 Okay, hello everyone. Thanks for coming this morning. It's my first time in Helsinki. I'm really excited And today I'm going to talk to you guys about onboarding Introducing people to your products and why it's usually the most important part of your growth strategy So a little bit about me I help companies scale so as an employee I led the growth team at Pinterest I started the marketing team at Grubhub as a growth advisor to Greylock partners a venture firm in the Valley And I started my career as an analyst at apartments.com and now I work as an advisor To a bunch of different consumer marketplaces network-driven businesses Places like Eventbrite, Thumbtack, Reddit to help them successfully scale their businesses So what we're gonna get into today is why focus on onboarding? Why do I think this thing is so important? To how you should think about optimizing your onboarding Some stories of how I worked on this problem at Grubhub and Pinterest And then how you should think about the the process of using qualitative and quantitative data to help your business in this area So to get started why onboarding? Why is this thing so important? If you've ever done a cohort graph how many people have actually done a cohort analysis for your business? Okay, some of you I'll show one later in the presentation This is where you watch people kind of try your product And you see how many people continue to use it over time and inevitably in most businesses You see the most people drop out at the beginning they try the product once and they never come back And then hopefully if you have a good product some people stay for the long term and build a habit So if you fix onboarding or if you improve it actually is the biggest driver of your long-term retention in most businesses but not only that Directly onboarding is the biggest driver of acquisition as well Most of this most scalable strategies for acquiring new users have to do with retained users Bringing other users into the product either they tell other people and and the product goes viral Or they create content that then you can distribute to a bunch of other people Or they just make you a lot of money which you can reinvest in you know sales or marketing So not only is onboarding usually the biggest driver of retention Indirectly, it's the biggest driver of acquisition as well And a couple things happen when you under invest in onboarding new users One is that people come in and they're just confused They're trying to figure out the value of this product and you're not teaching them how to get the value Meaning, you know, they churn and then they don't tell other people But what happens that's even worse is if you're not focusing on your new users Typically as product managers designers, etc You're then going to be focusing on your existing users who have somehow figured out how to get value out of this product Even though it's hard to understand and then they're going to want more complex features They're going to want you to build new functionality which then makes the product even harder to comprehend for the new users coming in So not only do you get kind of a first-time bad user experience for most people But the people who get through that end up indirectly making it worse for everyone else so The first thing you should think about if you're trying to improve your onboarding is You know, what's kind of the metric I should be measuring? so the the first step of figuring out that metric is What's the key action in your product that most correlates to someone receiving value from it and? How often would you expect someone to want to receive that value which I call the designated frequency and And for Pinterest it took some work to figure out the answers to these questions so for Pinterest you can scroll through a lot of images you can Click on images and go to where the content came from or you can save things for later by creating a board and What we found as we did research is you'd be scrolling through a lot of content because you have a lot of high intent but it doesn't mean we're actually showing things you want to see and You can click on content and that may make us think it's good But I'm sure we've all had the experience with clickbait where you click on something and it over promises and under delivers So saving something that Pinterest showed you ended up being the best key action That means Pinterest recommended something you liked it enough to want to see it again later So figure out the designated frequency a trick I like to use is just thinking about Before your product existed how did people try to solve this problem? So for Pinterest where did people go to try to find inspirations on topics? They cared about they bought magazines and most people got magazines on a monthly or weekly subscription So our original designated frequency target was monthly and then we later changed it to weekly For Grubhub this problem was a lot simpler There's only one thing you can do with Grubhub and that's order food have it come and have it be delicious So the key action was very easy for us. It was how many people ordered food online and Then of course people ordered food before Grubhub existed. They just used their phone to call the restaurant So we asked these people how often do they order food on Grubhub? And they would say once or twice a month so our designated frequency target for Grubhub was also monthly So if you have these two data points then you can build a cohort analysis and on the y-axis of this graph is the key action the percentage of people that are doing it over time and then on the x-axis is your designated frequency and If your business has product mark a fit you will see a flattening of this curve a Bunch of people in the darker blue will drop out They will churn and then some people hopefully will stick with the product and build a habit so the point at which you stop losing users out of your cohort is what we call the habit moment and The way you measure that is the habit metric How long does it take and what are the actions people need to do to get to that habit? So it's finding not only the key action and a designated frequency But what's that retention interval at which you can say if they're doing it this week? They're gonna do it next week or if they're doing it this month. They're gonna do it next month But people don't get to the habit of using a product immediately right it takes time So you have to work your way backward from okay these people who got the habit. How did they get hair and? the two Pieces that you need to figure out are the aha moment and the setup moment The aha moment is the first time someone experiences that value where they really get it and they appreciate the products that You're building and the setup moment is the work that both the user and you as the company need to do To make sure people can reliably get to that aha moment and I'll walk through a bunch of examples soon But first who should actually work on this problem who what's an onboarding team look like? This is not a marketing problem. This is not just a sales problem This is a cross-functional team product designers engineers product managers If it's b2b a lot of time sales will be involved sometimes marketing will be involved user research All these people have to be working together to figure out how to build the right experience So that people are gonna get the value of your product and keep using it Which begs the question? Where does onboarding end and the core product experience begin and The argument I would make is that your onboarding team should be in charge of the entire product experience Until people reach the habit moment, which is a bit scary at first, but when we switch to this Mantra of working for our onboarding team at Pinterest. We were able to have a lot larger gains and retention as a result Okay, so let's actually get to improving your onboarding. How do you do it? Well, it requires both people and metrics. You have to look at your data and you have to talk to users So, you know, here's some examples from Grubhub. This is a graph that I pulled of the number of online ordering restaurants on the x-axis when you typed in your address and on the y-axis your Conversion rate to purchase and what we noticed in almost every city we launched in is at a certain point There's this step change where the conversion rate doubles and this is the city of Boston and the conversion rate doubles at around 55 restaurants So when we saw this it was like, okay It's very clear if we can get to a certain density of restaurants Then people are much more likely to activate they're much more likely to retain So then the next question became Where are all the areas where we're not reliably getting people to these number of restaurants so that we can focus our sales efforts on Making sure that everyone who person who searches in Boston for example is going to get at least 55 restaurants and Doing this helped us increase our retention dramatically So this is an example of looking at the data But then we also talked to users and when we talked to users say, oh, why didn't you keep using Grubhub? Or why aren't you using it more often and they would say oh, it's expensive and we'd be like, okay Well, that's weird. We don't actually charge you we charge the restaurant What they meant is that restaurants had high delivery minimums and fees in some cases Which meant you'd have to spend a little bit more than you wanted to To enjoy a meal which meant that you weren't going to do it that often or you weren't going to do it again so what we did is we took this information and We went back to our restaurants and we said hey We think there's a lot of latent demand to order more often that you're missing because of your minimums and fees Why don't you try dropping those and we expect to see that there will be an increase in volume that will more than make up for the loss in margin on a poor order basis and For those of you who aren't familiar with the restaurant industry Restaurant industry is normally a very low-margin business But that is primarily driven by the front of the house You just can't pack more people into a place though slush is testing that today But they usually can always make more food in the kitchen so they can always pump more food out for delivery So delivery ends up being the highest margin business that a restaurant usually has so we convinced this random Costa Rican restaurant to try as they drop their minimum and they immediately jump to the number one restaurant in Chicago Which was one of our most popular markets And then we're able to build a case study around that To showcase to a bunch of other our other restaurants to try this and we're able to systematically lower minimums and fees as a result So how do you think about the setup moment and the aha moment for Grubhub? So if you go to Grubhub pretty much every screen will have something like this at the top Where are you give us your physical address and this is a pretty big ask for someone who's just coming to your website for the first time But restaurant delivery boundaries are not based on the city or based on the zip code or the neighborhood They're based on the street So we need to actually know your address to reliably show you good results So even though it's the big ask we make the user give us this information But then of course that alone isn't going to get you to the aha moment We then need to make sure we have a plethora of great online ordering results for you For every single address. So in this case, this is my address. Don't write it down But we have plenty of restaurants that deliver so the chance of me finding something and building a habit are very high for Pinterest this is what the Landing screen for a new user look like when we started working on onboarding. We're asking people to find friends. We're asking people To invite friends. There's all these images. Some are picked for you based on your activity Some are based on a topic you follow Some are based on friends that you follow people were so confused So as we looked at the data we saw that the people who got through this experience successfully Found something they liked enough to save it And when we started doing the user research, especially internationally, we were finding that pretty much every piece of content We were recommending was in English and People would say look I can speak English. I can read English. I don't want to I want to I want to read in my local language So we went to work on changing a lot of these emphasizing repinning emphasizing, you know, local content and Here are some specific examples So on Android where most people internationally we're signing up We radically simplified the sign-up process to get you to the aha moment faster So we started pre-populating your email address in your name We started removing password as a requirement and then age and gender are Skippable so then we can immediately get you into the core But then we would stop you and ask you to pick five topics because we need to know what you're interested in In order to show you a great feed So eliminated a lot of the extraneous steps, but forced topic Picking because that is the real setup moment and Then when you got into that feed for the first time we used to show all this information about the piece of content who pinned it where the image came from what the description was and People didn't know what any of that stuff meant and they weren't interested in it So on the right we switched to an experience where we're just showing more imagery and less textual content Which means more content on the page to browse at the same time more likely to find something you like more likely to repin More likely to retain We also started adding more user education and you know, there's a design that there's a design philosophy That's common in the Valley that says if a design needs education. It's a bad design My response to this is a design with education is better than a design that doesn't educate if your user needs to know what to do and You know education can help them figure out how to get value do not be afraid to do that So in this particular example, you can see on the left you land on the feed for the first time And we just ask you to scroll to see more When you stop scrolling that education changes to say, you know tap any pin for a closer look and When you get to an actually image Under over the save button, there's this pulsing icon Drawing your eye to that saying hey, here's the key thing you need to learn on this page And when you click on it, it explains the value of saving and all of this work increased activation rate dramatically essentially doubling activation rate over time for Pinterest So one of the things that you should think about for your business is how fast do you need to get people to that? Aha moment like how important is time to value? So for something where there's a lot of intent like doing your taxes if you don't do them you will go to jail People are going to go through a lot of work to make sure they get the value done which is completing their taxes, right? So the onboarding process can be fairly complex. It can be fairly long people are going to figure it out But for Pinterest Pinterest is almost on the entirely opposite end of this spectrum. Most people are signing up They've heard about it. They don't know exactly what value they're going to get They don't know if it's really for them. So we had to make the time to value really short Get them into the product value as soon as you can Otherwise they will bail they'll do something else and for Grubhub while the intent is high You you're hungry. You want to order food. There's also a bunch of alternatives to using Grubhub, right? You can go to the pantry and eat some chips You can go cook something or you can walk outside and find something to eat So if we weren't if we didn't have delicious food on its way to you in five minutes or less Chances are you're gonna find something else. So Grubhub was also on the left side of the spectrum But for a lot of B2B businesses the intent is so high that your onboarding can take weeks or months And people are still gonna deal with it and that's okay. You just need to figure out where you need to be on the spectrum Okay, so thinking through the quantitative steps to improving your onboarding The first is you have to define the actions that activated users and churn users are taking What are the people who activate in your product doing that churn users aren't and are there things churn users are doing that are making them Not successful and inevitably the first time you do this analysis The answer is anything people who do anything are way more likely to retain than people who don't do anything Which of course is not very helpful information But it points to this piece where all you can get from this analysis is correlation and you need to test Is this a survivorship bias? Of course the people who pinned 100 pins in the first day are more likely to retain That doesn't mean I can get everyone to do that or is it an actual insight that can help improve your onboarding So the only way to do that is to take the short list of things that correlate to people who retain and then start Emphasizing that in your onboarding seeing if you can get more people to do it And if you can get more of those people to do it does your activation rate increase does your retention rate increase for the business? On the qualitative side you need to talk to users both the people who churn from the product as well as the people who are Successful what are the people who are successful finding as the value of the product? And how is that different from the people who churn are the people who churn not Understanding that value or they just not a good fit for that value. We need to change your marketing habits to reach the right people And that's great for understanding the value of the product that you're building But in order to understand your actual onboarding flow these people aren't going to remember a lot of it They're not going to be able to give you great insights So to get that you have to watch new users run through the product you have to you know See what they're doing see if they're understanding ask them questions put a laptop in front of them Put it make them down that the app whatever and just watch them use it So we touched a bit on this earlier, but user research needs to be involved in every step of the way You just can't systematically improve your onboarding only looking at the data and I've seen a lot of teams You know forget to really involve user research as much as they can at Pinterest We essentially made them part of our cross-functional team So if you're thinking about okay, I want to go back and improve my onboarding What do I do this question of do you overhaul your entire experience or you know? Do you start optimizing and that question really depends on the stage of your company the earlier stage you are the more like you're just gonna try Something totally different in the confidence of what you're currently doing if your onboarding flow is working Then you're gonna start optimizing it by isolating variables and seeing if that's improving retention But if you're not confident if it's working or if you're hitting a local maxima Then that's where you might want to try something radically different and a be tested against what you've been doing and see what that difference looks like Okay, so this is just a sample of some of the stuff I've learned on onboarding I've actually worked to build an eight-week course on retention and engagement at reforge.com With Brian Balfour, Sean Klaus and Andrew Chen So if you're looking to get more into this topic, you know Feel free to come to reforge and sign up for the next class and thanks so much for having me