Streamline your document workflows with lead segmentation for R&D

How to create outlook signature

so I I'm Dan Rios and talked about customer segmentation or startups under this is primarily a 101 to Mohawk it's not like a 301 over lines over when they talk about slicing and dicing clearly granularly I imagine most of you at this point do not have too much of data we're really going to talk about the initial queries and segmentations you should run from five to optimize for growth with really small budgets right and to make the most impact so quickly about you know we get started I spent nine years in Silicon Valley various marketing roles for consumer tech companies everything from product marketing performance pumping and bus reach McGrath last couple of radios were split between Zynga and left I think I was one of the first mobile marketing hires and if their acquisitions spend across their entire portfolio of games less athletic side supports higher on the breath team came in pretty early AmeriCorps markets help them scale to 65 markets and run care so that was crazy ride unintended definitely intended distribution happen start out so the intent of Kennex primarily 101 we're really going to be looking for power law distributions 8020 s right to see if you can find power users provides really target and grow their talk a little bit about like what customer segmentation is from an academic standpoint and then what types of data you should be collecting right everything from a April data demographic data right little talk a little bit about psychographic data again I'm not going to go too deep into all that stuff but ever they want to focus on finding a real data then I want to talk about from applying these different data types right they actually get like a concrete segments and company view of your users look at lift and kind of see how we did that they're you know talk about Cigna as well kind of a case study from premium gaming so a little bit of this talk is actually kind of like me opening old wounds and shows the operations of the valley is like and generate a lot of learnings from working at both of those companies they want to tell you guys about how segmentation can work right across different verticals and across so look at the computer as well this is highly segmented and there's some pitfalls there right I want you guys to avoid those pitfalls the last time we'll talk about optimizing expansion but kind of figure out where your power users are coming from and how to go after them and I will close up with some recommended tools so what is it and why should you care so this is a definition from behind there we go but this definition from Salesforce comm customer segmentation is the practice of buying a customer base into groups of individuals there are simple and specific ways relevant to marketing at this age gender and Christmas events so also elsewhere stock knowledge really saying is all this is is dividing your user base but you see the body of your user base by different data points in each of gender interest spending habits so demographic data internal data for marketing purposes right using these noted those data points it's going to help grow your company by its help market to your customers right and so there's a lot more than just a gender and spending there's a lot of different types of data that you can actually divide your users fun so the primary source ever going to look at is April data so we'll talk about you know if you're a funnel right and engagement with your funnel right public standpoint we'll also talk about geographic data for some of you you might see really strong pockets of users in different zip codes or different cities something we'll talk about a demographic data for us agent gender also talked about Sligo graphic data just matter so why are they using it will actually see some of your power users and your non power users they'll use it differently like even if you only have one offering they might still use your product differently and the last thing close up that acquisition sores right so where are these people coming from okay so hot shave you care because this is your right now right I'm gonna look at all of your models but I definitely know that a lot of this or a lot of you guys but I probably have this problem for your earning some money and you don't know it and that's because you're probably not discriminating them across your user base as well as he should be so not all of your visas are equal I know you've heard that now multiple times for multiple people but I'm happy to like kind of belabor that point further that there's definitely differences in your user quality and you should try to find out you know those differences as soon as possible so you can use them we can optimize them for growth purposes so again the first type of data that I want to kind of go over is April 8th right and this is data that leverages your fun this is the most important source so your market analyst analytics that and you talked about yesterday that's the most important source of data that we're going to leverage here and I mean if you look at this model it's simply just an R funnel right so you guys already know what that is but I'm just gonna quickly go over it just for the point of circular learning and just think I managed it over the head with it you guys already know that acquisition and the top of the funnel right so driving new users whether they're signups or visits or installs activation this is key right it's your engagement metric right it's after you converting your new users into active users right so it's getting a user to actually engage with your product and experience that for value proposition whatever the hell that is right and so I lived I was like taking a ride for the first time right once they got them there and be active in enough ads it could have been playing a level for the first time in a new game if you're an e-commerce company it might be it might be in the user to make a purchase for the first time right retention is simply getting those users to repeat that behavior time and time again referrals of course is just getting your users to invite their friends and if you haven't monetize them already the activation metric that means just monetizing another but when you when you segment your users right you start blending customer segmentation the most important metric to look at the first thing you should look at is that activation a current price they launched a gas tactic look at engagement with that for action whatever it is so as you should soar your users find that occupation metric right so I'd lift we wanted to look at my company rights or our users taking across our entire user community if you're an e-commerce company you want to look at how many purchases your users or banking across your entirety surveys and look at this with that distribution see if you can identify the trends and I'm going to show you what that looks like right as this is like a really extreme example right where I have user count on the y-axis and I have total purchases on the x-axis and again you know this is sticking with the ecommerce example right where you know I'm looking for basically patterns based on frequency of usage right so total purchases is my core activation method except or method that I'm looking for see if I can identify patterns right in this example is really extremely right I have two clusters here you probably can see the numbers but have a group of users who purchased one two five things and improve the users super just 21 to 25 minutes on the right hand side here right and so what you want to look at when you see stuff like this if you are know what's going on here I'd like you might think that the people of purchasing 21 to 25 things right that they actually have a high or Ltd but they may or they may not right like they might be purchasing really really low-quality things and purchases the people to saying one five banks might be purchasing really high ticket value things right so that's really important excuse me so frequency plus how your users use your product that's why we services your product leads to a customer segmentation once you absolutely build out those segments you want to identify who these people are right like what are they doing and who aren't it what you really want to know who they are and for that I turn to your audience insights Facebook has really really great data on age gender location income profession and education so you can upload any segment of your users leveraging Facebook audience insights and don't give you all this data right they'll tell you exactly who they are sorry about this all right so when you combined it to you get different segments right it's well lived you look at our right volume right and we look at that across our entire user base right we did notice different patterns so we saw people use it in every single day right and one second we solved for a commuters rights of people using it every day and they were obviously using it to go to work and come back right they were jamming between 25 and 35 years old and these kids like leafy men like as always to do the ellipse grin they were super high income right they generated about or they earned about $100,000 a year and their one year in an Ltd was back 250 bucks and now that that one get NL TV that 250 I'm sorry I shouldn't call that out as aggressively that's just to kind of show you the value that this users like it's not the actual user value but it's just to show you that how valuable they could be so that was one user segment the next Easter said one that we found where the bar offers right so these are people that used to live primarily on the weekend they were 18 to 35 years old they skewed slightly female they generated their net Ltd was about and it was less than a hundred bucks again that's a hypothetical number it was not the exact LP people but favor they roll a lot of them and they use lists i barely has designated drivers on the weekends right and this was the biggest thing it's included everybody from college kids professionals to young professionals and again disputes like email because listen right next thing i want to talk about was primarily avoiding pitfalls and not understanding user behavior enough to basically avoid burning money so gonna talk about thinka and going well hunting and singer essentially freemium gaming is comprised of a really tiny audience that leads to a lot of mana station right so 1% of your audience actually leads to 95% of the modernization and when you look at that again something like this ok this is exactly what it is and so if you don't understand that and you by users indiscriminately you're wasting a lot of money right this is what I was getting but I was getting to when I was saying like don't burn money don't waste your money because if you buy users across the board without understanding that some users are a lot more valuable than other users then you're gonna turn cash right and so let's think it did I want a lot of people gaming companies do you spend in a penalty right they buy users they buy in slopes they buy signups they don't actually type path to monetization right and you want to type at the modernization we want to make sure what you're spending money and when you're looking at your path you're actually looking for paying customers like you're not looking for installs and I'm looking for signups you don't care about that you have to want a paying customer right and so when you do your unit economics make sure that in your time you the path is someone who actually eight is pierre-paul might actually paying for something so I talked about like how not to spent your money and I will talk about how you should spend your money but and how you can optimize your ad spend right and so I have a couple of scenarios here right here I have my first scenario where you're after you are making the distinction between power equal between customers and users right so you're only looking at your caffeine or LTV on a customer level so you're only looking at people who actually pay for something yeah here the uneconomic right interacting pretty rosy but I do want to show you how you can make them even better right so unit balance here LT give your average customer 15 bucks tak would be average customers $10 right these are pretty rosy right so a thousand dollars in marketing spend would lead to 1,500 a dress code right which is 500 dollars in profit for you which is in fact a second is the fifty percent return on your decimal pretty sweet scenario two I want to be even better right and this is what I really want to show you as I drill down to scenario two and I kind of go over payback windows and stuff like that if you actually look for power users right you can find on your LTV of your power customers there's a lot more so again this is hypothetical right I this isn't exact data which is very very common so LTV a group our customers could be 50 bucks chooses the reacts with your average customer I seen that time and time again especially when it comes down to ecommerce you're Catholic your power customer I'm throwing the era's like fifteen fifteen dollars this is just again I put that with a number it could be higher to be lower than your calf of your average customer but I just kind of want to show that to you right so a thousand dollars in marketing leads the 3300 introspect right and that's basically 2,300 dollars in profit two months I get 230 percent ROI so you just put googled your ROI right next I want to talk about payback windows right so you saw that scenario where we found our settlement those users are a lot more profitable but and you but you might be thinking that that person area wasn't that bad right might be thinking well actually is narrowing a wand and we're like I was generally 50% in profit right wasn't that bad for me but actually generate a good amount of profit that was pretty sweet right but I do want to go over a payback wages because power customers generally have a lot to repay back windows right so your power customers generally have in this example I have your power festival company a month payback window your average customer could have like a six to nine month payback window right and so if you have toward our payback windows it's a lot easier for like any growth marketer to come in it spend a lot more aggressively the love it payback window is the less you can spend up front so it takes a lot longer our net money back right so when you identify power customers you want to look at their payback window right that's really really important so don't take your uneconomic not like basically face value drill back down into the actual payback I don't see how long it'll take you to earn that money back because that's important and I'll help you grow your business faster by investing a lot more often okay last thing I want to talk about attribution so I'm not going to go in and geek out on like all these different tools because I think like Andy and maybe a couple people talking apart over that but you do want to know where these people are coming from right and so tools to use everything from Mixpanel Google Analytics for actually depends you know whether you're primarily whatever mobile branch is primarily mobile Mixpanel Google you can use for web or mobile attribution what these guys will allow you to do is basically figure out what you can also perform better which channels to perform worship and allow you to double down right if you double down on channels and you can also attack me right detective targeting Yucatec created and you milk they double down on that on ads and message right so allow you to break in higher-quality users from like specific channels and with specific campaigns alright so I just want to stop up here and I just kinda quickly talk about everything I went over there so you want to spend versus buying it I don't know how frequently they're actually engaging with your app whatever it is whether it's purchases whether it's rise whether it's playing level on a game and then you want to answer that house in the wines right how are they engaging with my boss why are they doing it are they taking a ride every day you know to go work and come back they just using acid estimated driver are they buying expensive items are they buying clothes like low value items how why are they actually using I brought up and then you want to marry them new on top of that right so who are these people right you know and again it's with audience and sites allows you to do that really quickly and then lastly you want to optimize ad spend for powered musics to maximize the ROI with limited budget right so you want to leverage your ad spend to target these people right and only those people if you can identify those power users and I'm going to close out all the above will help you grow a lot faster and before I before I let you go to open stuff to questions I just want to follow these tools again but I recommend so fixed panel web analytics amplitudes great for mobile in app analytics Facebook audience insights yeah that'll give you demographic data autopilot awesome for CRM optimization segment IO and particle user kind of like the the one they like these data hubs I like to call them the one API integration tool all API intuitions so if you plug in a segment on Farkle you're basically plugging in anybody else really quickly and easily that's it guys that's all I have hopefully you got the idea that power users didn't matter but their work a lot more that you can actually get paid back on your investment a lot faster using power reducers or identifying power users and trying to bring them into your funnel I'll open up the questions now went over some of that really fast got a little nervous for whatever reason but I'll go put it up so if anybody has any question it's love you know I'm also almost to go to slide you to see if there's anything in there questions yet any questions general I'll bet may be a strange question but I'm just thinking about me have you found it here experiencing the limitations and power users so you're thinking about the case of everywhere because it's a great product but well there's only so many of them right so I think like in the case of Twitter they're flatlining under breath right now right in fact they've got all the pieces they're gonna get and plus the patients are in so now like there are strategies to expand what we need to try to find new power users in different pockets but they're pretty much saturated in the u.s. so yeah you go saturate them over time and then you can always try to hunt the number with via new channels but most likely we in Twitter skates are so whole they probably already got all the pieces they're going to get so yes I mean you see zephyr okay Full Nelson yeah so much about direct consumer surveys triack impacts information outside of say Facebook we've certainly found that not everybody it's hardly accurate so the question is really do you find other information from direct survey help I was going to call it Saturdays I recommend it some companies use their days especially with like our psychic identified but all these assessment systems such an easier tool to use that they give you more data and multiple don't like responding to surveys the engagements usually low so you usually have to incentivize those surveys are you sending out like instead of eyes there times that my service usually non incentivize our users Howard customers trying to understand there be that might be on that are currently not just the basics of age and gender and the websites they make is that television shows for magazines there okay yes to get data that Facebook one have if you want to get like data specific to your business I think it's probably make sense surveying stuff develop after that I was kind of getting a pretty broad brush and pretty broad stroke but like this general demographic data and where you can get that so it's kind of like demographically at one on one right like age gender location right that's pretty much wraps and everybody but if you are not specific product use case stuff and is there any definitely works okay well other stuff anything else whenever that pretty fast so I followed s for that but anything else around how is it or what not I do not see any other questions so I will let it go at that guys I am my email stand off ever starts out tomcee to make questions feel free to reach out if you are cohort analysis what about it for analysis see change me they other things this is so power we uses is a type of colored so like if you look at color definition so you can provide that for could be anything we made users with / time who can be used to spawn by age gender location of their in right so we're just a group of users the distinction in terms of how do you want to find those new abuses how you want to divide those drug abusers is up to you most code most people use open else's anonymously with time right so it could be a group of users that you acquired in January of this year right and then you look at their behavior over time to kind of identify what they're doing and whether they're power issues that they're not is that the question in terms like how do you look at color in general or I just want a little more guidance in terms of what you want probably cover it's a really brought if you had it go yeah so if you identify a segment of power he says the baby is think from like your your other users right yeah most likely they would be subletting you have a glide and filbert you just say you've hired in January like everybody that you acquired in general right so maybe I don't know so you did 100 times a day three thousand times and ten percent of the more power you subscribe they were spending a lot of money so maybe they were spending like $1,000 a day right you're generating $300,000 most people you could look at those people as like high value payers and just segment out those users and look at their activity as a separate color okay everybody else well so that looks like it looks like it's it I think from the floor I don't I don't see any other questions but again if you guys want to talk about LGBT populations you know economics power users not our users it could be Jeff Daniel 500 startups here okay basically Monday we so you me anytime thanks guys