The Foolproof Formula for Finding Product-Market Fit

give that for Raul all right good morning everyone thank you for getting up so early for this I'm the founder and CEO of superhuman well we make the fastest email experience in the world our customers get through their inbox about twice as fast as before and many see inbox zero for the first time in years now this is the story of how we built a product market fit engine products market fit is the number one reason why startups succeed and the lack of product market fit is the number one reason why startups fail but what really is products market fit Paul Graham the founder of Y Combinator would say that product market fit is when you made something that people want Sam Altman until recently the president of Y Combinator would say that products market fit is when users love your product so much that they spontaneously tell other people to go and use it but it is perhaps Marc Andreessen one of the best venture capitalists of all time who said this he put it in the most vivid and compelling way you can almost always feel it's when products market fit is not happening customers aren't quite getting value users are not growing word-of-mouth is not spreading the press reviews a mediocre and the sales cycle takes too damn long but he said you can almost always feel it when products market fit is happening customers are buying as fast as you can add service you're hiring sales and supports as fast as you can reporters are constantly calling you about your hot new thing money is piling up in your checking accounts and investors are camping outside your house this is indeed a vivid and compelling definition and one that I was staring at through tears in the summer of 2017 you see it seemed so uh national so subjective what do you do if by this definition you do not have product market fit indeed I began to wonder can you measure product market fit because if you could then you could optimize it perhaps you could systematically even numerically increase product market fit and as it turns out the answer is yes you can measure it and you can optimize it but before I share how let's wind the clock back by nine years to 2010 nine years ago I started this company rapportive we built the first Gmail plugin to scale to millions of users when people emailed you we showed you what they look like where they worked their recent tweets links to their social profiles we scaled rapidly and in two years we were acquired by LinkedIn now during these four years I developed an intimate sense of email I could see Gmail bizarrely becoming worse every single year becoming more cluttered consuming more memory using more CPU slowing down your laptop and still not working properly offline and then on top of that people were installing plugins like our own rapportive but also boomerang mix max cleared that yes where you name it they had it and each one of those plugins took these issues of clutter memory CPU offline performance and made each one of them dramatically worse so we decided it was time for change we imagined an email experience that is blazingly fast where searches are instantaneous where everything happens and a hundred milliseconds or less an email experience where you never had to touch the mouse where you could do everything from the keyboard and fly through your inbox an email experience that just worked offline so you could be productive from anywhere and an email experience that had the best Gmail plugins natively built-in and yet was still somehow subtle minimal and visually gorgeous and so in the summer of 2015 we opened up our offices this was all fancy sign and we started to write code in the summer of 2016 we were still writing code and in the summer of 2017 we were still writing code and I felt this intense incredible pressure to launch both from the team but also from within myself after all our last company had started scaled and being acquired in less time and here we were two years in and we still had not launched but no matter how intensely I felt this pressure I knew that a launch would go very badly it would not be the Marc Andreessen story I did not believe that we had products markets fit and I couldn't just say this to the team these are hyper ambitious super intelligent engineers they poured their hearts and their souls into the product it's not what they wanted to hear I needed a plan and so I started my search for the holy grail for a way to define products market fit for a metric to measure products markets fit and for a methodology to systematically increase product markets fit I searched high and low I read everything I could find I spoke with all the experts and I found this guy Sean Ellis Sean ran early growth at Dropbox LogMeIn an event bright he coined the term growth hacker and as vivid and as compelling as Marc Andreessen definition is it is still a lagging indicator by the time that money is piling up in your checking account and investors are camping outside your house guess what you already have products market fit now what Sean found is a leading indicator one that is benchmarked and predictive simply ask your users this how would you feel if you couldn't no longer use the product and then measure the percents that say very disappointed after benchmarking hundreds of startups Sean found that the companies that struggled to grow almost always had less than 40 percent very disappointed but the companies that grew the most easily they almost always had more than 40% very disappointed if more than 40% of your users would be very disappointed without your product and you have initial products markets fit this example shows that same question put to 731 slack users 51 percent of whom said they would be very disappointed without slack 51 is greater than 40 therefor slack as products markets fit now you might argue you don't need a fancy question to tell you that but the purpose of this example is to show you just how hard it is to beat this benchmark this metric is more objective than a feeling it predicts success better than the Net Promoter Score it is not only the best way to measure products market fit with it you can build your very own product market fit engine and this as I will show you can automatically generate your roadmap for you step one is survey sends these four questions to all of your users how would you feel if you can no longer use the product what type of person do you think would most benefit from this product what is the biggest benefit that you get and how can we improve the product for you now you should wait until your users have had the opportunity to experience the core of your product this could be two or three transactions it could be a period of time at superhuman that we wait for three before we send this question there when you get your results analyze the answers to question number one and they will look something like this these are the actual results for super-human from the summer of 2017 as you can see only 22% of our users would have been very disappointed without our product we did not have products markets fit now this may seem sad but I at least had a way of explaining our situation to the team and moreover I had a cunning plan for increasing this number which brings us to step two which is to segment now we want to understand who are the people who really love our products and for this I like to use a concept of the highest expectation customer which is an idea I learn from Julie sue pan Julie ran early marketing at Airbnb Dropbox and many other great companies the highest expectation customer is the most discerning person in your target demographic they will use your products for its greatest benefit and they will help spread the word and most importantly others aspire to be like them because they see them to be clever judicious or insightful for example the Dropbox hxc is trusting they're technically savvy they want to save time and at the end of the day they simply want to know that Dropbox has their back when it comes to saving their life's work I'm an example of a Dropbox hxe I'm sure many of us here here are also the air B&B hxc doesn't simply want to travel they want to belong they want to experience Paris as if they really live there and Airbnb rose to its early success by catering into these taste makers and these influences so how do you create your own eight EXCI easy go back to the survey results take all the people who said they'd be very disappointed without your products remember these are the people who love your product and analyze their answers to question number two who do you think this product will be best for this turns out to be an incredibly powerful question because happy users will almost always answer it by describing themselves using the words that matter most to them you can then turn these words into a rich and detailed description of your highest expectation customer meet Nicole she is the superhuman hxc she's a hard-working professional she may be an executive a founder an investor or a manager she works really really hard often into the weekend now she feels she's productive but she's self-aware enough to realize she could always do better yet she's struggling to find the time to improve of course she does a ton of email on a typical day she'll read hundreds of emails and on a busy day she may send 80 or more now this is the most important part it is critical for her to be responsive because if she's not she could block her team damage her reputation or miss opportunities she does aim to get to inbox zero and she might do so a few times a week however once in a while perhaps once a year she'll have to declare email bankruptcy by the way if any of you have to do that let me know we can help she generally has a growth mindset and this is the good news for us the bad news for us is like many she has a fixed mindset about email she's skeptical that software can help improve her and make her go faster once you have your hxc let's get back to these survey results and assign a persona to each one and now here's the magic we're going to take the people who love the products the people who said that be very disappointed without it's in this case founder manager executive and business development and we're going to use them to narrow the market and the simplified example that means we focus on these and we deliberately ignore sales success engineering and data science and just by segmenting our product market fit score increases by 10 to 32% we're not quite at forty percent yet but this is a huge improvement for just two minutes of work and the next step then is to analyze now we know who really loves our product next we want to answer two questions why do people love our products and what holds people back from falling in love with our products to understand why people love our products once again we go back to the survey results we pick the people who said they'd be very disappointed without it the folks who love it and we analyze their answers to question number three what is the main benefit to you let's take a look at some examples processing email is much faster I get to my inbox in half the time the app is crazy fast it makes me feel like an actual superhuman faster responsiveness navigation so much faster than Gmail more efficient with my time do email more quickly speed aesthetics I can do anything from the keyboard I like to take these comments strip out all the filler words put it on a word cloud print us out large stick it on your wall and then it's as clear as day people happen to love our products for the speed for the focus and for the keyboard shortcuts and then we head back to this chart we want to increase the size of the very disappointed crowd well first of all and as painful letters as it is we have to ignore the not disappointed crowd they'll be very vocal they'll be very loud but they are so far from loving our products that they are essentially a lost cause you can build everything that they ask for and they still would not be very disappointed without your products they're too far away that leaves the somewhat disappointed crowd can we convince some of them to fall in love with our products the answer is most certainly we can but the question is to whom do we listen to because for about half of these you can build everything that they asked for and yet they still won't love your product and worse you'll end up with a muddled and an incoherent roadmap so how do we know who to listen to well here's the magic take the main benefit from before in our case speed and use that to further segment this segment you now have the somewhat disappointed people for whom the main benefit did not resonate and the somewhat disappointed people for whom the main benefit did take the former group and politely disregard them because you will build everything that they asked for and they still won't love your products because the main point of your product does not resonate but the other group we pay very special attention to those because something and probably something small is holding them back if you build it they will fall in love with the product and you can convert these people who are only somewhat disadvantaged without your product into being fanatical evangelists for what you do so how do we figure out what to build we take these set of people and we analyze their answers for question number four which is how can we improve the product for you well now several layers deep in this rabbit hole so I'm going to take it from the top just one more time we did a survey we figured out who love the products we figured out the main benefits of the product we figured out then who sort of slightly loved the product we then used the main benefit to segment those people who slightly loved the products into the set of people for whom the main benefit resonated then the main benefit didn't we then take the set of people for whom the main benefit resonated and we looked at how we can improve the product for them you can mathematically prove should you be inclined to do so that this will work you then look at their comments go through the same process you can create a word cloud and back in the summer of 2017 our biggest problem was that we did not have a mobile application that of course has since been resolved and after then the the insights become less obvious and more interesting search integrations calendaring attachment hound handler so now you know what people love now you know what objections people have the time has come to build step four is implement you want to spend half your time doubling down on the things that people love in our case that means your speed more shortcuts more efficiency and even more refined sense of aesthetics and the other half of your time systematically overcoming the objections to people loving your product if you only spend time doubling down on what people love as vision driven teams and founders are tempted to do you will never increase the size of the very disappointed segments but if you only spend your time systematically overcoming objections as data driven teams are tempted to do then competitor will quickly overtake you and so it is rational to spend half your time on one half your time on the other your roadmap automatically writes itself and in many senses this is optimal so now we're building things that people love were systematically overcoming objections the next fifth and final step of the engine is to track there are no silver bullets but this framework will work as you're building things constantly be surveying new users we serve a weekly and then aggregate into monthly and then aggregates again into quarterly the first quarter when we started this after segmentation you'll recall our product that product market fit score was 33 percent a quarter later it was 57 percent 47 percent the quarter later it was 56 percent and a quarter after that it was 58 percent 58 percent of our users would be very disappointed without super-human this product market fit engine really does work it is not only the best way to measure products market fits it gives you a methodology to systematically increase products market fit and if you want it can even automatically generate your roadmap for you now with that I hope I've inspired you to at least try this in your own companies and if you do please let me know I would be more than happy to help and with that I think we have time for some questions but before we do I would like to ask my co-founder of a vet to stand up so you can see him back stand up just to slightly area he's even more handsome in real life than in black-and-white photos so if you have questions please find him or me and we'll start here yes very helpful could you talk a little bit about how you might apply this much earlier so in your case you had spent about two years developing you had something to actually deploy to users to start to get feedback what about in the very early stages when your ID ating on possibilities and trying to determine where the value might actually be before you even have a lot built can you do you have any examples of how one might use this framework much earlier in the sort of implementation before even getting to implement did you actually had implement as number four but you actually had implementation as number one zero yes great question so the question is what do you do before you have users how can you tell whether you're making progress ideally someone on your team and if you're the product person yourself can be a proxy for the kind of person that you're building for we deliberately chose a business where at least one of us in fact several of us more most importantly me as the person making the product decisions could really feel for who those users are and before there are numbers before there are metrics that comes all down to feeling I used to be a video game designer back in the day and there's a question if you're a video game designer that you ask every single day which is is this fun it's a really interesting question because you can get deep into the rabbit hole of what fun is but the point is that you're asking an emotional question and there is no quantitative answer and so if a superhuman the question I was asking myself was does this feel blazingly fast and does it make me feel special as a user and it took us about three or four months to get to the point where there were flashes of brilliance and then the whole thing will fall over and it will be broken for three days and then a week later I'd feel that special moment again for like 10 seconds and the facts that every now and then I could sporadically feel it gave me hope to keep going but that will only carry you so far at the time when you can start surveying users I highly recommend you switch over from a retrospective method to a quantitative method yes here hello thank you so much by the way for this I learned a lot I guess my question is what is the critical mass either of users or of customer beyond which you believe the accuracy of your model get more and more accurate for example I am dealing with b2b so I won't have a critical mass of customer willing to give me feedback so at which point of time I would say let's wait for the 10 B customer and I know that this will help you fine-tune the method right so how many people do you need it turns out that you get statistically directionally accurate numbers with as few as 40 respondents you don't actually need that many now you mentioned b2b b2b is an interesting case where you have buyers who aren't necessarily the end users for us our buyers are the end users the surveying framework is simpler if you happen to be running a b2b product I would suggest running two parallel streams of surveys one for your end users because in this day and age they of course have to be happy but also one for the buyers because they will be extracting different value from your product and you're looking for at least forty people forty respondents in both sets of surveys next question doors thank you I'm blown away because every year at Summit this is my third one I meet someone I talked to someone that I didn't even know existed and no offense to you I did not know you existed however has an incurable entrepreneur you are now my best friend okay going out tagging on to his question because as maneuv started so many different things you realize one thing and that is that at the early stages the sheer energy and infection that you have for your product carries over to the people that you hand it to and they tell you that it's the it's the greatest thing in the world and you you go about it with full confidence that you have the product fit only to find out that the person who has not met you and has not been infected with you or enthusiasm doesn't have that so going back to the early stages again because that's where I find where most of the mistakes are made how can I present this data unemotionally and get a better response so the question is how can you present the data unemotionally I'm not sure why you would want to do that you know I mean the best presentations are left brained right brained the Hostin cult it's everything deploy everything you have you only have one chance to make the company work why would you pick half a story [Music] [Music] oh maybe I misunderstood you're talking about presenting to customers I see okay got it understood I would not present this kind of data to customers however there are collateral side-effects that you can gather here one of the most important is the two questions who do you think this product is best for and what is the main benefit there's a rule of thumb in marketing which is that the best marketing company you can possibly have is the copy that your customers write for you and this happens to be a trick for collecting that on mass now you still need a copywriter you still need a marketer to squeeze it down to really pity sentences but once you do you can then use that in your marketing copy so that's something that we do very much so you choose okay so about a year ago I was experiencing a forma if you're missing out because people are telling me this product is so good and I signed up and nothing happened because you had a waiting list and then a few weeks ago a friend referred me to theoretically to you or maybe someone works you and I was chatting with you on email or I don't know I actually was chatting with you and and so I'm curious if you could describe your your referral process strategy and and why it took so long people to get in the email chain and and why did I talk to you after referral but not coz actually applied a year ago so like how does that all work what's the reasoning behind all that Floyd buddy old pal I'm pretty sure it was me they don't think okay so it's a wait lists are a funny old thing people assume that our wait list which is quite substantial was about two hundred and fifty thousand people on it is because we're trying to drive foam we're trying to drive high post Cassity but that actually couldn't be further from the truth the purpose of the wait list was to pick the people and remains to pick the people who will be most successful with the product if you're building the kind of product like ours that lives or dies by word of mouth and another untold secret is that in these days that's every product then it's critical that as many of your users as possible are going to have stellar experiences even if that means you say no to 99% of people now email as a category has a tremendously large surface area imagine trying to rewrite Word or Excel it's that complicated so we can only support an a small number of people to begin with not from a scalability perspective but from a feature coverage perspective and we slowly grow that over time as fast as a start-up can do and then as we have new functionality that makes it accessible to more people on the wait list then we invite you on now there's a twist and there's a twist because we found through data that referrals massively outperform anybody else if your friend refers you to suit the human you are more than 15 percent more likely to fully convert over from whatever you were doing previously to using our product and there's many reasons for that there's a social obligation there's social proof there's the feeling of getting into this cool thing and that's all great the thing is we didn't design for that it just turned out to work that way we take some questions from the back hi mm sorry I'm not in the tech business so I saw weed and in Canada where you know the regulations there's it's not highly differentiated I can't imagine that anyone would say I would be devastated if I couldn't buy your pre-rolled joints they'll buy someone else's pre-rolled joints so I've been sitting here because I would love to figure it away so you get the question I mean I guess it's a commodity in some way you're building brand but I don't think anybody they might love your brand they might love your product they're not gonna say they're like gonna die if they don't have access to your product great question and you answered your own question which is brand if you had the means why would you buy a Lamborghini or Ferrari all the other way around they're both go about as fast as each other well maybe you care about the environments need buy a Tesla Roadster s the rule about as good as each other but very specific reasons to buy each and so for superhuman we picked a brand orientation and a segment of the market and I believe that you could do that for weed as well I would go super deep into branded marketing I am but I guess it's applying this idea of surveying and working with and having if it wouldn't be 40 percent saying they would be very disappointed I guess they would say their yeah okay I take it back [Laughter] hi thanks for that presentation I recently went through YC where we were told to do some of your practices and I have a CPG company so kind of piggybacking off of her I have hair products about 25 different products and when I asked this question some of them aren't answering just based off of the product that they tried and so how much you alter the survey to include different types of products that people are trying I'm and identifying what it is that they like about your brands or your products there's a really fascinating question I think I would have to understand the brand in more detail to coherently answer if it's the kind of thing where you expect people to pick one product line and stick with it then I would say they would have used that shampoo or conditioner or whatever it is at least two or three times but if on the other hand you think it's more of an eco system product well you have to buy into the whole idea that everything in your bathroom is coming from you then I would only survey that subset of users and in practice if I were doing this I would observe that I actually don't know and I would survey everybody but try different segmentations so I'd collect all the data and I'd look at the segment of people who have only bought one you and I would look at the second two people who have filled their bathroom with stuff from you and if they behave materially differently you know that your actual business lies in an ecosystem product I'm not one killer app you want to pick the next one let's see this gentleman Robbie excellent presentation Rahul really I also didn't know you existed so wonderful you're my new friend because we all struggle about that magic product market fit just that insight that there is a leading indicator is just brilliant brilliant I wanted to double-click on your idea about no waitlist that first of all is super human free and does that concept of a waitlist work for paid products or only for you know products that are free you know this points are beginning to question myself whether I exist so I think wait lists work exceptionally well in either case a great example of a very popular app right now with a really powerful wait list is Robin Hood you don't have to pay to get in you pay as you use it but the wait list mechanics because the core product is so incredible work fantastically well now the other question was around charging it's actually critical for superhuman that we charge for any number of different reasons because we charge and by the way everybody pays everybody pays $30 a month this is a premium product aimed at people for whom email is work and work is email and for that set of users that actually feels cheap but compared against the landscape of software as a service it may appear to be expensive and one of the main reasons why we charge is to create some degree of commitment you have to self-identify as somebody who says yes I have an email problem and yes I want to solve that and yes I'm going to turn up for an onboarding call for those of you who don't know ever one who's a superhuman also does a 30-minute live video call with one of our wonderful onboarding specialists because we find that the people who make it through this are those who wanted it anyway and the people who couldn't be bothered to put in a credit card or to turn up for a video call didn't either value their own time or have the growth mindset earnest around becoming faster to make the product worthwhile so I always always recommend charging because it leads to the creation of much more committed customers let's take some questions from that side of the room Harrell I'm one of those guys that were dying superhumans stopped existing our congratulations first do you lead growth at superhuman and second do you see any disadvantages for focusing on different segments of the market you have founders managers and something else and there's some people who would say to focus just on founders I no longer lead growth at superhuman we're about thirty eight two people and I'm trying to win the for those who are going through a similar thing I've actually learned a lot of people have done this before we're in the process of scaling very quickly and so I'm trying to you know build a machine that works the my contribution to growth was very early on where we came up with this framework and I said it was always going to be three things it was going to be PR there was going to be virality and I was going to be content marketing and that's because in 2014 I analyzed all of the ways that a company could grow and I identified that these are the only three things that would ever work for us and so I lay down the guardrails and now we have a great growth organization working around around those the growth organization so it's unusual compared to most companies because we do this onboarding process the question was how is the growth organization put together we have at this point coming up to 14 onboarding specialists remember these are the people who take their time spend most of their weeks to onboard you on to superhuman but you need the folks who will feed those people and so we call those customer acquisition specialists we have roughly two equivalent full-time people on that some of you may have interacted with aha she's half robot half Saha but she's very very good at her job and she feeds the funnel for the onboarding specialists and then below that you need a team of people who are there to support now in a normal organization you would call this customer support but I've always hated that phrase I hated the idea that you know they were seen as a cost center and not as a profit center so we in fact call this organization the Delights team and everything that they do is to create delight for users that's actually how how they're assessed and so the folks on that team are called to life specialists and that that's sort of the three layers of the growth organization now into this we're about to hire third demand generation for content marketing and then of course for management to tie the whole thing together let's list you one more question then we can close you wanna pick our oh I actually do know who you are my question for you is that so you have a say at SAS product whets reoccurring service but for let's say a construction worker or let's say a design firm or let's say like a tutoring company that is pretty much outcome based one time you use it when would you implement the survey would it be throughout the process or once they get the end of the product I think it comes down to the frequency of the transaction so for a tutoring service I'm guessing here it might be once or twice a week maybe would I be right in guessing that okay let students twice a week so after about two weeks of consistent you supa you've had about four transactions to me that would be the right time to ask the child probably also the parents in a separate survey assuming they're the one paying for it would you be disappointed if you or your kid could no longer use this product the key criterion is not before the user has experienced the main benefit now if it's a year-long construction project that might be really hard you might have to wait a long time but if you're selling tools metaphorical pickaxes and shovels there's probably a time within that period when you can make that ask all right let's give her a round of applause

Channel: Summit

Views: 10,912

Keywords: summit, conference, ideas, talks, performances, gathering, community

Id: BhDj_C4Qzhk

Length: 42min 6sec (2526 seconds)

Published: Mon Mar 08 2021