many different channels does the lead use in order to interact with the client so let's take a look at the process but this is the rapid miner process for the segmentation and I just ran this process to show you the results so we are using a cluster model visualizer to summarize tetras choice here reading the data and selecting attributes using PCA which explains 80% of the only those attributes which is 80% of the variance and the reason normalizing it supplying it to the clustering algorithm and using trust model visualizer so let's take a look at the results here are those five clusters with cluster zero being of the smaller size and cluster 3 being of the largest size we can actually look take a look at the centroid table to understand the characteristics of the of each of the cluster and this basically creates at the center of each of the cluster based on those actually shapes applied and we can further use the model visualizers to actually summarize each of the clusters so let's take a look at at the presentation as to how we have summarized the clusters and what were the different personas found so these are those five clusters with the first lustre being kerning is highly interactive and inbound based on the observations this cluster was very responsive and also proactive and asking questions on his pointing to the email and there was a high and a high conversion ratio was observed with the conversion ratio of 3.9% the last two clusters are named as pretends to be involved in minimal interaction because their response rate was low and generators also knows that that's 0.5 so we used the cluster model to cluster a new lead come into the system and tag them ing to these ing to the behavior classify into one of these categories and this was very useful for the agent to understand or behavior of the eat and understand how they should be engaged in the future or what could be expected of them so let's move to the next use case which is channel attribution so once when the lead and they're like please coming from multiple channels and it was important for the client to understand which of the channels are performing better or where are the best quality of leads coming from and what is the effect of one child over the others we have used a microbe change which is a concept of simulation basically reading one child at a time from the system and studied the effect of that channel already conversion or or the rate conversion of other channels and let's take a look at the data first okay okay so this is the data and we had to arrange the data to create a lead journey only path for the different channels of the lead has touched during its journey and you can see this is the lead ID that is the first lead and this is arranged in ascending order of the reactivity which is this is the different channel that the need has touched and let's take a look at the design of the process and okay I just run this thing so basically the output of this is three different results that the first results that explains the removal effect that is when a channeler is removed from the system what is the effect on the lead conversion and the first column here is the channel we have 11 different channels this surprises but sorting get teleplay so the removal effect you can see channel number one has the highest total in effect that means of turn number one is removed from the system it would affect 91% of elite followed by channel two which would effectively six percent of the beat and so on the other one is to understand when does the channel appear in the lead journey with the China appeared his hostage point or whether it appears in the middle of the journey or whether it appear in the left its point fastest point is just before the conversion so every channel is given up value based on where it appears in the lead joining so China number one it really appears everywhere it looks like that appears yeah so each each and every channel has a value and now across these different variables its first stitch last stitch I mean your touch let's take a look at the results I just go back to the presentation and we can see this area so you can see on x-axis we have different channels it is channel 1 to channel 9 and or the y-axis we have the count of the total amount of leads which have appeared where doctor channels channels appeared his first touch point as the lastest point and there's a lineage point I need it the color here represents its time you can see the child number 2 is very popular and it kind of appears everywhere in the lead journey and channel number 1 you're actually appears most of the time as a last touch conversion so child number one in looks like it's very important because it's the channel which could have influenced the decision of the lead so when we study the removal defect the channel number one has scored the highest and it so it here it means that if channel move one is removed it would affect 94% of the leads followed by China number two which would affect or 65 percent of the leads to do this we have use only those leads which have converted in the past and this was very useful in understanding the behavior of the lead coming from different channels or importance the channel and further we have also studied the transition matrix trailer transition matrix to study the effect of one channel over the other this is basically a correlation type metrics and need for example if you look at China number seven it of 0.8 the correlation between channel 7 and channel number 10 is a 0.85 that means 85% of the leaves who come to channel 7 would transition to channel number 10 this kind of study was very important to understand the effect upon China over the other and if it is removal Houston how it would affect the other channel okay so let's move forward to our next test case which is lead scoring and this was one of the most important to use case because I can mention before there were many leads coming into system and attack being important for the client to quantify the leads based on the probability of the lead converting and categorize the leads into a VC priority or based on in society with a being of the highest priority and C being of the lowest priority and we have scored each and every lead on a scale of zero to hundred and categorize them as a high-low or ABC and we have used logistic regression for classification we had we obviously hired our class and balance issue because any conversion rate was very less and we have this module to boost the minority class we had a lot of features in the system so we have used virtually index with each a selection and selected top top n number of features which explained 80 percent of the variance the data so let's take a look at the process okay so this theorem is used for beat scoring and we have used lead demographics literally the cabinet literally the activity data for scoring and we had 72 regular attribute so there was a need to do which is election and let's take a look at it our design graphically in a process and like this one we take sign so we had built it on those leads which will offer a spam of broad nature taking only importantly it's for a modeling that is the leader jet converted in the past of the leadership which have not in my past and which have a good amount of data and I just go inside here to show you how we have done feature selection and here you can see the selection to select only top coding features based on the arrow importance and we have used GLM whoredom training the model so let's take a look at the results gonna take some time I already have the results again here okay so the output of this is a score on a scale of 0 to 180 each lead and this explains the probability of the leading about time this score is correlated to those optic it's correlated to the applications were converted in the past those kind of profiles which are converted in the past would get a higher school plus the leads which are active in the system would get a higher score someone which is having a good profile but are not only in a system or stale and not moving further in the process will be given to ask where go back to the presentation ok against each and every lead we get we provide a score of 0 to 100 and this is coming from tableau you can see that the leads you can see two scores here you can see needs coin applications course application score is a score based on the Committee of the leader correct affinity of the lead and I will explain that in the next section and you can see that this lead firstly it has elite score of 65 and set it has a lead score of 72 and the score is correlated to the satellite status of the lead when its open engaged or outreach of God by charter and also the stage of the lead as to in which stated it is whether it's in the known stage or whether it's marketing qualified needle with its sales qualified lead this was very important for the agents to understand the to prioritize their lead and to understand what the need would become looking why not and how much time should be given to the lead it's good and once they didn't understand how to prioritize their lead they also wanted to understand with what kind of products should be offered to these leads to ensured conversion and or anyone in order to do that we have used descriptive statistics to understand what kind of profiles have Ethan do - what kind of product and we have created this correlation table in tableau and you can see that program number one that is computer science you're on the x-axis you have different programs that have different card or different products and you here you see different you see the count of lead to have opted for this program and the kind three colors here represents the category of the lead with green being the highest quality lead looping mid quality orange being the least quality and program number one that is computer science is highly trading because most of the leaf coming in system are octal for computer science program and studied in the past followed by a mechanical engineering and so on and we have supplemented this result with a score on the acquisition route that is from where these leads are coming and you can see that social media is the top one channel from where what's the nature coming but the quality of the lead coming from the channel is mixed nature with 50 persons having with Pauling in the B category in 50 percent polling in the C category whether we give a score on a scale of zero to hundred against each and every application and there are two leads here for example each second doughnut you can see that they have these two applications have the same Li score because it's essentially for the same person that they have different application score which is basically a product affinity score that is of 2018-19 and they are owned by two different ages that's Alex and Alicia and by looking at the application score we can see that Alex has a higher chance of closing the little compared to Alicia okay so the next year's case is discount recommendation and there was a need for the agents to understand how much this con should be offered to the late in order to close close the lid close the deal and not all is required to score in certain profiles of the need where price sensitive we were looking for it offers an insight of a certain profile so definitely not looking for price but they were looking for other things like the rank of the University etcetera and in order to determine what is the optimal amount of discount that should be opportunity had use descriptive analysis by studying what has happened in the past and what amount of discount will result into what percentage of conversion so let me take you to the results this is coming from tableau and we have here and comparing two regions that is China South Asia and on the x axis here you see different programs that is programmable fun to program them to pipe it is compared across China in South Asia and these three different segments represents the quality of the lead or the practive it highly given low and the colored berries here represent the range of discount offered in the past the blue represents no discount light to the precise length and 5k dark orange represents between 5k to 10k in light orange represents more than 10k and clearly it is visible here the China is the list price sensitive over South Asia and most of the leads are converted without discounts but program number three and program number four in south asia region have enjoyed certain amount of has come and they to this kind of tells the ages as to which program should be offered what amount of discount which is related to but is correlated with the region and this has helped them optimize the sales cost this study was very important for the client to understand which kind of programs should requires what kind of offers and incentives in order to promote their sales in these are the five use cases but there are many of the use cases which are their anti predictive marketing analysis and with us they will take you through the key takeaways and other use cases and over to you mother said thank you very much so that was great Thank You Nikita for taking us through the use cases and explaining the inside derived from each of them we've seen how predictive marketing analytics can be an essential piece of marketing journey being sides through PMA can expedite your efforts in winning and osawa but that is just one half of the battle keeping them engaged and ultimately retaining them is the real long-term success our team and analytics have created this long-term success story for organizations of Consular our PMA solution has practically simplified their marketing efforts and their goal of delighting their customers with a personalized service is now simpler to achieve then ever based out of Texas we are in business since 2004 and served over 300 customers our services are not just limited to predictive marketing we are in fact a brand under which more than 400 technology professionals data analytics and scientists throughout the globe collaborate to help our clients navigate the digital transformation all this by leveraging our advanced analytics big data and a IBM's automation to easier PD process the credit of course also goes to our partner rapid miner and its products potential to build clean and interpretable code that helped lead a science team design a solution that has a decreased time to market which I believe is the most appealing feature for any client so thank you again for joining with us today please feel free to get in touch with us to know more about and lakes and their services and products we offer thank you