Building & Scaling Products: Lessons Learned from Four Years and 8,000 Customers – Des Traynor

0:00[Music] [Music] i'li talking about ten lessons learned specifically ten lessons learned since July 1st 2011 why did you lie first 2011 but that's actually when we launched intercom way way way back what basically probably I remember the day very clearly I was running from client meeting to client meeting and in the middle of the day and refresh Twitter I know it's like 100,000 people who felt like we're tweeting about intercom and the turn said Jason free to tweet event which was a which is basically you know fantastic on the day and then obviously queued up as [ __ ] over support requests so this is a sort of lessons learned talk today

0:52intercoms 240 people over 8,000 customers you know significance double-digit revenue growth month two months things have gone really well we are as was said at two officers Dylan in San Francisco and I'm kind of approaching this from what do I know know that I did not know then so some of this is lessons there in the hard way some of it's things that hey [ __ ] we try to them actually worked and hopefully you can tell the difference so it's kind of hard to cost it's a cost of my mind all the way back to the things that mattered versus things we pissed away a

1:23lot of time on that didn't actually matter in the end but I'll do my best to try and dissolve them the first thing this is probably like the most fluffy abstract but in my opinion genuinely the most important point is this idea of vision when we started intercom we had a very clear idea of the problem we were trying to solve for those of you who are into history of SAS you might remember a product called exceptional that was a that was our original product and what exceptional we had like thousands of users we could you know we know as a funny side not I often tell people there

1:56was a street in San Francisco where we had more customers than we did in all of Ireland which maybe told us we should consider a different office but we had this problem where we had all these customers and we had no way whatsoever to know who the hell they were I'm talking to them was a nightmare that if you can remember back in the days before stripe and good tools like that the way we actually talked to her customers was we'd go into at PayPal export to CSV sync that up with like some sort of analyst we should tell us who was actually using the product not just who's paying first

2:25and then import it all into like MailChimp Perconte monitor and said that a male and then you get like 150 replies in to your personal Gmail and they try to sort you all that it was a [ __ ] nightmare so hence we settled this intercom idea so why does vision matter I think without a vision the companies I worry about when they don't vision is that they tend to flip-flop and they don't have a good sort of you know base set of axioms or calculus that lets them make quick decisions it kind of reminds me with the Cheshire Cat scene and Allyson wound around where Alice comes to a fork

2:57in the road and the Cheshire cats there and she says well where should I go left or right and he says well where are you trying to go and she's like I don't know and he says well then doesn't [ __ ] matter which way you go unless you have a vision you don't necessarily know what it is you're trying to do and as a result every single decision will paralyze you because you don't have any sort of guiding light and to bring this to practical terms or when I talk about a vision I consider to have like three sort of core components firstly it's like an opinion about what the future

3:24looks like and that's this is important but it's like you have to look at all the technological trends that are happening at the same time not just the things you're gonna do but what else is going on and then the sort of the new tech that you're gonna bet on or the new assumptions you're gonna bet on so if you're gonna build a business based on intercom or based on stripe or based on something like that are you you're confident that you're gonna bet on them for the next 5 10 15 years or whatever and then lastly an idea of the change that you actually want to make so like

3:52given that all this stuff is gonna happen all these technological trends are solid the world's moving a certain way what do you want to do to change it and it can be different if you like you were in commerce or if you're a messaging app or if you're marketing or if your churn or customer success or whatever it is always a set of things that you should think about in the future so if you take around the more like marketing marketing is getting more personal and personalized it's now more important than ever that marketing can prove its value it's now more important than ever that marketing

4:19is metrics driven and you know we're seeing it shifts towards people who are advertising to start engagement rather than conversion so you're seeing a lot of brand advertisement of the internet which we didn't see before I like all of these trends are important need for you to understand for you to even begin to tackle your world and your world might be in might be project manager Andrew might be time tracking but even at that these things are changing as well slack is fundamentally changing the nature of project management you know remote work is changing the nature of like time tracking and all these sort of things

4:48like you need to think about these implications as they move forward an intercom our vp of product paul adams he refers it is frequently asked his 666 roadmap which is a it's not as evil as it sounds he says like your first six weeks should be pretty well planned and like you know task driven and you're in six months should be you know significant bodies of work and then six years should be a very strong opinion about what the world will look like in six years time and it's important to have all three said you can basically know where you want to get to know how

5:17to chunk it in bodies of work and then also know what to do today so we talk about tech trends it's interesting because there are like you know genuinely you know the world of technology is one where we just you know trove everything out and reinvent it every few years what trends actually matter to your business like this product you could call it a flop we call it a multi-billion dollar flop but everybody what I refer to it and it's not something that actually matters in your world if you're like you know is this relevant is the connected home relevant right this idea that like that

5:52Porsches and toasters form a coterie called the intercom of things or it was wearable so any of these things actually matter in your world and if they don't that's fine but that's your opinion and you might think yeah well you know there's I'm working on a finance app it's just piece of sass its Web only that's cool but like just bear in mind the teaser to trends are happening and if it turns out that they do matter that's gonna be a big problem for you if you've chosen to ignore them the biggest trends that actually does impact the majority of the businesses here from what little homework I did is this idea

6:22of the decreased cost to do anything online so order to compute to store to access anything whatsoever it's become cheaper now than ever as is always the case in tech to do is to set up anything so like you can't have a case where you have like four people running a multi-million or occasion that you know tens of million dollar business and that's because of trends like this and what that kind of gets us is just like explosion of software for example if you got all you know and single up here is a messaging app they all think they're reinventing the world of messaging I'm not even joking someone

6:56took the pain second time of color coding their entire home screen and I appreciate it I just copied that image wasn't my job but genuinely like I like some of these will go on to become like 18 billion dollar businesses like whatsapp some of the mobile become niche players for like specific uses but the one thing I always say when I see this like you know like three years ago I used to talk about photo ops being the new taking post Instagram photo ops was what everyone flocks to straightaway and the one thing they always seem to be lacking is a sense of what actually what

7:26change is what are they actually doing are they just copy carrying something that they saw work for example you know are they just looking and say oh you know Josh got to like 50 km or or by doing metrics on stripe I'm gonna do metrics on stripe and get to 55 cameras doesn't necessarily work that way you have to have a sort of a bigger idea but while you're doing that bigger idea to me is best captured as like why do you exist other than to make money and making money as I was the important it's what we're all here for it in a sense right well not here here if you want to

7:59thank one of you back at work but but what what is it you're actually shooting for and the reason I think this matter is this because if you stick with like goals or targets or if you go for like you know I want to be the number one or the top 10 or whatever the danger is one day you'll wake up when you'll hit that target and then you don't know what the [ __ ] you want to do and genuinely that will happen like every you know for sure every metric or goal I thought intercom could never hit we've hit and I've got lots more do we haven't hit yet you know

8:31optimistic we'll hit hit them too but like if your goal is 50 km or or and that's all you're thinking about what do you do today that happens all right I know that sounds like a first real problem in a sense but it really comes down to like what it is you're actually shooting for so I think it's really it's can be a quite an inspiring thing for you your your teammates your company if you intend to grow on to actually have a sense of the bigger picture and for us it's like we want to make an internet business personal stripe want to increase the GDP of the internet and

9:02when you look at projects like Atlas they were and he did you can see that that's a she is the guiding light instacart want to be the best way for anyone to get food or pay people to go to the store for you uber want transportation as reliable is running water but the key point is like I do genuinely think it will help you with your content marketing will help you speak about your product it will help you recruit if you want to recruit it'll help you inspire your customers if you actually tell them this bigger picture of what it is you're shooting for if you come straight out in your blog

9:32and say my entire goal is to hit 20 km or or whatever that's cool but just don't expect that to resonate we're not even older than you and your bank balance and that is a risk so there's a lot of value in having a larger picture when you have that larger picture it's obviously imperative to start small one of my favorite laws from systems engineering is called goals oh and girls are simply states that any complicated system that works has evolved from a simple system that works any complex system design from scratch will fail without saying is that even if you wanted to build a bloated piece of [ __ ]

10:05you have to start small yeah even if you want to end up in this big sort of monolithic mess you got to start small if you start big you're broken from day one so it's interesting when you look at even the biggest low deist things we could imagine they've all started small so if we look at sales for us today sales for us today have a couple of products you might say they've you know they have a reasonably large suite but when you look at where sales force started it's pretty pretty funny this is sales force when they launched and those who were really good eyesight will notice that they're giving

10:39away a zig Zigler tape to anyone who signs up for a trial and the fact that the cassette tape actually tells you how long ago that was because DVDs are probably expensive that was for me for premium trials or no wouldn't be invented yet actually so you know they did start small though word today is like most people's poster boy for being like the big bloated piece of software I know I always remind people don't you know check yourself before you criticize word I think it's the second most successful software product of all time ever so yeah they could do a few last tabs but like really you know but

11:14if you look at word where it started this is what word literally way it's the earliest screenshot I could find and it looks a lot more like notepad but they got there right em so I really believe that starting small with a simple thing is incredibly important that's hardly a new met new message - but maybe the way the way will go true this what I should be interesting it's really hard to solve especially a small team where you can't afford external research or it's really hard to solve a problem you and Nate Lee have yet to experience yourself so if you you know if you've yet like if you've never run a

11:53business building software for people who run a business is really hard because you don't have the domain expertise like you know if you've never acquired customers building software helps will acquire customers is really hard because you literally have you're totally blind to the problem so as a result your solution is not well not gonna be well defined it apparent it you know you know well when you can like understand a problem so intimately that you can distill it the specific workflows and you can build features that enable those workflows to happen your first release in this regard should be the smallest piece of software that solves the problem on at that point you

12:28should stop and genuinely put your text editor away when you think you've solved the problem the reason I say this is because if it all fails either the problem isn't real alright as in it might have been something that frustrated you but it's not generalizable so so you haven't nailed it in that regard you don't understand it so you're tackling a real problem but your understanding of it isn't there you literally thought you had a working solution and no one cares or your execution of it soaked so you didn't actually you know turns out there's a lower you met you offered a low-quality offering that no one stuck around for

13:03and obviously there's a fourth category here which is you forgot to tell the world that you launched or whatever but generally distribution in the sort of in there like 50 to 100 signups is like relatively easy to get these days with like web sites like product launch with things like Twitter even would like to network in this room you could definitely find like 20 to 50 people if none of them can stick around and keep using it like after a couple weeks you should really stop there I like the danger is you'll try and feature your way out this because especially if you're like a technical founder that's your go-to move

13:36for everything there's no problem that Emacs console right and our VI or whatever but him it's a it's it's the most tempting thing it's much harder to think well maybe I need to do my homework a bit better I was just a side note like a product we almost launched way back was this is the pre intercom we actually we're kind of on a knife edge of who we build intercom where we will be build this no it was you laugh when I see it it was have you ever taken a photo of a whiteboard yes you probably have right maybe you haven't maybe that's what the [ __ ] problem was but

14:14so we built this like a whiteboard or tool and we call it a white border and it was basically was an iPhone app plus a web app that basically every single time you took a photo would automatically go and be shared of your teammates if any whiteboard you ever did and you know we thought it'd work and we ended up getting into this sort of what I call like execution death spiral so that's death spiral looks like this I'm sure you've seen this before but lucky no one was using our product so we said well what features will make a user product and people are trying to be

14:42helpful so they give us lots of lots of suggestions as features we go and build those features and our payback for building those features is of course that no and uses their product and and you can genuinely piss away a good few months of your life chasing this and it's not fun right because users aren't like it's very very hard for for customers especially people who like if they are within your network it's even harder for them to give you a very honest answer of hey that problem the iPhone camera is good enough and you're like oh you know it can you know they really feel like they're hurting you so

15:18you have to be careful at who you ask for feedback which brings us to our next point which is you have to manage your beta deliberately beta periods are hard and like there's a whole earth to it that most people just blitz true totally ignoring that like there's a lot going on here and they're making a lot of decisions they kind of think who else wrong betas okay so the deal is how do you run a beta you put a shits on a page live you make some sort of vague offering you collect LOD email addresses go build a product Farmall dima or not spam its targeted

15:47the direct mail you email all the people I mean it genuinely isn't spam in that case they've actually asked you mail old people they all bounced our office we're doing right and like that's basically the most depth of thought that actually gets into most famous there's a few different problems people make along the way a small list of target users beats a long big list of that includes a lot of non customers you should really be careful the healin is your beta because you only want people in there who are actually the people do you think should credibly use your product and you shouldn't let the notions of who we can

16:22have 100 users by Monday actually influence that because you only want features sorry you only want feedback from people who are actually credible users and this is why whenever I see people doing something like we're gonna give away a free Apple watch to everyone who signs up whatever a give away a free Apple watch and you'll get people who want Apple watches they're probably not interested in your invoicing solution you know give away a free iPod and you'll get people who aren't there if you happen to be building something for people who want Apple watches that's great so for example if you're making apple watch straps or wristbands or wherever

16:53perfect giveaway an Apple watch they're the exact people you want but for most of us we're not doing that we're building software that's targets a specific business needs in order to Adriel fall into it this comes back to the visioning a little bit but if you make a big vague pitch about your product what will you get you'll get a big wide range of users right this product here is basically saying next to nothing about what it does how's it cool name I just invented it for this presentation obviously but but I'm if you do this what you'll get is a lot of people who are kind of interested in

17:29photography reimagined whatever the hell that is and that gets you a lot of feedback so you'll get feedback like yeah well I want to sell my photos or I want to export an integrand we're all better filters there I need to collaborate my earth director and you're gonna have these sorts of feedback all over the place right those four people are looking for four totally different products I'm none of them are probably looking for to photography reimagined or whatever the hell it does and but if you keep taking this feedback and this is this will feed your sort of feature bias of like yeah well we're just a few

17:56features away from being ready you're gonna go and build this beautiful mess like you you'll build this wide variety of features and you're gonna end up with like all this [ __ ] basically has sprawled on each of these ones here maybe I was like five users who actually care about it but that's like you know it specifically when you're in the world of like let's start with a small team and let's it revenue quickly this is the worst way to go about it another danger relating to this is just you'll always feel a one feature away it will always feel one feature away from going live

18:27and don't rely on your customers to tell you when to launch your customers are never gonna be like hey Liss and i'd love to end my free beta period and I'd love you to start charging me now very rarely you get those emails more so like you know sitting back waiting for your customers tell me I think you've got the product perfect they just won't write occasion like some of the more conscientious entrepreneurial users will but like hey heads up this is actually really good or whatever but for the most part you're never gonna get this email and you'll never feel ready I really like this quote by round three says like

18:58iterating means like you know keep going until it's great not just keep going because you don't have a good definition of what doing looks like and that's the danger there's there's an iteration cycle it's compounded by the fear of actually going live and testing the reality of your hypothesis that makes it real scary this is what I like I like people who can like it there's only really two ways to exit a beta in my opinion you can exit by Scopus and you're gonna say I'm gonna address all these features and we're done or you can exit by time is in November first it's going hell or high water that's

19:30happening right and the best ways to think about this come from people like Steve Jobs whatever who will say like this I love this definition of the DVD burning product in our SX where he said basically here's the app it's got one window you drag your window you click the button that says burn that's it and it was a very very simple definition of what otherwise would have been a monstrosity of a product and his advice was it has one feature it burns DVDs the next problem please and when you get that sort of clarity that's when you know your product is ready the funny ones have started in that life out here

20:03I've one of my favorite quotes ever is a it was in Tina Fey's book Bossypants when she was explaining to Laura and she was like well look we you know I've been working on the scene for a while it's just not quite ready and he's it ready the show doesn't go on because it's ready it goes on because it's 11:30 on Saturday night which is it's just a solid point like you don't get to [ __ ] around you know it's like 11:30 the viewers cheer tune in whatever you got is whatever is gonna go I like I really believe that you can accent my feature set or he connects with my

20:31deadline but above all you gotta exit one reason that really matters to me is the beta periods are really lossy and people don't realize the cost of being in beta and we certainly like when we had our beta period intercom yeah you name it like Airbnb Dropbox tasks rubbish they all signed up and then we it took us like seven months to get it like oh hey we launched a product at that point half - boo emails had like left the company and stuff like you know we lost so much people don't realize you would never design a signup flow that looks like this right visit a website

21:04you know like a confirm your email wait a few months get it get in the invite it's it like it you could not imagine that more a higher attrition flow for getting people to use your product done let's make them confirmed or email for something that they don't yet know what it is and then let's wait a few months and then that's mailed them again and try to re-expand ourselves you know it's it's a it's not a great period to be in so be careful the attrition is just high and most frequently what we see is like you know I'm sure we all get this we've

21:36all gotten some mail where it's like welcome your trial of snap cycle is they're ready and you're like what the [ __ ] is snap cycle and you don't realize you signed up for it like you know seven months ago and you saw a beta page on product hunter something like that right and they're like they're like yet we're ready are you ready and we're like ready for what like you know so like that's the danger I think sort of a connected point is like during these beta periods I genuinely think like you're better off staying relatively quiet a bit about your product it's it's kind of easy to

22:03let your mouse and kind of income in conjunction like your ego ruin away or as you start telling people I've solved all the world's problems wait until you check it myself - were there danger here is like first impressions do count if you ever intend to get coverage on buy coverage I say press here it's just as true for product - hunter hacker news or anywhere you actually want to share it go talk whatever you'll only get really one stab at that and lastly the less you say more people listen so if you're forever talking up this product people eventually tune it out and they're like

22:35dude look mail me when it's done but so let's say we get you all that and people actually do start signing up I genuinely believe and we have an entire team in intercom that works on this specific problem onboarding it's literally the most important thing you can get right after your product is letting people get to experience the magic of your product and the most common phrase we use and I've already heard it twice at this conference is like you know we're in this dogfooding period and I believe it you're probably dogfooding your software but there's a difference in dogfooding your product as it as you think it's

23:11gonna be used on dogfooding signup so the one thing I often say when people look they are forever sharing all these beautiful screens of what their product actually might do one day signing up for your app is the only thing that everyone's guaranteed to do it's literally the only float you can guarantee everyone will go true right it's the highest leverage thing you can work on in that regard because everyone will go true and yet most product owners to sign up for Apple once and when they did it it looked like this right the risks here are like your tur ah you know as you build out your features and

23:53you complete your your idea of the product and what I can do whatever tier you've worked on for people who look land in your product it will go in a date because you're pitching features that are no longer important under you've worked on more important [ __ ] since then that you're no longer talking bit in your onboarding your docks look where today your welcome email is probably out a day that's probably asking to take the wrong first step or asking not to do the coolest way or the best way to integrate or whatever your video so go today all your marketing material goes at a date and intercom

24:20will genuinely one of the probably one of the most significant projects in the history of the company was we looked at sign up finally and this is why we built a whole team around we looked at what was happening when people signed up we were trying to work out why is it like take a room like this everyone here signs up for intercom why is it so many of them bounce and you know if the answer lies a lot in fact they're not all necessarily engineers and the way the sign of flow we had designed required you to be an engineer so what and where people who aren't engineers

24:52have typically is credentials for a third party product like a MailChimp or Mixpanel or a stripe or someone that they also have CSVs of their user list because they tend to email their customers they tend to be in charge of the other half of the business we built a CSV importer for intercom and it improved our conversion rates by 48 percent and that like you know to put in perspective let's say conversion was two percent it went to like three percent or ever that's a huge huge impact we've compared against people who av-test green versus red buttons or whatever like this is huge and we wouldn't have

25:26got that realization if we didn't obsess over what this new experience was like as a whole talk about that on our blog you can have a look at but the the thing I'd really encourage you to think about here is that like real dog fooding means you're actually eating a new tin every day that's a dog stay alive you know and they don't eat one thing once two years ago and then you know live off so students since them so never stop signing up for your product genuinely as I said we literally have a team that does this every day every week all dating about is how can people who hit

25:56that email address and click sign up how can they get us successful as possible when you're out there you're gonna start looking at the market and you'll you'll be tempted to like look in a gutter product category is there look for people who are directly competing with you so if you do invoicing or time tracking or churn reduction or funnel analysis or metrics you're going to look for other people doing the exact same thing I think these are my real competitors that's not always the case and in fact in most of my experiences being it's rarely the case that your competitors are exactly there your competitors tend to be something

26:28somewhat more orthogonal I often ask people like when the flowers compete with chocolate right one of them is in like the confectionery industry and you'd ones in the horticultural industry right they're two different categories but we can all imagine a scenario where we've said we need flowers or chocolate right so as you buy butter them together supposed to complementary goods but they do compete if you're trying to send a thank-you or standard congratulations or saying I'm sorry or whatever that's when you're literally deciding one or the other and yet I doubt anyone in like in Hershey's is like we need to get better at beating at these 12 roses right when

27:03does Twitter compete with games you're bored on your commute to work you know again Twitter thinks you know it would be simple for Twitter to think that like their computations Facebook but it's not it's only thing that gives a quick hit of interesting information in space to three to five minutes or what really competes with TripAdvisor all right is it like Yelp and Foursquare or is it like asking a human out in the street hey where should I go for a drink around here these are all sort of questions you know like they're like trees sort of examples that won't necessarily be like what why you shooting a bit of the way I

27:34would encourage you to think about it is like you have like direct competitors which are people trying to address the same market with the same type of marketing at the same type of product there are indirect competitors and there are like sort of tertiary competitors tertiary creditors are tricky because they'll have conflicting offerings like for example McDonald's compete with gymnasiums all right they don't realize that what they do it's like you know it's like if you believe in one of these parts you're unlikely to believe in theater okay or you're just very very disciplined um but like every single note taking up I've used has lost has lost its fight or

28:09sorry not note-taking really kind of tasks it's lost its fight to me putting post-its on my monitor that's literally my number one way of getting things done and whenever I quit on like Wunderlist or to Dallas or any of the things I never asked hey what why didn't it work out I'm like you know it was just always easier to write a post and stick it on my monitor and you know we could get into it and we could sort of say well what is that like it's a direct real-world notification if you like it stops me from doing order work until I've got this thing doing I'll never

28:35forget it it's a you know it's it's a ferocious competitor for a lot of the products in this room email is a ferocious competitor to I frequently email myself ideas for blog posts there are whole products that think that I should use their solution for this and they're probably right but like it's important to work out what is it that makes email so strong for these things on in general like if you know I on some sort of law of averages I'd say most of you have emailed as competitor in some sense just plain old email and it's worth thinking about why that is and when your customers quit already

29:10definitely onto a competitor or they go back to their old habit which was like actually I'll just email the customer manually or actually I just like send myself an email as a reminder or actually I just email people instead of slacking or whatever so when you have competitors it's always useful to know what makes people switch so there is a weird moment in in the world where like a customer's going along there but daily daily business one day and they caught some of your marketing maybe they saw your product the driver and for those of you who've acquired customers they literally made subconscious or conscious choice I'm going to change my world so

29:48that I can start paying you money for your software that moment is best regarded as being some sort of switch product this switching moment every time a customer adopts your product they are quitting something else they're like quitting on a bad habit or they're quitting on post-its under monitor or to quitting on email or to quitting on not doing something but they're changing something about their life and the best way to think about that is this nature of the switching moment and just four forces like so I don't know if any of you are familiar with the jobs to be done method but like this is a really interesting way to

30:21think about it what makes somebody switch there are four forces at play on the left-hand side here what we have is your current life most people don't switch most of you who are using mock books here you're not going to switch right people don't regularly change products your life will be a nightmare if every see if ever you have to like go from first principles and every single decision you make so you just hard code some things I drink Pete's coffee I use them I phone I use a MacBook whatever the forces are play here the two green ones are the forces of progress the first the first

30:53of these is what is [ __ ] about what you're currently doing right and the second is what is exciting or attractive about this new offering and then below the line are the forces of stock the stagnant forces these are what keeps the world away it is what am I scared about about adopting this new product and then what do I actually like about where I am today and when you're trying and it's really worth doing this like sit down with a pen and paper or whiteboard and work out what are our competitors our real competitors why do people stick with them what what attracts them to us

31:25what are they scared about from us and what pushes them away from their current for us well the current solution sorry and what you want to do is maximize the top forces you want to say yes your solutions [ __ ] ours is [ __ ] amazing you shouldn't be scared of it and also all those reasons that you're still using it they're all irrelevant right and the best campaign ever that did this was this one I'm a Mac I'm a PC they were so strong on this idea of PCs how viruses they're slow you have to change them every two years you have to reboot

31:58them all the time Mac's have no viruses they are amazing they are fast they're beautiful looking and then like they went at the other side don't worry all your all your office tools that all works on Macs you can do spreadsheets on Macs all the things you like about windows that you can have all dem - they addressed every single one with such a broad media campaign of every angle and like low and behold when they started this campaign it wasn't true but like now I'm looking at probably easily a hundred opera logos in front of me here today one interesting thing to notice is this area does this like sort of

32:32principal business call it the 9x effect most most of you over value how good your product is by bite reacts and meet it and most customers tend to over value what they have by about 3x so your the tables are against you you think your you have is amazing and they think that they have is disproportionately better than it is so in this case here if you look at the new new is like slightly a better offering than old but the 9x effect makes it like not not relevant basically it's it's what makes it pretty pretty tricky so this is like you know I like you have a pet belief I've know if

33:11you ever had Paul Buchheit in a bit like you know 10x is what it takes to really motivate a change I you know it's like pseudo [ __ ] calculus whatever but I'd like to think that this is what's going on here to actually be really better you need to be so significantly better such that people's resistance to change and your own inflated sense of what you've created or like actually washed out because the product is so much stronger so people are switching to your product you're talking to customers and they're asking what are you gonna do next so product roadmaps are in our interesting and dear to my heart is probably

33:46something we've talked about and thought a bit more an intercom than any other single topic it's you know when you start playing at Europop it's really easy to justify new work and it's really hard to justify revisiting old work and the forces up later are very obvious new shits interesting and there's no legacy code you have to worry about right you got to do whatever the hell you want um the the problems here are that you're not thinking about it from your customer's point of view you're thinking better from what's up phone what's a fun thing to code or what's the funding to design or what's a phone workflow to add

34:21if you have a product and most you do it's really important to be able to do this I've been talking about this matrix for about two years so I won't go on too long but there are ways to think about your features that are pretty simple how often people use it and how many people use it and the simplest way to think about this is you know if you take something like whatever screen people see in your product right after login that's something that all of the people use all the time so making a small improvement there versus coming up with the world's best reset password flow is like

34:58disproportionately different reset password is some of the people some of the time you know post login is all of people all of the time so it's really important you can do this and when you're thinking about your features and what work you actually want to work on it's really worth asking yourself am i working on the most important things for my customers in the top left here at the hazard science this is the area where you end up building small niche features for customers who might even be promising you a big like I'll pay $500 mode if you add this one thing if it's just them if it's just those customers

35:38you're gonna end up building what I call consulting where that is you have a solution but you've basically turned into a freelancer who has like maybe 80 customers and each one it has a different app and you just plug them all together into one and that's a really dangerous place to be an especially if you're a small team because you don't be able to maintain at all you're you know you you know basically still for death true attrition of like consistently everyone wants their own features worked on a good way to think about your roadmap is like you should have four types to work in your product number one

36:13is you should be improving the features you already have and you can measure that to an increase in solicited customer satisfaction what do you do that true NPS are qualitative feedback or whatever way you want to do it number two is you should be getting more people to use features you've already built getting people to use features you've already built more and lastly and this should be the edge case not the default case when you come into work building new features to attract new workflow so you can hopefully grow your customer base or at least grow your revenue and that's genuinely how I think about it

36:43when you're tackling this there's a lot of trade-offs with feature work that you have to consider one of them is like the trade-off between quality importance satisfaction and frequency and this is especially important when you're actually looking at your existing product quality is basically asks how good is this thing today importance is how important in the users workflow is it is it like a nice-to-have or is it like something that's actually core if it breaks you know the whole product goes to [ __ ] how satisfied are day today and then how often do they use it and honey to retrieve these can give you an

37:17indication of how important something is to work on if it's important thing that people use frequently but they think it's bad yeah that's something you want to work on however if it's not an important thing but people are pissed off at it it doesn't really matter and I like these are all just ways to think about it right another way to think about is I Tony Luke's framework and he is like the importance satisfaction sort of index right and what he says is that basically you need to score features by how important their on score features but I have satisfied your customers are and that tells you where the

37:46opportunities to improve are so this will stop you from working on [ __ ] that either isn't important already you're already nailing pretty well both of those things are like are always tempting right literally the highest leverage thing you have as a founder as a founding member of a product team is picking what you actually do and what your company works on if you're doing if you're not picking it right or if you're stuck by any sort of naive bias of that let's do the new phone [ __ ] you're throwing away the opportunity that you created yourself so when we talk about getting things used let's say more or by

38:24more people a target feature might be here right maybe 60% of people use it maybe on average they use it like once a week or some other and maybe in your world once a week isn't enough right obviously you can't get people if you're invoicing you how can people the invoice anymore done they need to invoice right so there's obviously real-world pressures here as well but it's good to ask yourself how could we trigger more usage and if you're familiar with any of near Royales work he is some pretty solid material on how dashi it you know get users into more habitually using your product for keys things and he

38:57talks about things like how can we make it more of a habit how can we create more triggers what if it notified slack every time this happened or what effect we sent a morning mail that that told people here is the tree actions you need to complete or whatever what sort of rewards can we do how can we make it to default that people do this how can we make it usable in more context what if it was like easy to do on the phone what if you could send an SMS would do it automatically whatever how could it be used for more purposes these are all good ways to ask how can we get

39:22people using a feature more and then the other side is when you want to increase adoption of a feature what's stopping people from using it today and it's easy to say like it's easy to take customers first answer here but it's really important to think about it from a like from what the classic sort of the Japanese method would be like the five wires right understanding the real why so they'll say I don't really want I don't need it and you say well why don't you need it because you believed it that they needed it that's why you created it they say I don't get any value I don't

39:51know why well it doesn't present well meetings oh why well I want to send her in a doc and present a PowerPoint file and what you've given me as a screenshot of some stuff let's go right what's limiting there well why why why didn't we build a image or a report or something that could be used here oh well because the data is like hidden mind all the UI controls or whatever okay so now you have an actual hypothesis right this is the deep reasons not using this is because they the jobs to be doing is help me impress somebody in a meeting by sending around

40:20a report and you're not enabling that because it did think that saying around is a screenshot from the drop-down menus which looks [ __ ] so that's one example of getting to the real why and that's an actual real example from my career what you do is you do enough of this and you'll get to root causes for a lot of reasons why things aren't getting adopted by all your users it's fair to say not everyone needs to use everything you've built but it's also fair to say that most people should be using most of the things you've built if you're not there you have a problem so when you did

40:53what you do is you get these sort of barriers to adoption that's what I would call them and you rank them when you resolve them and something I often say to people it's like it's tempting to believe that like every single one of these things needs more lines of Ruby or knockout or whatever that's not always the case sometimes you just need better education or you just need to communicating things better you need more tool tips you need slightly simpler UI whatever but you can easily communicate here's how you can do this thing properly or whatever I'm not saying that you can communicate yourself your way out of everything but it's you

41:28know it's a good test before you go and do a of programming work to say well what if we just tell people had to do this it's a great way to validate your hypothesis of yes I need to rewrite this feature or simplify it or whatever one thing I would say we're oh drops is it's really tempting anytime you have a long list it's really easy to a dude a classic sort of order list which is here's a list of work we need to do and here's how long each team will take here's our team and we'll just do that and what that basically ignores is a small thing

42:06called how much customers actually care and I look I read the story from the Google Docs team where they did one of these exercises where like let's just add all the problems and let's see what we work on and they would hit a load of improvements they made it easier to do page formatting and all this sort of [ __ ] and and still customers weren't adopted like what the hell is going wrong so they tried a different approach they brought in 100 users and they basically said here's everything we need to work on here's some fake dollars and we want you to put fake dollars into the

42:39cups of the features that we you want us to work on and a really interesting thing happened all of the money went on to a formatting bold italic and underlined which is basically how most people people still today don't use like heading features in Microsoft Word nor do they certainly don't use some Google Docs and so all of the P all of the money went there the best part was one of their target customers went into their own wallet and took it another hundred and put that in there as well right because what they didn't realize was this thing right just cuz it's on a list it doesn't mean all lists are equal

43:10right people can care of disproportionately more about one thing and that's okay and if they do that that's the thing you should tackle it's very very easy to lose priority in an ordered list and and just you know be mindful of that only time you're putting a list of things out there make just number two and number one could be light-years apart when you're shipping new features which as we said shouldn't be the default every week because he was gonna grow massive product you don't necessarily want you need to think about how you market the most features actually flop right most new features flop including all the areas including all of

43:47ours unless you take a real extra effort to make sure and communicate what the value is my favorite example of this comes from Microsoft where they sell fish to boast about something microsoft noted that turns out everything that our users are asking for in the next version of Office we already have in the current version of Office so they put out this press release as if to be like well one of us yeah congratulating the fact that they had basically you know created a monstrosity of a product where everyone was asking for stuff that was already there but what's going on there but what's going on is is kind of

44:22straightforward they're you know they're rolling these things out but they're not doing anything to try and get people to use them they're not like for example some some standard things they're not selling what the feature does instead of selling what it's composed of they're like and this ground-up reruns let's just take the following things rather than you can now achieve X for your business it's selling the capability that your product provides is way more impactful than selling the ingredients it's composed of when you announce a new feature you always have to follow up in context and get people using it you have to fight really hard for usage here so not

45:03include things like you know where if it's there in the product there's a right way to pitch it you know call it them but like here here's the new thing we added it's awesome it's gonna let you do X or whatever okay the alternative here is like you know sending us you know a badly targeted message to everyone about something that you haven't seen announce it in context is way more important probably the biggest crime here we all commit is we push a huge amount of effort into telling like we blog post customer email everything saying here is our new thing and that works for all of our current users but

45:33then a whole new set of users go and sign up and they didn't read that blog post and they didn't get that email today they don't know it's there so it's buried somewhere and the only way you would ever ever found it is if you happen to be around on launch day which is a [ __ ] way to be right and then lastly you have to be like fastidious or passionate but just chasing their feedback so if people aren't using your product you need to find out why if people aren't using it you need to find out why I need to be following up if somebody uses the tree

46:02times and never again you're like hey just FYI we're gonna be working on this soon and I noticed you tried it out but like you didn't get you know you didn't stick what happened what went wrong what does it not do you need to get this feedback huns consistent he back into your roadmap it's easiest thing in the world to do is move on to the next feature well it's like the worst thing you can do and I would say like that at different points in your company in your company your roadmap and priorities will shift when in the early days you're for sure your talk about doing new features

46:31once you have new features out there you're talking about adoption once you have adoption you're talking about improvement this is how you curate a roadmap and it took us a long while to get to this sort of started working but it serves us really well intercom was a lot of software and to basically manage to consistently work on each piece it takes a lot of discipline the last lesson that I found incredibly valuable in the past few years has been this idea of consistently resetting your opinions it's true that startups move ridiculously fast even even if you're deliberately telling the world we're gonna take this slow we're gonna take

47:12three years whatever the speed of decisions and the speed of like progress is still quite quite high in comparison would say that I could coffee shop where it's a motor type of business and every decision leaves it's like is you going down a certain path right so when you look at say like the history of uneven companies it this could be intercom we made some decisions early on every day every day we make more and more decisions will we do this will lead it up should we should we hire this person should we grow this department the reality is like when people tell you the story after company they tell you the

47:42path they took but they don't tell you all the decisions that they made they so you get given this kind of linear sort of story and it's often presented like a linear story two people tell you if it was a twosie like it was you know if you look at say movie like the social network was like well he had a fight with his girlfriend so you set up Facebook I know he's a billionaire you know but like I'm but it's a none similarly like you know if you like talk to them to like even you know modern companies that we all look up to and

48:10admire what they'll give you is they're not say you know that genius moment we had that was actually a coin toss we were actually gonna do something else like you know people want to present this idea I like it being a fait accompli like it was like of course we were always going to succeed we're magical and because like you know and the theater and knock-on effect of this is that because we when we rehash our stories we tend to think it is like linear guaranteed positive progress we tend to think the future will look like that too so we kind of plan it the next

48:39set of roadmaps ago of course we're going to launch to do features and then once we launch them we're never gonna eat revisit them cuz we're gonna get it right straight away and then we're gonna move on to the next feature you know so it can really bias you this idea of forgetting about like the things that the decisions you made along the way because some of those decisions weren't necessarily correct and not I mean probably the ugliest chart you'll see off the entire micrograph is this one yeah but the point being you're getting smarter about your business every single time you move on you're getting smarter

49:09and smarter and smarter you know a lot more today than you did two years ago and you when you started you all along your path and some of those decisions you made way back they're not the right decision anymore when you get new information you can change your opinion it's okay you know you're not you don't look there's a weird thing where people are go we'll treat years ago they as you said the blonde I'm like yeah three years ago I did say that I thought that was true damn I think this is true now there's a weird pressure you're under like to like to present yourself as just

49:45like permanent cast and stone person but changing your opinion is a really valuable thing if you knew them well you know now would you still have built our feature we're just sort of ship that integration or chose that architecture design that screen that way hired that person to open that office you can keep going on these things all right but the reason I can encourage people to reset their opinion about its custom mistakes you know about today but I are not busy correcting there are the ones you're making every single day you're just remaking those mistakes over and over again and that's as true in your

50:19business as it is in your product as it isn't your roadmap as it is in your life you're saying in the beginning your talk that during your beta you shouldn't necessarily be listening to which features the users find important because they'll just tell you you know to be nice or whatever but the later the talk you're saying that we should you know let the users prioritize which features they want so how do you when does that switch happen where it does make sense to do that so the difference with a beta period is you you're like you're working through a speculative idea so what I said was like

51:01specifically people who sign up if they come from a wide variety of backgrounds and they're trying to use a product for a wide variety of things it's a real dangerous place to be to let that drive your roadmap it's a much better place to be to let people who are using your product the way you intended this be used and using features the way you have used to take their input on where you need to improve features it's like it's kind of like this DeLaughter is grounded in reality and the former is granted and sort of hypotheticals but I think like in general like I'm kind of I've often

51:31said people like you know if you want to know how to improve your calendar talk to people who are using your calendar every day don't talk to people who've never necessarily used your calendar at all or people who don't even know do you have a calendar and ask them what you need to work on like it's it's always that the pressure is always grounded in reality and reality usually looks like usage the alternative is when you're talking to people who say won't adopt a feature first of all be sure that they're actually getting this job done elsewhere before you listen to their feedback because you know if they're not

51:59doing it at all like let's say you have a like a payroll system and you're trying to work out why they won't do it well maybe they're just not doing payroll at all but you need to find out hey but before I ask you about r1 I need to work out what you're currently doing so at the very least you're grounding yourself in the fact that they do have this problem in the air solving your true some what it means then it's a better conversation however why not ours so for me that the whole thing is just grounded in reality grounded and usage grounded in real world problems that are

52:25actually being solved today can you talk a little bit about changing plans and pricing as your product evolve Zand managing user expectations yeah yeah so I think there's a whole talk you could give on pricing with it there's a I believe that when you you know there's a few core beliefs that I have a better pricing that that I would encourage people to consider one of them is your product will we get better over time better it's best to find it's more valuable so I do believe you should reconsider your pricing just like a as I said you should reconsider absolutely everything about like your product in

53:11your system if you find yourself moving off market what you'll find is by oak market I mean I say attracting larger customers or attracting customers who expect more of you in terms of availability service the elongated sales processes any of that sort of stuff you definitely need to like have a way to capture more value they're probably the most broken thing that this industry can tend to do is offer unlimited on cup say Enterprise plans which is basically giving the biggest discount you can to the richest customers you can which i think is totally broken so like there's a few different things at one I think

53:47you should whatever we consider your pricing you should absolutely plan on changing your pricing it's it should not be set in stone now we can talk about grandfathering and migration and all that sort of stuff but like but ground zero for me is like you need to plan that your pricing will change you don't know everything when you've set out your first set of pricing you literally don't know how to product to be adopted or used or utilized or devalue a little deliver so you need to plan on changing it and that there's implications there in terms of how you actually say integrated or what sort of pitches you

54:14make you should definitely charge more than you but we think you should I find out consistently true it's unlike plenty of people have said this before and some of people in this room have given much better versions of that pitch and I have but I think we are creating software that is being used by businesses around the world businesses will spend routinely you know $100 on taxis in a week they'll spend $85 on like coconut water and you're coming in okay $19 a month for how they had a like managed your roadmap it's you know I think there's a real weird discordant thing where people believe

54:52that going cheap will get better adoption that's I've never seen that to be true whatsoever I think there's a you know the best way like so one of things we learned with the intercom was like when we started we had one sort of monolithic you could you could buy intercom and that was actually quite prohibitive for a lot of our customers because they didn't want to use everything they wanted to use specific flows and yet we found ourselves trying to capture all the value we'd built for people who only wanted to use a sliver of the value so we made it possible for people to buy specific things so they

55:29could just buy our support product or buy or acquire product wherever and that was really it was really impactful for a very simple reason more people were able to acquire asserted opt our solution because we you know because we didn't feel the pressure being like we're we're displacing MailChimp and Zendesk and like you know lark and all these tools we should be square our price to be similar it's somewhat similar to all of those things together when we got out of that mindset I got into a mindset of let's let people buy what the piece of the product they want to buy that was a healthier place for boto us and our

56:00customers the sort of the generalizable principle there for me is like make sure you're not trying to capture value that your customers don't experience but make sure you captured the value to all your customers experience unlike this there's lots more I could say about that like an intercom we ended up building around pricing engine um or sorry our own billing solution reason for that was like I believe that if your product is unique in the world it's a genuine you know you should not just simply think our product you know you know on one hand you're marketing selling the world we have a whole new solution to the

56:34problem but then in the other world you're saying but when we go to market we're gonna go gold silver bronze or like small medium large whatever we're gonna adopt the same pricing model at the entire industry house I don't think that's necessarily smarter I think you should consider changes like metered billing you should consider things like you know different types of products or like or different routes to adopt one or one piece of your product or whatever all of those things are worth during upfront especially in a business especially for boots are after we're like your we're like revenue is actually heavy you know its core it's one thing

57:03if you have millions of dollars of funding you can actually afford to fuss around and you know delay these challenges for a while but like if you're looking for revenue upfront you need to have good ways to capture it so I mean I could go on and on and on about this but does that kind of address go rob has a question first then we have another one over here check-check alright I have a question for you you mentioned that it's a common mistake for folks to launch a feature and then just do a simple email and then anyone who signs up after that never gets that

57:38right so obviously that's mistake is there a tactic or something that you guys have done it had intercom to kind of introduce you know people who sign up after that to your features yeah sure so I think every time you add like there's a checklist that we kind of go to after we add a feature to the product and then includes all the boring [ __ ] like is it explained in the docs is it on the marketing side all that sort of stuff have we reached out to all previous customers who asked for this and let them know in your house there's obviously that's where the checklist is

58:08useful but then part of that checklist is how core is this feature to somebody experiencing the delight of intercom so so random example we recently shipped a facebook integration that means we can integrate your Facebook page and pull in all your cameo messages there and other to the support inbox like it's a kind of question for us is how much we want to pitch that but because of where it's located it's in like settings integrations they aren't in types of integrations and in facebook no one's gonna still mill into it right so the decision we make is how do we decide at what point to pitch this on

58:43how is it more important on say making sure people send an auto message or you're like uploading email template right there's a lot of things you kind of have an idea of what you want customers to do so we basically have a sort of you know type of framework where we work out what you know and we call them P numbers and everything like P Zero is like every needs to know about this forevermore and that we like to say you know a new route to sign up or new super feature I like Petri will be like we have this but people should you know they're gonna go looking so like

59:13maybe like a Salesforce integration it's like v3 but the tactics specifically is like when you know how important it is versus other things you want people to do you should have a plan that says okay like is this something people need to adopt in their first week month whatever and you see I sort of have a sense of a time horizon of when is this important and then you you basically have a message that pitches what the capability that you're going to enable is so again not selling like we have a Facebook integration what's but what's selling like keep all your Facebook messages along side you use masters and keep it

59:46all in sync you know then it's just a case of but where does this lie in our priority system versus order stuff that we want people who've yet to adopt this feature to know about and it goes into it like one of our campaigns which is you have to be released feature but you'll see of it and it'll just get like you know they'll get an in-app message if they're in the product they'll get an email if they have gone a specifically long period of time without still adopting it or ever but like the under criteria for who gets this is people who we credibly believe

1:00:14can get value from this and people who are well yeah that's Bracy number one it's like who do we think should be using this that isn't and how long as an acceptable wait time versus all the interment messages we want to send to say oh and by the way should our Facebook and then the last piece is when you have that message live you have to look at the success rate of that message so we know for example to Facebook teeny like because it's it's a free add-on I think it's like 75% of people who get that message will go ahead and do the Facebook connection if they have a

1:00:39Facebook page and then for you 25% we want to work out what the hell we got wrong like what would have made this easier why aren't weren't using it what else are they using it said all that sort of stuff so that's kind of too full run true it's a bit tactical be kind of got to do and we have time for one more question over here it is hi on the question of onboarding I'm curious if you have any thoughts or expresses around re onboarding may be dormant users or dormant customers people that you're looking to re-engage who does that in your company or do you have any

1:01:09special nuances or techniques around that for sure a few a few random thoughts and this ism it's a hard target ik people who know your product does and used your product and I can quit your product it's very there that's a much worse position to be in when it go sermon on somebody who's never heard of your product all right you have to kind of reignite a lot so what we typically do so we used to like try and do manual reach it's I think genuinely in the earlier days when we like we ran our customer numbers were in 200 so whatever we would reach at people

1:01:43who would quit and work out what went wrong I'm like I don't I'd say like maybe like five percent of the time we'd get them back but the other ninety five percent we get learnings and learnings is really what the bigger volume I in my opinion of these two things however today when we wrote a substantial new Alec like I said earlier like P 1 or P 0 type stuff we will take that opportunity to mail the larger group who after who've yet to unsubscribe if it's like and say here here's a few more capabilities we have now maybe this maybe this now is that a

1:02:13tipping point where you can come back to us but like obviously when someone quits as well it's it's important to know the context or the reason why they quit so if they quit because we didn't have an email templates feature yeah we'll follow up when we have that feature but if they quit because you know their business fails you know there's no point in following up and being like hey well if we know Vemma sorry about your business but we can never do email templates you know so like so it's a so I guess the generalizable lessons here is like understand and like we do this and

1:02:39intercom we record consolation reasons on to user records so understand why somebody quit or didn't use your product at all that's one whole piece and then the second piece is like when you're following up make sure you have something new to say because you're actually trying to encourage a whole new set of fresh eyes and you won't get off for something tiny so you neat you know you're better off doing when you're something bigger or a new offering for them that will actually you know trigger a refresh refresh [Music] you