YT YouTube2Text

YouTube2Text → Transcripts → How To Talk To Users | Startup School

How To Talk To Users | Startup School — Transcript

by Y Combinator · 3,540 words · 510 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:01foreign
0:03[Music]
0:09my name is Gustav and I work here at y
0:12combinator as a group partner I've been
0:14here since 2017. before YC I was a
0:17practically to Airbnb and actually a YC
0:20founder back in 2007. today I'm going to
0:22talk about how to talk to your users and
0:25your future customers here's what I plan
0:27to cover today why the best Founders
0:29talk to the users throughout the
0:31lifetime of the company two how to find
0:33your users and how to talk to them three
0:35what questions to ask them and what not
0:37to ask them and four how to turn your
0:40conclusions into an MVP most people in
0:42the world have the idea on how new
0:44startups to form completely wrong they
0:46think ideas of new products is something
0:48the financials come up with on a lazy
0:49Sunday or a late night coding session
0:51you probably know it doesn't work this
0:53way this scene as you can see here here
0:55is a scene from The Social Network movie
0:57it's a great movie but they got a lot of
0:59things wrong on how Facebook actually
1:01got started this is Brian chesky Brian
1:03is the co-founder and CEO of Airbnb on
1:06the left side of the photo is Amal Amal
1:08was the very first guest on Airbnb they
1:10actually spent a weekend together in San
1:12Francisco when they'd be kind of watched
1:14you notice the difference between this
1:16photo and the previous photo well this
1:18photo actually is a real photo it
1:19includes a two-way conversation with a
1:21real customer great Founders talk to
1:23Future customers before they even have a
1:25product in fact the very best Founders
1:28in the world learn directly from their
1:30users throughout the life of their
1:32company but that's again if you have
1:34correctly identified who your users are
1:36you should learn from them throughout
1:38the lifetime of your company why you
1:40might ask
1:41users and customers will keep you honest
1:44they are the only stakeholders actually
1:46paying you anything if anyone will tell
1:49you the truth it will be them so my talk
1:51today is about how to do this well so
1:54let's talk about another m b story in
1:562010 um and I remember this uh myself
1:59Brian chesky the CEO of Airbnb he'll
2:02decide to launch an experiment he was
2:04going to give up his apartment and live
2:06in 50 different airbnbs over the next
2:08couple of months you might ask yourself
2:09why would anyone do that so much hassle
2:12most people read the articles about this
2:14experiment because he did announce it
2:15they missed a point
2:17um I personally remember this being kind
2:19of crazy by doing this Brian will
2:21literally get a chance to talk to every
2:22one of these 50 hosts each single day on
2:25this trip and one incredible source of
2:27honest feedback on your product without
2:28the Airbnb hosts MB would not exist
2:31exist today what Brian Joe and Nate did
2:33so early is deeply understand the
2:35motivations
2:36um and the feedback from their hosts
2:38still today Brian and Joe gets calls
2:41from hosts on their personal cell phone
2:43numbers I remember them telling me that
2:45that on their first day of adworks at
2:47Airbnb
2:48why do they do that well it's the number
2:50that they put on the website and share
2:52with all of their hosts early on they
2:54weren't hiding behind Anonymous email
2:55they wanted to hear from the customers
2:57at the time this seemed almost radical
3:00most Founders would spend their time in
3:02front of the computers trying to find
3:04scalable growth channels like Google ads
3:06unfortunately most Founders still don't
3:09build personal connections with their
3:10users today behind behind info at emails
3:13or don't do not reply emails you don't
3:17believe me you should just do a search
3:18in your email for do not reply at and
3:21see how many people don't want to talk
3:23to their customers so let's start by
3:25trying to figure out who I should talk
3:27to it's always easy to interview people
3:30that you know or people who are on your
3:32network they're simply the most likely
3:33to respond but perhaps can be less
3:35honest in their answers to not offend
3:39you don't be afraid of reaching out to
3:41them and don't be afraid of rejection
3:43the second group here is co-workers or
3:45former co-workers they can be a really
3:47great source they might know the topic
3:49really well and if you're building
3:51software for startups you can easily and
3:53easily navigate yourself to someone
3:55who's a potential User it's also very
3:57common that the early users you speak to
3:59are outside of your personal circuits or
4:01your co-workers circles when asking YC
4:03Founders how they found their early
4:05users some of the most common answers
4:07were LinkedIn red forms slack or Discord
4:10communities and sometimes in-person
4:12events so to illustrate this process
4:15um a little better I'm going to start a
4:17new mock startup this is not a real
4:19company I'm just making it up but I'm
4:21going to go through the process as I
4:22would have if this was a real company
4:24and first I want to understand if the
4:26problem is real my hypothesis that I
4:28have here or kind of like the idea space
4:30that we sometimes call as NYC is that
4:33companies want to reduce their carbon
4:35emissions but for whatever reasons don't
4:36do that so my plan is very simple I want
4:39to first interview some potential
4:40customers to learn about the problems
4:42and motivations around carbon emissions
4:44and three try to understand what an MVP
4:46will look like MVP is a minimal viable
4:48product to do this research I plan to
4:51start by reaching out to Founders CEOs
4:53CFOs at startups or bigger companies I'm
4:57also going to be shocked to people on
4:58LinkedIn that have titles like
5:01carbon or climate or sustainability
5:04basically Wars that gives me hint that
5:07they might be relevant people to talk to
5:09so what am I looking to learn well first
5:11I'm trying to understand do the company
5:13that they work for care about their
5:15carbon emissions and two why do they
5:18care and why do they not care if they
5:20don't care and three who in that company
5:23might be the one who cares the most
5:25about this uh and why all right so so
5:27here are two examples of Outreach
5:29messages so the first one is a LinkedIn
5:31message to a CFO who I used to work with
5:33at Airbnb the message will go something
5:35like this first I introduce myself I
5:37mentioned something about our time at
5:38Airbnb then I say I'm starting a new
5:40project briefly describe it but not too
5:42much in detail and I just asked for 20
5:44minute phone call or video call that's
5:46it and here's another example of a
5:48message to someone I don't know it's
5:50also on LinkedIn to the same group that
5:52I mentioned earlier it's similar but
5:54geared towards people I don't know so
5:56the message will and the tone will be a
5:57little bit different all right so once I
6:00get on the phone or once again on the
6:02video call let's talk about uh how to
6:05actually interview potential customers
6:07so first you want to make sure that
6:09these interviews are over video calls
6:11phone calls or in person you can learn a
6:13lot more from a five minute video
6:15interview than 500 or 5 000 sugar
6:18responses
6:19secondly you want to build some Report
6:21with interviewee they're going to give
6:24you answers to questions that maybe no
6:25one have asked them before so you need
6:27to make sure that they can trust you and
6:28feel comfortable in this situation next
6:30uh the most important thing here the
6:32most most important thing here is to not
6:35introduce your idea until maybe at the
6:37end of the call or maybe not at all
6:39doing this too early can bias them in
6:41their answers your role here in this
6:44interview is to listen not to talk
6:47there are some great techniques to get
6:49people to open up and talk about their
6:51problems and one good one is to ask
6:53follow-up questions that are open-ended
6:55for example tell me about that if you're
6:58not recording the interview you should
7:00at least take notes as much as you can
7:03you need to translate the recordings to
7:05notes anyway so you might as well take
7:07notes in the in the actual interview I'm
7:08going to play an example video here on
7:11uh how a potential customer for my
7:14carbon emission company and how that
7:16conversation will go like Hi how are you
7:18um it's been a long time since we worked
7:20together Airbnb
7:21yeah great to see you
7:23so I don't know if I told her this but
7:25I'm thinking about starting another
7:26company and I have a couple questions I
7:28wanted to ask you about this era that
7:30I'm interested in uh does your company
7:32montevue Inc do you care about your
7:34carbon emissions
7:36we do actually care and we use a
7:38consultant who creates these PDF reports
7:40once a year
7:41interesting uh what do you do with a
7:44report
7:45I mean right now I don't think we really
7:47take much action on it unfortunately
7:50well why not
7:51it's just really hard to know what to do
7:53with the information
7:55um the information is really dense and
7:57um uses words we don't really understand
8:00tell me more about that
8:01well we hired this consultant
8:04um they asked us to fill out this word
8:06doc with information but most of the
8:08data that goes into it uh seems
8:10inaccurate and I just feel like we need
8:13another expert to figure out what to do
8:15with the data
8:16got it got it
8:18um but why is it important for your
8:20company at all to track carbon emission
8:22well our
8:29um and I think there's some regulations
8:30coming up too
8:32you notice here how I never introduced a
8:35product or a solution I only ask
8:37questions to learn about the problem
8:39that Amy's having I don't even say what
8:41I'm going to build
8:42that's how you want to do it here are
8:44some very specific questions that you
8:46can use in your interviews one tell me
8:48how you do X today by X of course I'm
8:51referring to a specific problem a task
8:53or goal that you are trying to solve two
8:55what is the hardest thing you are doing
8:57about x three why is it hard four how
9:01often do you have to do x five why is it
9:04important for your company to do X make
9:06sure you dive deep into this question
9:07like understanding that motivation is
9:09going to be critical here and six what
9:11do you do to solve this problem for
9:13yourself today it's really important to
9:15understand what they're doing
9:18now if you can watch them do this that's
9:21even better if they can screen share
9:23during the video call or show you their
9:25laptop as you're doing the interview and
9:28show you how they actually get the PDF
9:30report what they do with it
9:32um how it looks like
9:33that's the ideal State you want to
9:36deeply understand their behavior not
9:37just what they're saying but what
9:39they're doing because changing the
9:41behavior will be hard you can also ask
9:43follow-up questions and please learn how
9:45to make use of these ones it's very hard
9:47to get someone to say everything in just
9:49one answer to one question get
9:51comfortable with asking questions like
9:53what do you mean by that or can you tell
9:55me more about that or why is that
9:58important to you here are some questions
9:59that you should not ask like there are
10:01ways you can kind of derail this
10:04interview will you use our product
10:06they're probably going to say yes but
10:08they're not going to mean anything to
10:09you the answer
10:11two which features would make product X
10:14better
10:15remember their job is to tell you about
10:17the problem your job is to think about
10:19what features to solve that problem so
10:22asking them about your job isn't really
10:24going to help you yes and no questions
10:27it's not really that helpful for you you
10:29want them to explain and give you as
10:32many concrete and real examples as
10:35possible so just simply asking yes no
10:36questions is not ideal here and four
10:39um how would would a better product X
10:42look like to you how would a better
10:44Carbon emission counting software look
10:46like to you uh many of the people aren't
10:48product developers they're not software
10:49Engineers they're not designers they're
10:51not product managers this is not their
10:53job their job is to do a different job
10:55and you are the one trying to understand
10:58what is the challenging part about their
11:00job
11:01and finally
11:03um asking two questions at the same time
11:05that's going to confuse them it's going
11:06to confuse your answers so this is when
11:09you're eager to get um get answers it's
11:11very easy to do this this thing but try
11:13to avoid it but I want to highlight one
11:15of the things I mentioned here there's
11:16one big danger during interviews and
11:18that is to focus on features instead of
11:19problems
11:20it's very natural for you to do that
11:22your bra your brain is thinking through
11:24all the ways you're going to solve this
11:26problem
11:27but really what you're here for during
11:28this interview is to deeply understand
11:30the problems not actually come up with
11:32Solutions that's for for us for a
11:35separate event so users generally have
11:38good problems but also generally bad
11:40Solutions so I'll give you two specific
11:42examples so early days of Gmail users
11:45were asking
11:46um Paul Buchheit the founder of Gmail to
11:49view both the Inbox and the actual email
11:52that they were reading on the same
11:53screen now why would they ask for that
11:55well the reason was simply that Gmail is
11:57too slow and people did not want to wait
11:59to load each of the emails in the inbox
12:01they want to see everything at the same
12:02time similarly in early Airbnb a lot of
12:05guests want to have the phone numbers of
12:07the host so they can call them we're
12:09like why do you need that it just turned
12:12out that they didn't fully trust the
12:13platform like we had not displayed
12:15enough trust in Airbnb so they needed to
12:18feel comfortable uh with getting to know
12:19the host before it actually
12:21um making the booking secondly
12:23um users don't
12:25have incentive to say no to really any
12:28additional features if you ask for this
12:29feature or this feature this feature
12:30they'll probably say yes to everything
12:32right you however are in charge of the
12:34projectations and you have a lot more
12:36incentives than they have to figure out
12:38which are the most important of the
12:41features so let's talk about next steps
12:43so you've done your um five or ten of
12:45these user interviews you've written
12:47down your notes let's let's talk about
12:49what you do then one you want to make
12:51sure you have notes from all each of the
12:52interviews you can use sticky notes or
12:54some similar software to organize these
12:56learnings you can bucket them in
12:58different buckets that kind of describes
13:00different problems and see which one is
13:01actually the thing that matters the most
13:03and then you want to write down your
13:04conclusions what are you learning from
13:06all of this you want to use all of this
13:08information to create a hypothesis or
13:10what the solution to this problem might
13:11be don't over intellectualize this
13:13process really you just want to start
13:14assigning MVP as fast as you can but you
13:16want to make sure you have accurate
13:17information when you decide the MEP and
13:20then you'll want to test it with the
13:21same users it's also important to know
13:22if the problem you're solving is
13:24actually valuable what I mean by this is
13:26that people will value the solution to
13:29the problem enough to pay for it or the
13:31problem itself solving the problem is
13:33valuable so here's some good tips to
13:35determine if it is one are people paying
13:37money for other Solutions in the space
13:39today are the people that make the PDF
13:41reports for carbon emissions accounting
13:42are they getting paid well if they are
13:44getting paid that's a good sign that
13:45this is probably a valuable thing do
13:47people already have solutions to this
13:48problem they are very happy with even
13:50though it might appear Basics to you for
13:52example like Excel or Google
13:54spreadsheets is actually competitors to
13:56many many many hundreds of startups
13:58actually quite a formidable one to move
13:59someone off Excel a Google spreadsheet
14:01you need to make the experience of your
14:03solution dramatically better they're not
14:05going to pay you hundreds of dollars a
14:06month for something they can almost as
14:08easily do in Google spreadsheet or Excel
14:10and three you want to evaluate how easy
14:12it is to sell to this audience and this
14:14is one thing that I like to do a lot
14:16with the companies I work with for
14:17example selling to Plumbers or
14:19contractors is notoriously difficult
14:21compared to selling to startups the
14:22first two groups just don't change their
14:24tools or the software very often versus
14:27startups are open to trying new things
14:29all the time so if you are solving a
14:32problem is valuable you still have to
14:33think about whether it's going to be
14:35easy to sell this particular audience
14:36once you have an MVP prototype you can
14:39start by showing it to users even if
14:41it's just a design prototype in Envision
14:43or something like that you can just
14:44click through we used to do this at MB
14:46all the time we would go downstairs in
14:48our in our office and we would find some
14:50people that were waiting for an
14:52interview or waiting for getting a
14:54coffee or something and we just show
14:55them the latest version of the app this
14:57would be a Envision prototype on our
14:59phone it would be Spike hand them our
15:01phone and and show it to them this is
15:03stuff that we haven't even built yet the
15:05key thing but if you're doing this
15:06tactic is to not tell them what to do do
15:11not tell them what to do just watch them
15:12play around with it you can tell them a
15:14specific goal like try to make a book in
15:16Airbnb or try to do a search with dates
15:19and even b or something like that but do
15:21not tell them exactly how to solve each
15:23screen remember if this was a real
15:25product you wouldn't you wouldn't be
15:26standing there next to them to give them
15:27advice another good trick when you're
15:29doing this either when you're listening
15:31or having them or recording them is to
15:34have them speak their mind while they're
15:35doing it have them describe exactly what
15:37they're thinking on each of the steps
15:39maybe the words that they're reading
15:41maybe that will give you hints of what
15:42words that they understand and which
15:44words they don't understand or maybe
15:45what specific screen means or what
15:47purpose is and after you've done a bunch
15:49of these interviews and then maybe MVP
15:52protest sessions it's a great idea to
15:54keep these interviewees involved
15:56throughout the process
15:57one common idea that a lot of wasting
15:59companies are using is to create a slack
16:01instance or a WhatsApp group for with
16:03their customers now you want to make
16:05them feel special right you want to make
16:07them feel like they have exclusive
16:08active access to the Future world
16:11changing product so think about how
16:13you're wording or or how you communicate
16:16this WhatsApp group that you adding them
16:18to it's a great idea to keep showing
16:19them your product as they are
16:21progressing like kind of give them this
16:23exclusive feeling that I'm the first one
16:24to see this product assets progressing
16:27and also as you are
16:30reacting to things that they're saying
16:32you can ship a new a new screen with a
16:34new version of the product
16:36um and that builds trust with it with
16:38them they now know that you react to
16:40their feedback very fast finally some of
16:42them will love connecting with other
16:44people doing the same thing as they are
16:46they don't get to do this very often and
16:49um you are the one enabling that if you
16:51do all these steps you will collect the
16:53right information
16:54um from your future users you will keep
16:56them involved and transfer the
16:59information uh and that allow you to
17:01transfer the information into your first
17:02prototype and MVP so to summarize here's
17:05what I covered today why the best
17:07Founders talk to their users throughout
17:08the lifetime of the company
17:10how to find your users and how to talk
17:12to them
17:13what questions to ask them and what not
17:15to ask them and finally how to turn some
17:17of those conclusions into an MVP thank
17:20you everyone
17:21[Music]

About this transcript

This page contains the full transcript of How To Talk To Users | Startup School by Y Combinator, generated from the public captions YouTube serves with the video. The transcript has 3,540 words across 510 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.