How To Talk To Users | Startup School
Y Combinator2022-12-01
YC#Y Combinator#yt:cc=on
143K views|1 years ago
💫 Short Summary
Gustaf Alströmer, a Y Combinator group partner, discusses the importance of talking to users and future customers for startups, covering how to find users, interview them, and turn conclusions into a minimal viable product (MVP). He emphasizes the need to learn directly from users throughout the company's lifetime, using Airbnb's early experiences as examples.
✨ Highlights
📊 Transcript
✦
Good founders talk to users throughout the lifetime of the company.
00:01Talking to users is essential for learning and growth.
Users and customers keep the company honest and provide valuable feedback.
✦
Potential users can be found through personal and professional networks, as well as online communities.
03:25Interviewing people within your network may result in less honest feedback.
Early users can be found in various online platforms and in-person events.
✦
Interviews should be conducted over video calls, and it's important to build rapport with the interviewees.
06:11Focus on listening rather than talking during the interview.
Example of a potential customer interview about carbon emissions.
✦
Good interview questions to ask potential customers.
09:45Ask about the current process and challenges.
Understand the importance of the problem to the company.
Observe the customer's behavior and ask for clarification.
✦
After understanding the problem, the next step is to create a hypothesis for the solution and test it with the users.
13:10Ensure the problem being solved is valuable and users are willing to pay for the solution.
Keep the interviewees involved throughout the process and make them feel special.
00:01 foreign
00:03[Music]
00:09 my name is Gustav and I work here at y
00:12 combinator as a group partner I've been
00:14 here since 2017. before YC I was a
00:17 practically to Airbnb and actually a YC
00:20 founder back in 2007. today I'm going to
00:22 talk about how to talk to your users and
00:25 your future customers here's what I plan
00:27 to cover today why the best Founders
00:29 talk to the users throughout the
00:31 lifetime of the company two how to find
00:33 your users and how to talk to them three
00:35 what questions to ask them and what not
00:37 to ask them and four how to turn your
00:40 conclusions into an MVP most people in
00:42 the world have the idea on how new
00:44 startups to form completely wrong they
00:46 think ideas of new products is something
00:48 the financials come up with on a lazy
00:49 Sunday or a late night coding session
00:51 you probably know it doesn't work this
00:53 way this scene as you can see here here
00:55 is a scene from The Social Network movie
00:57 it's a great movie but they got a lot of
00:59 things wrong on how Facebook actually
01:01 got started this is Brian chesky Brian
01:03 is the co-founder and CEO of Airbnb on
01:06 the left side of the photo is Amal Amal
01:08 was the very first guest on Airbnb they
01:10 actually spent a weekend together in San
01:12 Francisco when they'd be kind of watched
01:14 you notice the difference between this
01:16 photo and the previous photo well this
01:18 photo actually is a real photo it
01:19 includes a two-way conversation with a
01:21 real customer great Founders talk to
01:23 Future customers before they even have a
01:25 product in fact the very best Founders
01:28 in the world learn directly from their
01:30 users throughout the life of their
01:32 company but that's again if you have
01:34 correctly identified who your users are
01:36 you should learn from them throughout
01:38 the lifetime of your company why you
01:40 might ask
01:41 users and customers will keep you honest
01:44 they are the only stakeholders actually
01:46 paying you anything if anyone will tell
01:49 you the truth it will be them so my talk
01:51 today is about how to do this well so
01:54 let's talk about another m b story in
01:56 2010 um and I remember this uh myself
01:59 Brian chesky the CEO of Airbnb he'll
02:02 decide to launch an experiment he was
02:04 going to give up his apartment and live
02:06 in 50 different airbnbs over the next
02:08 couple of months you might ask yourself
02:09 why would anyone do that so much hassle
02:12 most people read the articles about this
02:14 experiment because he did announce it
02:15 they missed a point
02:17 um I personally remember this being kind
02:19 of crazy by doing this Brian will
02:21 literally get a chance to talk to every
02:22 one of these 50 hosts each single day on
02:25 this trip and one incredible source of
02:27 honest feedback on your product without
02:28 the Airbnb hosts MB would not exist
02:31 exist today what Brian Joe and Nate did
02:33 so early is deeply understand the
02:35 motivations
02:36 um and the feedback from their hosts
02:38 still today Brian and Joe gets calls
02:41 from hosts on their personal cell phone
02:43 numbers I remember them telling me that
02:45 that on their first day of adworks at
02:47 Airbnb
02:48 why do they do that well it's the number
02:50 that they put on the website and share
02:52 with all of their hosts early on they
02:54 weren't hiding behind Anonymous email
02:55 they wanted to hear from the customers
02:57 at the time this seemed almost radical
03:00 most Founders would spend their time in
03:02 front of the computers trying to find
03:04 scalable growth channels like Google ads
03:06 unfortunately most Founders still don't
03:09 build personal connections with their
03:10 users today behind behind info at emails
03:13 or don't do not reply emails you don't
03:17 believe me you should just do a search
03:18 in your email for do not reply at and
03:21 see how many people don't want to talk
03:23 to their customers so let's start by
03:25 trying to figure out who I should talk
03:27 to it's always easy to interview people
03:30 that you know or people who are on your
03:32 network they're simply the most likely
03:33 to respond but perhaps can be less
03:35 honest in their answers to not offend
03:39 you don't be afraid of reaching out to
03:41 them and don't be afraid of rejection
03:43 the second group here is co-workers or
03:45 former co-workers they can be a really
03:47 great source they might know the topic
03:49 really well and if you're building
03:51 software for startups you can easily and
03:53 easily navigate yourself to someone
03:55 who's a potential User it's also very
03:57 common that the early users you speak to
03:59 are outside of your personal circuits or
04:01 your co-workers circles when asking YC
04:03 Founders how they found their early
04:05 users some of the most common answers
04:07 were LinkedIn red forms slack or Discord
04:10 communities and sometimes in-person
04:12 events so to illustrate this process
04:15 um a little better I'm going to start a
04:17 new mock startup this is not a real
04:19 company I'm just making it up but I'm
04:21 going to go through the process as I
04:22 would have if this was a real company
04:24 and first I want to understand if the
04:26 problem is real my hypothesis that I
04:28 have here or kind of like the idea space
04:30 that we sometimes call as NYC is that
04:33 companies want to reduce their carbon
04:35 emissions but for whatever reasons don't
04:36 do that so my plan is very simple I want
04:39 to first interview some potential
04:40 customers to learn about the problems
04:42 and motivations around carbon emissions
04:44 and three try to understand what an MVP
04:46 will look like MVP is a minimal viable
04:48 product to do this research I plan to
04:51 start by reaching out to Founders CEOs
04:53 CFOs at startups or bigger companies I'm
04:57 also going to be shocked to people on
04:58 LinkedIn that have titles like
05:01 carbon or climate or sustainability
05:04 basically Wars that gives me hint that
05:07 they might be relevant people to talk to
05:09 so what am I looking to learn well first
05:11 I'm trying to understand do the company
05:13 that they work for care about their
05:15 carbon emissions and two why do they
05:18 care and why do they not care if they
05:20 don't care and three who in that company
05:23 might be the one who cares the most
05:25 about this uh and why all right so so
05:27 here are two examples of Outreach
05:29 messages so the first one is a LinkedIn
05:31 message to a CFO who I used to work with
05:33 at Airbnb the message will go something
05:35 like this first I introduce myself I
05:37 mentioned something about our time at
05:38 Airbnb then I say I'm starting a new
05:40 project briefly describe it but not too
05:42 much in detail and I just asked for 20
05:44 minute phone call or video call that's
05:46 it and here's another example of a
05:48 message to someone I don't know it's
05:50 also on LinkedIn to the same group that
05:52 I mentioned earlier it's similar but
05:54 geared towards people I don't know so
05:56 the message will and the tone will be a
05:57 little bit different all right so once I
06:00 get on the phone or once again on the
06:02 video call let's talk about uh how to
06:05 actually interview potential customers
06:07 so first you want to make sure that
06:09 these interviews are over video calls
06:11 phone calls or in person you can learn a
06:13 lot more from a five minute video
06:15 interview than 500 or 5 000 sugar
06:18 responses
06:19 secondly you want to build some Report
06:21 with interviewee they're going to give
06:24 you answers to questions that maybe no
06:25 one have asked them before so you need
06:27 to make sure that they can trust you and
06:28 feel comfortable in this situation next
06:30 uh the most important thing here the
06:32 most most important thing here is to not
06:35 introduce your idea until maybe at the
06:37 end of the call or maybe not at all
06:39 doing this too early can bias them in
06:41 their answers your role here in this
06:44 interview is to listen not to talk
06:47 there are some great techniques to get
06:49 people to open up and talk about their
06:51 problems and one good one is to ask
06:53 follow-up questions that are open-ended
06:55 for example tell me about that if you're
06:58 not recording the interview you should
07:00 at least take notes as much as you can
07:03 you need to translate the recordings to
07:05 notes anyway so you might as well take
07:07 notes in the in the actual interview I'm
07:08 going to play an example video here on
07:11 uh how a potential customer for my
07:14 carbon emission company and how that
07:16 conversation will go like Hi how are you
07:18 um it's been a long time since we worked
07:20 together Airbnb
07:21 yeah great to see you
07:23 so I don't know if I told her this but
07:25 I'm thinking about starting another
07:26 company and I have a couple questions I
07:28 wanted to ask you about this era that
07:30 I'm interested in uh does your company
07:32 montevue Inc do you care about your
07:34 carbon emissions
07:36 we do actually care and we use a
07:38 consultant who creates these PDF reports
07:40 once a year
07:41 interesting uh what do you do with a
07:44 report
07:45 I mean right now I don't think we really
07:47 take much action on it unfortunately
07:50 well why not
07:51 it's just really hard to know what to do
07:53 with the information
07:55 um the information is really dense and
07:57 um uses words we don't really understand
08:00 tell me more about that
08:01 well we hired this consultant
08:04 um they asked us to fill out this word
08:06 doc with information but most of the
08:08 data that goes into it uh seems
08:10 inaccurate and I just feel like we need
08:13 another expert to figure out what to do
08:15 with the data
08:16 got it got it
08:18 um but why is it important for your
08:20 company at all to track carbon emission
08:22 well our
08:29 um and I think there's some regulations
08:30 coming up too
08:32 you notice here how I never introduced a
08:35 product or a solution I only ask
08:37 questions to learn about the problem
08:39 that Amy's having I don't even say what
08:41 I'm going to build
08:42 that's how you want to do it here are
08:44 some very specific questions that you
08:46 can use in your interviews one tell me
08:48 how you do X today by X of course I'm
08:51 referring to a specific problem a task
08:53 or goal that you are trying to solve two
08:55 what is the hardest thing you are doing
08:57 about x three why is it hard four how
09:01 often do you have to do x five why is it
09:04 important for your company to do X make
09:06 sure you dive deep into this question
09:07 like understanding that motivation is
09:09 going to be critical here and six what
09:11 do you do to solve this problem for
09:13 yourself today it's really important to
09:15 understand what they're doing
09:18 now if you can watch them do this that's
09:21 even better if they can screen share
09:23 during the video call or show you their
09:25 laptop as you're doing the interview and
09:28 show you how they actually get the PDF
09:30 report what they do with it
09:32 um how it looks like
09:33 that's the ideal State you want to
09:36 deeply understand their behavior not
09:37 just what they're saying but what
09:39 they're doing because changing the
09:41 behavior will be hard you can also ask
09:43 follow-up questions and please learn how
09:45 to make use of these ones it's very hard
09:47 to get someone to say everything in just
09:49 one answer to one question get
09:51 comfortable with asking questions like
09:53 what do you mean by that or can you tell
09:55 me more about that or why is that
09:58 important to you here are some questions
09:59 that you should not ask like there are
10:01 ways you can kind of derail this
10:04 interview will you use our product
10:06 they're probably going to say yes but
10:08 they're not going to mean anything to
10:09 you the answer
10:11 two which features would make product X
10:14 better
10:15 remember their job is to tell you about
10:17 the problem your job is to think about
10:19 what features to solve that problem so
10:22 asking them about your job isn't really
10:24 going to help you yes and no questions
10:27 it's not really that helpful for you you
10:29 want them to explain and give you as
10:32 many concrete and real examples as
10:35 possible so just simply asking yes no
10:36 questions is not ideal here and four
10:39 um how would would a better product X
10:42 look like to you how would a better
10:44 Carbon emission counting software look
10:46 like to you uh many of the people aren't
10:48 product developers they're not software
10:49 Engineers they're not designers they're
10:51 not product managers this is not their
10:53 job their job is to do a different job
10:55 and you are the one trying to understand
10:58 what is the challenging part about their
11:00 job
11:01 and finally
11:03 um asking two questions at the same time
11:05 that's going to confuse them it's going
11:06 to confuse your answers so this is when
11:09 you're eager to get um get answers it's
11:11 very easy to do this this thing but try
11:13 to avoid it but I want to highlight one
11:15 of the things I mentioned here there's
11:16 one big danger during interviews and
11:18 that is to focus on features instead of
11:19 problems
11:20 it's very natural for you to do that
11:22 your bra your brain is thinking through
11:24 all the ways you're going to solve this
11:26 problem
11:27 but really what you're here for during
11:28 this interview is to deeply understand
11:30 the problems not actually come up with
11:32 Solutions that's for for us for a
11:35 separate event so users generally have
11:38 good problems but also generally bad
11:40 Solutions so I'll give you two specific
11:42 examples so early days of Gmail users
11:45 were asking
11:46 um Paul Buchheit the founder of Gmail to
11:49 view both the Inbox and the actual email
11:52 that they were reading on the same
11:53 screen now why would they ask for that
11:55 well the reason was simply that Gmail is
11:57 too slow and people did not want to wait
11:59 to load each of the emails in the inbox
12:01 they want to see everything at the same
12:02 time similarly in early Airbnb a lot of
12:05 guests want to have the phone numbers of
12:07 the host so they can call them we're
12:09 like why do you need that it just turned
12:12 out that they didn't fully trust the
12:13 platform like we had not displayed
12:15 enough trust in Airbnb so they needed to
12:18 feel comfortable uh with getting to know
12:19 the host before it actually
12:21 um making the booking secondly
12:23 um users don't
12:25 have incentive to say no to really any
12:28 additional features if you ask for this
12:29 feature or this feature this feature
12:30 they'll probably say yes to everything
12:32 right you however are in charge of the
12:34 projectations and you have a lot more
12:36 incentives than they have to figure out
12:38 which are the most important of the
12:41 features so let's talk about next steps
12:43 so you've done your um five or ten of
12:45 these user interviews you've written
12:47 down your notes let's let's talk about
12:49 what you do then one you want to make
12:51 sure you have notes from all each of the
12:52 interviews you can use sticky notes or
12:54 some similar software to organize these
12:56 learnings you can bucket them in
12:58 different buckets that kind of describes
13:00 different problems and see which one is
13:01 actually the thing that matters the most
13:03 and then you want to write down your
13:04 conclusions what are you learning from
13:06 all of this you want to use all of this
13:08 information to create a hypothesis or
13:10 what the solution to this problem might
13:11 be don't over intellectualize this
13:13 process really you just want to start
13:14 assigning MVP as fast as you can but you
13:16 want to make sure you have accurate
13:17 information when you decide the MEP and
13:20 then you'll want to test it with the
13:21 same users it's also important to know
13:22 if the problem you're solving is
13:24 actually valuable what I mean by this is
13:26 that people will value the solution to
13:29 the problem enough to pay for it or the
13:31 problem itself solving the problem is
13:33 valuable so here's some good tips to
13:35 determine if it is one are people paying
13:37 money for other Solutions in the space
13:39 today are the people that make the PDF
13:41 reports for carbon emissions accounting
13:42 are they getting paid well if they are
13:44 getting paid that's a good sign that
13:45 this is probably a valuable thing do
13:47 people already have solutions to this
13:48 problem they are very happy with even
13:50 though it might appear Basics to you for
13:52 example like Excel or Google
13:54 spreadsheets is actually competitors to
13:56 many many many hundreds of startups
13:58 actually quite a formidable one to move
13:59 someone off Excel a Google spreadsheet
14:01 you need to make the experience of your
14:03 solution dramatically better they're not
14:05 going to pay you hundreds of dollars a
14:06 month for something they can almost as
14:08 easily do in Google spreadsheet or Excel
14:10 and three you want to evaluate how easy
14:12 it is to sell to this audience and this
14:14 is one thing that I like to do a lot
14:16 with the companies I work with for
14:17 example selling to Plumbers or
14:19 contractors is notoriously difficult
14:21 compared to selling to startups the
14:22 first two groups just don't change their
14:24 tools or the software very often versus
14:27 startups are open to trying new things
14:29 all the time so if you are solving a
14:32 problem is valuable you still have to
14:33 think about whether it's going to be
14:35 easy to sell this particular audience
14:36 once you have an MVP prototype you can
14:39 start by showing it to users even if
14:41 it's just a design prototype in Envision
14:43 or something like that you can just
14:44 click through we used to do this at MB
14:46 all the time we would go downstairs in
14:48 our in our office and we would find some
14:50 people that were waiting for an
14:52 interview or waiting for getting a
14:54 coffee or something and we just show
14:55 them the latest version of the app this
14:57 would be a Envision prototype on our
14:59 phone it would be Spike hand them our
15:01 phone and and show it to them this is
15:03 stuff that we haven't even built yet the
15:05 key thing but if you're doing this
15:06 tactic is to not tell them what to do do
15:11 not tell them what to do just watch them
15:12 play around with it you can tell them a
15:14 specific goal like try to make a book in
15:16 Airbnb or try to do a search with dates
15:19 and even b or something like that but do
15:21 not tell them exactly how to solve each
15:23 screen remember if this was a real
15:25 product you wouldn't you wouldn't be
15:26 standing there next to them to give them
15:27 advice another good trick when you're
15:29 doing this either when you're listening
15:31 or having them or recording them is to
15:34 have them speak their mind while they're
15:35 doing it have them describe exactly what
15:37 they're thinking on each of the steps
15:39 maybe the words that they're reading
15:41 maybe that will give you hints of what
15:42 words that they understand and which
15:44 words they don't understand or maybe
15:45 what specific screen means or what
15:47 purpose is and after you've done a bunch
15:49 of these interviews and then maybe MVP
15:52 protest sessions it's a great idea to
15:54 keep these interviewees involved
15:56 throughout the process
15:57 one common idea that a lot of wasting
15:59 companies are using is to create a slack
16:01 instance or a WhatsApp group for with
16:03 their customers now you want to make
16:05 them feel special right you want to make
16:07 them feel like they have exclusive
16:08 active access to the Future world
16:11 changing product so think about how
16:13 you're wording or or how you communicate
16:16 this WhatsApp group that you adding them
16:18 to it's a great idea to keep showing
16:19 them your product as they are
16:21 progressing like kind of give them this
16:23 exclusive feeling that I'm the first one
16:24 to see this product assets progressing
16:27 and also as you are
16:30 reacting to things that they're saying
16:32 you can ship a new a new screen with a
16:34 new version of the product
16:36 um and that builds trust with it with
16:38 them they now know that you react to
16:40 their feedback very fast finally some of
16:42 them will love connecting with other
16:44 people doing the same thing as they are
16:46 they don't get to do this very often and
16:49 um you are the one enabling that if you
16:51 do all these steps you will collect the
16:53 right information
16:54 um from your future users you will keep
16:56 them involved and transfer the
16:59 information uh and that allow you to
17:01 transfer the information into your first
17:02 prototype and MVP so to summarize here's
17:05 what I covered today why the best
17:07 Founders talk to their users throughout
17:08 the lifetime of the company
17:10 how to find your users and how to talk
17:12 to them
17:13 what questions to ask them and what not
17:15 to ask them and finally how to turn some
17:17 of those conclusions into an MVP thank
17:20 you everyone
17:21[Music]
🎥 Related Videos
a16z Podcast | Things Come Together -- Truths about Tech in Africa
a16z Podcast | The Infrastructure of Total Health
The Robot Lawyer Resistance with Joshua Browder of DoNotPay
a16z Podcast | Bots and Beyond
Design Sprints as a Tool for Organizational Change
a16z Podcast | Valuing Today's Fast-Growing Software Companies
🔥 Recently Summarized Examples