YT YouTube2Text

YouTube2Text → Transcripts → How To Start A Dev Tools Company | Startup School

How To Start A Dev Tools Company | Startup School — Transcript

by Y Combinator · 5,473 words · 773 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:01[Music]
0:09hello my name is Nicola D and I'm a
0:12group partner here at YC before that I
0:15was the co-founder and CEO of algolia
0:18algolia is a search API used by millions
0:21of developers to build a great search
0:23experience in their apps and websites
0:26and it's relevant today because today
0:28we're going to speak about death tools
0:30and how to start a def tools company so
0:34here's what we're going to cover first
0:36we'll discuss about the funding team and
0:38how to find your idea then how to start
0:42the company from the prototype to the
0:44MVP and finally we'll dive into the go
0:48to market including sales and marketing
0:51for def tools but first what is the def
0:54tool so a def tool is software that is
0:56used by developers to help them build
0:59product
1:00and that includes a lot of aspects from
1:03coding testing debugging documenting
1:07deploying running it's all of the gamut
1:09of aspects of that job and it's a broad
1:12category then you can think about idees
1:15like vscode API like strip twio or
1:19algolia my company library and
1:22Frameworks like react not GS or
1:24longchain or even infrastructure or
1:27cloud services like AWS or vers so and
1:30many other tools Docker terraform data
1:33dog GitHub you get it it's very diverse
1:37and at YC we've supported many of these
1:40Dev tools hundreds of them and we
1:42already have two public companies gitlab
1:45the open source GitHub and peder Duty
1:48that started as a simple alerting system
1:51and now is used by half the 14 500
1:54companies and many others you probably
1:56know about strip Docker Hoku super
2:00segment poog Apollo airb and also alonia
2:04of course and we've learned a lot from
2:07them so let's dive in and start at the
2:10founding stage well first you need a
2:13founding team and the key aspect of
2:16death tool is that you are going to
2:18build a technical product for developers
2:21obviously you need to be a developers
2:23yourself most of our co-funding team for
2:26Dev tools are all developer teams
2:30and the great thing about death tool is
2:32that developers are actually using de
2:35tools every day so when you are building
2:37a de tool you are basically helping
2:39yourself in your action of building a
2:42product so let's take a step back and
2:44chat quickly about what are good and bad
2:47death tool ideas and as a profess let me
2:50start by telling you that with the
2:51Advent of llms and a in general it's
2:54actually very difficult to know for sure
2:56today what's a good or bad idea the bad
2:59ideas the ideas of yesterday could
3:01actually be possible to make successful
3:03now thanks to llms some ideas are
3:07actually more difficult than others
3:08developers every day are going to work
3:10on things they are actually not going to
3:12like like documentation QA testing and
3:15these become obvious ideas for them to
3:18fix for others the problem here is that
3:21they are not the only one to think about
3:23it and that creates a lot of noise and
3:25that's why there are so many of these on
3:27the market do you really want to build
3:29the next yet the next QA tool for other
3:32developers H I don't know another way to
3:35look at it is the difference between
3:37build time and runtime ideas so Doug QA
3:40testing all of these ideas are build
3:43time ideas things tools you're going to
3:46use when you building the software it
3:49happens that these ideas are mostly nice
3:52to have you could build your product
3:54with them what's more interesting are
3:57runtime ideas why because at run time it
4:01becomes really critical it's must have
4:04products if you bet on a on API for
4:07example you cannot run your own product
4:10if it's down that's why it's more
4:12critical and it's way better idea for f
4:15tools the other advantage of these
4:17runtime ideas is usage because the more
4:22your customers are going to grow usually
4:25the more they're going to use your
4:27product so your incentives are align
4:30think about stripe for example if their
4:32customers sell more product they make
4:34more money everyone is happy much better
4:37idea another quick category uh that I
4:40can touch upon is libraries and
4:42Frameworks these can be greater is
4:44usually open source but they can also be
4:47very challenging to monetize think about
4:50pandas for example very difficult to
4:52make money out of that the one way you
4:55can monetize these ideas is by offering
4:57a hosting service think about next GS
5:00and Verso for example okay last point I
5:04wanted to touch on this idea which is
5:07the llm and AI Trend actually with this
5:10change in the world there are many new
5:12death tools being created and it's
5:14awesome it's so exciting to work on the
5:16de tool that's going to help these
5:17companies building nlms but it's also a
5:20time where the market is very nent so a
5:24obuse great idea is llm observability
5:28well uh it's obuse right anyone building
5:31with LM they want to know how they are
5:33doing they want to be able to evaluate
5:35them guess what many other companies
5:37think the same thing and that's why you
5:39have dozens if not hundreds of such
5:41companies out there it doesn't mean that
5:44it's a bad idea but I think that if you
5:46are starting in that space you need to
5:48have a clear idea of how you are going
5:51to differentiate yourself or are you
5:53going to win against all of these
5:54competitors okay which leads me to a few
5:58mistakes people often do the first one
6:00is waiting for the perfect idea you are
6:04going to wait for a very long time let
6:06me go back to this llm observability
6:08ideas I would actually encourage you if
6:10that's the idea you have just start
6:13that's okay even if you don't
6:14differentiate yourself right away that's
6:16okay you learn as you go and eventually
6:19you'll be able to pull it off or you'll
6:21change your idea which leads me to my
6:23second mistake sticking with the wrong
6:26idea for too long you may not realize
6:29but
6:3050% of the companies in YC eventually
6:34pivot from their first idea it's huge
6:37and it's also true for death tools
6:39finally last mistake thinking you need a
6:42business founder in the team I know that
6:44what I start when I created agolia we
6:46need up creating agolia with two Tech
6:48coers but it crossed our mind hey we
6:50don't know how to sell we need to hire
6:52or to find some other co-founder who
6:54knows about sell no it's actually ways
6:58understand you think to s to learn how
7:00to sell your product actually we check
7:02the numbers
7:0474% of YC def tool companies had only
7:08Tech co-founders and that's compared to
7:1145% for all of the other companies okay
7:15so now that you have a team and an idea
7:17where do you start well there is only
7:20two things you have to do build and talk
7:23with users there is no specific order
7:26you can actually start by talking with
7:27users because they're going to help you
7:29figure out out if your initial IDE is
7:30good or you can start by building a
7:32prototype so you can get easier feedback
7:35from users let's speak about that
7:37prototype the most important thing here
7:40is not to over engineer it you should do
7:44that quick and dirty and I know that it
7:47can be very difficult for some
7:49developers out there especially for
7:50experienced Engineers who Pride
7:53themselves on robustness and scalability
7:56of their code but the truth is that the
7:58most important thing at that stage is to
8:01iterate as fast as you can all of th I
8:05would assume that you'll throw away 90%
8:09of all the code you write so your goal
8:12at that stage is just to identify the
8:1410% of what you building that's actually
8:17valuable as fast as possible and then
8:20later on you can refactor this 10% only
8:23and avoid wasting a lot of time as
8:25youate of course you'll want to speak
8:27with users and get feedback it's
8:29completely okay to show them a prototype
8:32please don't wait to have a perfect
8:33product just show them a prototype and
8:36start Gathering feedback as early as you
8:38can and that will eventually lead you to
8:42an MVP a minimal viable product and this
8:45word the V is important valuable you
8:48need to provide value to your customers
8:51it's much better to be 10x better on a
8:54very tiny thing for someone who cares
8:58about it well working for a niche is
9:00completely okay once you have real
9:02customers who love your product it will
9:04be very easy to expand from there let me
9:06give you the example of algolia well
9:08today algolia is a very broad large like
9:12full featured search engine right but
9:15when we started to be honest it was just
9:17a glorified autocomplete very minimal
9:21but that product was so much better than
9:23whatever else was available at the time
9:26that some people cared enough about it
9:29and when when we started demonstrating
9:31it we didn't have much to show actually
9:34I still remember our first customer that
9:36first meeting when I did demo using a
9:39command line to index uh content uh we
9:42didn't have any API client yet we didn't
9:44have any admin UI yet just a command
9:46line and then a very simple web page to
9:49show the search that was enough to close
9:51a $2,000 a month contract okay you have
9:55built something you can show users so
9:57the next step is to talk with users
9:59I know that many developers many techers
10:02actually introvert and they don't like
10:04speaking with users or with customers
10:07like they don't like sales well you may
10:10not realize but as you are building a
10:12death tool you have a huge Advantage you
10:15know your audience you are your audience
10:18you are a developer yourself you speak
10:19the language of your customers you are
10:21uniquely qualified to understand them so
10:25you'll see it actually will be way
10:27easier than you think please don't wait
10:29wait for the product to be ready you
10:31want to learn ASAP right away if you are
10:34solving a real problem then people will
10:36be ready to pay for but how do you find
10:39these users two things outre and
10:42launches let's speak about Outreach
10:44quickly death tool co-founders often
10:47plan to have a bottom up adoption model
10:50where individual developers are going to
10:52come inbound try the product and convert
10:54or maybe becom their Champions U at
10:57bigger companies well I have some bad
10:59news for you nobody knows you yet so to
11:04get there first you need to get people
11:07to know about you and there is no other
11:09way to do that than outreaching and
11:10finding them yourself but you need to be
11:13smart about it start with your network
11:15your previous colleagues or classmates
11:18and then you expand from there um
11:20friends of friends you can leverage
11:22LinkedIn you can find like specific HS
11:24over there but please please personalize
11:28your messages
11:29like don't copy what you can see out
11:32there sent by marketers actually your
11:35developers I'm sure you hate these
11:36messages so don't do the same just ask
11:39yourself would you be excited to open
11:42your message then ask your co-founder
11:45ask other developer friend would they be
11:47excited to open your message and then
11:49read your email and then possibly act on
11:50it well you simply iterate until they do
11:54now let's speak about launching you can
11:57launch many times and that's what you
11:59should do here to get more
12:02attention the best place to launch for
12:06death tools I would AR you is Hacker
12:08News Hacker News is a community of
12:11intellectually curious people many of
12:14them are developers and there is that
12:17section in hacka news called show Eden
12:20that's the perfect place to launch and
12:23gather feedback on what you are building
12:25again like when you we writing like kind
12:28of this Outreach messages please don't
12:30do marketing don't try to sell your
12:33product just explain plainly and simply
12:36what's new and interesting in what you
12:39are building that will be very engaging
12:41for the community and they'll share
12:43comments to you engage with these
12:45comments don't let them unanswered try
12:48to engage with people asking questions
12:50some of them are going to be haters and
12:52be rude well you should engage with them
12:54too but your goal is not to convince
12:56them it's just to convince any other
12:59readers who are going to read how you
13:00reply to these haers just be classy here
13:03one of the great example here is segment
13:06um they simply did an experiment by
13:09launching their next idea on Hacker News
13:11and it blew up they got hundreds of
13:14votes and so many comments that they
13:17knew they were onto something and that
13:19became their idea another example is AMA
13:22olama helps companies run llms on their
13:26own machines well it started as a
13:28comment on a comment on another post on
13:31Hacker News and just from that they got
13:33a few people interested and and then
13:37shortly after they launched their
13:38product it complete took off on Hacker
13:41News they got like again hundreds of H
13:43votes and they knew they were on do
13:44something and they didn't stop there
13:47they launched again and again every few
13:49months each time they had like new
13:51things to share and that continued to
13:53fuel their growth uh creating a lot of
13:57excitement in the community Hacker News
13:59best place to launch the last thing I
14:01wanted to touch upon here in that
14:03section is you should do things that
14:06don't scale so again you're here to
14:09learn as fast as possible from your
14:11users what they like or not about your
14:14product a great example of a team that
14:17did things that don't scale is stripe
14:19early on the stripe team were going to
14:22their customers small startups and they
14:25help them Implement their own product
14:27sitting side by side front of the
14:29computer coding with them that was not
14:32only a great way to make sure that users
14:35were going to be satisfied we going to
14:37be actually excited about what they were
14:39building but it was also a great way to
14:41gather invaluable feedback very early
14:45from their users so it's completely okay
14:48to do things that don't scale if they
14:50lead to faster learnings okay and here
14:53are a few mistakes to avoid so remember
14:56the goal is to go fast and trade fast so
14:59please don't choose a text tag because
15:01it's cool because you want to learn it
15:03no you should took it because of your
15:05expertise with it so that you can
15:07iterate as fast as possible other
15:09mistakes we we touch not talking with
15:13users please get that feedback as fast
15:16as you can which also means avoiding
15:19overbuilding before getting that
15:21feedback maybe a small thing here be
15:24careful about that feedback be careful
15:26about not misunderstanding some of the
15:28developer feedback back for example they
15:31may share they may tell you how you
15:32should beat your product you don't care
15:35you just want to know if they would use
15:37your product or not and why or they
15:39could tell you hey you know what I could
15:41build this in a week I don't need it
15:43well guess what you just build it
15:45yourself in a week you you should not
15:47mind about that just move on and
15:49continue improving the product last
15:51mistake some teams do is hiring too
15:54early very early on you don't even know
15:58if your product uh is providing value
16:00please don't hire anyone until you have
16:03convinced yourself you're working on the
16:05right idea let's now talk about go to
16:08market and the first question you should
16:11ask yourself about go to market is your
16:13business model what explicitly should
16:16you be open source open source has
16:18become one of the main go to market for
16:21death tools and it definitely works as a
16:23business model as many companies have
16:25proven like data breakes elastic ashic
16:27cob gitlab all aome great companies I
16:30would argue that you must be open source
16:33if you are working on things like a
16:34library or framework that's what today
16:37expecting for the market developers
16:39wouldn't Implement on top of a new
16:41framework that is not open source you
16:43should also be open source if you are
16:45dealing with data especially sensitive
16:47data a new database would need to be
16:50open source or an API connecting with
16:53sensitive data in a CRM or maybe an EHR
16:56needs to be open source you could be
16:59open source but probably don't have to
17:01be open source if you are working on
17:03other things like an API or maybe an
17:05application like a butt cracker maybe in
17:07that case it's not as important but
17:10still could be your differentiation open
17:12source has quite a lot of benefits your
17:15customers are developers well as you
17:18know already developers prefer working
17:20with open source tools it's also going
17:22to help you create more awareness
17:24through the community it can be also
17:26your differentiation like you could be
17:28the open source version of X think about
17:32post and amplitude or super base and
17:34Firebase or airb and five TR great
17:37approach speaking of airb you could also
17:40receive contribution from the community
17:43in the case of airb they got a lot of
17:44connectors contributed from the
17:46community but I would be careful on that
17:48one most cases you shouldn't count on
17:51them because the quality of the
17:52contribution may not be up to your
17:54standard and then you have to deal with
17:57the contributors and that could be more
17:59actually more difficult than you think
18:02but in most cases I would say that you
18:03should not count on them good
18:06contributions are actually pretty rare
18:08and last benefit but not the least is
18:10the trust and it's actually very
18:13important for large Enterprises in the
18:16Enterprise real take the example of met
18:19Plum for example who is building a open
18:21source EHR the fact that they are open
18:24source I would argue probably help them
18:26shorten their sales cycle with
18:28Enterprises by a year or more there is
18:32one thing you need to think about
18:34however when you are creating an open
18:35source project maybe not at the start
18:38but pretty soon it's
18:40monetization if you cannot monetize your
18:43product you don't have a company at some
18:46point even if it's later in the life of
18:47the company you need to earn money there
18:50are many variants out there but it's
18:53good to look at the common approaches
18:55some of the common ones are hosting
18:57having a cloud offer so your customers
18:59may use your open source free
19:02self-hosted but they could also pay for
19:05you to host it so that they don't have
19:07to maintain it sometimes Cloud version
19:10also have like some additional features
19:12like Team Management another one is
19:14being open core in that case you usually
19:16have a a special offer typically an
19:18Enterprise offer that have advanced
19:21features that are not in the open source
19:23version or maybe they're on a def
19:25license common features are typically
19:27the ones that are going to to be
19:29necessary for Enterprises like SSO o
19:32lugs Disaster Recovery slas these kind
19:35of features and finally some open source
19:38companies charge for support and
19:41services but I would actually discourage
19:43you to follow that monetization pass
19:46because your incentive becomes to build
19:48the most complex product possible so you
19:50can charge for support and services
19:53which is a bad incentive and if your
19:55product doesn't need too much support
19:56then people are going to turn and not
19:57renew their support contract and what if
19:59you are not open source well for these
20:02Dev tools they are usually following one
20:04of two approaches either they usage
20:06based like apis aloia too stripe and so
20:11they charge based on the usage their
20:14customers do uh possibly they have a
20:16volume discount or Enterprise specific
20:19options that are on the side and the
20:21second approach is having some specific
20:24plans typically like a good better best
20:27approach good is going to be the self
20:29serve option um that's going to solve
20:32the problem of Developers for a very low
20:34price for the better options usually are
20:36going to be interacting with engineering
20:38managers who are going to care about
20:40collaboration that's going to enable
20:42teams still can be self- served and the
20:45best option is higher up you are talking
20:48to the CTO it's all about security audit
20:50logs Disaster Recovery SLA and in that
20:54case it's usually a sales Le approach
20:57let's speak about sales so as a quick
21:00reminder from what I was saying earlier
21:02don't fear Outreach even if most death
21:05tools are going to end up being bottom
21:07up and rely on inbound really heavily no
21:11one knows you that first and so you need
21:13to start with Outreach that period of
21:16founder Outreach is how you start as you
21:19get more well known that's when you'll
21:21get inbound usually you are going to
21:24have at that point more Self Serve leads
21:27and then eventually you end up hiring a
21:29sales team to sell to bigger icons
21:32Enterprise so should you hire a sales
21:34team well I would argue that you should
21:37wait as long as you can start by selling
21:41yourself don't delegate away the truth
21:43is that you are the only person in the
21:46company today capable of selling your
21:48own product if you as the founder cannot
21:51sell your product nobody else can as a
21:55rule of Thum I would wait to get to
21:58about 1 million AR before hiring your
22:01first salesperson and even then try to
22:05hire people who are technical or at
22:07least understand developers because
22:10developers don't hate speaking with
22:12salese actually remember our first sales
22:14peoples at aolia they decided to change
22:17the title on LinkedIn and instead of
22:19like being sales I can't EXA they called
22:22them themselves product specialist and
22:25the nice thing here is that they
22:27actually had to own that stages they had
22:29to understand the product intely so it
22:31was kind of like a great incentive for
22:33them to be better at product and a great
22:36way for them to open doors with
22:38developers who don't want to speak with
22:40sales another example post hog did you
22:43know that team the sto is also their
22:46sales leader not a business leader their
22:48sales leader is also the CTO is see
22:52sales as a nuring problem which is an
22:54awesome approach for a death tool if you
22:56want to learn more about sales you
22:57should also watched pit talk about
23:00Enterprise sales another startup school
23:02video so is it different to sell a de
23:05tool from selling any other product yes
23:08it is you should have build a sales deck
23:11developers like don't want to waste
23:13their time reading through a sales deck
23:16you should show not tell just two
23:20demonstrations I think that's the best
23:22uh sales approach you can do for de to
23:26at aloia we didn't have a sales deck
23:28before we got to 10 million AR so even
23:32our sales like proper account exacts we
23:35are just showing
23:36demonstrations uh when they were selling
23:38and that was working so much better for
23:40everyone you should lean completely in
23:43the tech aspect of your product for
23:45example at Alia we sell a lot to
23:47e-commerce players and what we realized
23:49is that when the buyer is not Technical
23:53and want a turnkey solution we always
23:57not always but most of the the time we
23:58actually lose the deal but when the
24:01buyer is technical or if tech has a lot
24:04of influence in the decision we win way
24:07more often so we realized by leaning
24:10into that aspect of our product we are
24:12an API we are not like a turnkey
24:15solution well we ended up like finding
24:18our ICP and having better success
24:20selling okay and the last thing that is
24:22also true in with inev tool is that most
24:24of them are bottom up what that means is
24:26that a lot of people are going to s
24:28serve even within
24:31Enterprises so instead of going top down
24:35just pitching your product directly
24:37sales should also check if some of the
24:41Enterprise employees are already using
24:43the product and they should then focus
24:45on helping management understand what
24:47the product is already doing for them
24:50and how much more it could help them if
24:53they were to pay for more like new seats
24:56or features or additional usage okay so
24:59the last uh topic I wanted to touch in
25:02this video is developer marketing your
25:06goal is to create a lot of awareness for
25:08your product because you want inbound
25:11the first thing you should do is find
25:13your community so it could be Hacker
25:15News or some subreddit or some Discord
25:18out there that is related to what you're
25:21working on but be careful your goal is
25:23not to sell your product just yet here
25:26just be helpful just establish yourself
25:30as the expert and of course people are
25:32going to start knowing you and trust
25:34your brand your company that will help
25:37you down the line uh people will think
25:39about you when they have that problem
25:41launch of that's very important we
25:44discussed about uh the example of AMA
25:46before uh but that should also be done
25:49for the life of your company some
25:52companies even do launch weeks and have
25:54been very successful for that it works
25:56great for open source for example super
25:59Bas is doing that once a quarter buring
26:02all their news in one week a quarter to
26:04make as much noise as they can other
26:06things you can do make your
26:09documentation a first class citizen
26:12documentation is often an after start
26:14for developers they prefer coding than
26:17actually documenting their code I would
26:19argue that documentation should never be
26:21an after it should be part of the
26:23product documentation is marketing
26:26that's how people are going to interact
26:28with your product that's actually the
26:30first place people are going after your
26:33homepage and maybe the pricing page
26:35stripe and other de tools have really
26:38raised expectation there developers have
26:41become very demanding that something you
26:44should invest in at algolia we decided
26:47that the feature was not ever done until
26:50the dock was done and documentation
26:52should be written by developers you know
26:55Alia even when we Cruise a team and at
26:57some point Point uh we created that
26:59documentation team their role was not to
27:02write the documentation for the
27:03developers they were here to help
27:05developers write better documentations
27:07for example they built a tool where it
27:10would automatically insert API keys for
27:13people who would be logged in so that
27:15when they copy pass the code it would
27:17work right away out of the box uh which
27:19made the documentation ever more useful
27:21for our customers speaking of things
27:23that developers should do I would
27:25include support there you should make
27:28support a huge part of your marketing
27:31your users are developers right
27:34developers hate speaking to these Level
27:38One support people who don't know
27:40anything about the product they are
27:41supporting Engineers should do support
27:45and that's going to to lead to two great
27:47things one your users are developers and
27:50when they interact with developers who
27:52speak their language are going to be way
27:53more satisfied actually I have examples
27:56where someone would reach out to support
27:58and then the developers like the person
28:00doing the support was able to actually
28:02fix the bug and deploy in production in
28:05like 20 minutes and so that kind of like
28:07blew the customer away like they were so
28:09impressed so excited about the product
28:11and two it's going to help your own
28:14developers much better understand your
28:18customers needs because they're going to
28:20face the pain as they supporting them
28:23and that's a great way for them to be
28:25the better product taking the example of
28:26aloya again we actually waited until
28:30hundreds of employees before we built a
28:33dedicated support team and then we
28:36actually asked our best engineering
28:38manager to build that team support was
28:41never a second class citizen for us so
28:44developers are doing documentation
28:46developers are doing support all of that
28:48is your marketing should you even hire
28:50marketing team or when should you do
28:52that well similarly to sales I think
28:56founder should lead marketing
28:58for a long time maybe forever actually
29:00to be honest even at later stage I don't
29:03know a single death Tool company that is
29:06happy with their CMO when they have a a
29:09traditional marketing background why
29:11because traditional marketers they don't
29:13really understand the Persona they don't
29:15really understand developers only other
29:18developers do and developers really hate
29:21being marketed that please don't insult
29:24their intelligence you know that because
29:25you are developers right well early on
29:28at aloria we were actually asking every
29:31single engineer in the team to do what
29:33we called a marketing hack every single
29:36months a hack could be writing a simple
29:39blog post or could be speaking at a
29:42Meetup or could be building hack new
29:45search all of our best marketing at
29:47aloya was done by developers I would
29:50encourage you as you build your
29:51marketing team to make sure that you
29:54have a lot of developers in your
29:56marketing team because they are the ones
29:57who know how to speak to your customers
30:00you may ask what about Dev advocate well
30:03I love the idea because that means
30:05having developers in your marketing team
30:07but it's also a very IL defined role
30:10with often a lot of very fuzzy
30:13expectations so it's difficult to
30:15measure difficult to build
30:18accountability for so my recommendation
30:20here would be to wait and first leverage
30:24your existing engineering team and ask
30:27them to do some form of marketing or Dev
30:30advocacy and probably at Ser day or
30:33later it start to make sense to hire
30:35dedicated Dev Advocates even then think
30:37about hiring from within or from your
30:40committee for example if your open
30:42source could be contributors or active
30:45or helpful Discord members these are the
30:48best Dev Advocates candidates to join
30:50you so let's wrap up and recap what you
30:53said here start now please don't wait
30:57for the perfect idea because as you are
30:59going to build your product even if it's
31:00not the perfect idea you'll do the Reps
31:02you learn and that will lead you to the
31:05best idea if you don't start you'll
31:08never get anywhere build quickly please
31:11please don't over engineer you need to
31:13have that very fast iteration Pace where
31:16you can learn from your users as you are
31:19building crappy quick and dirty is
31:23completely okay early stage and then you
31:25refactor what actually matters spend
31:27time with with your users like there is
31:29no way you're going to learn if you
31:31don't speak with your users learn from
31:33them do things that don't scale go to
31:35their office spend time with them launch
31:37early you might have seen many of our
31:39videos where we say there is no shame in
31:41launching early people won't remember
31:43you if it doesn't work out but you could
31:45get invaluable feedback that will lead
31:48you to work on the right product to
31:50build the right thing that people need
31:53as you are building a de tool you should
31:56consider open source as your Market
31:58observe the market observe other players
32:01how they build their product and figure
32:03out if open source is the right thing
32:05for you but don't dismiss it think about
32:07it add before making a choice don't
32:10forget you are the best sales person for
32:13your product no one else is going to
32:15know your product or your audience as
32:17well as you do learning to sell is way
32:21easier than learning how to speak with a
32:23developer and finally you are also the
32:26best marketing person for your product
32:29you know the language of developers you
32:31know how to speak with them you know
32:32where they're spending their time the
32:34community you know how to access this uh
32:37audience okay that's wrapped up for me
32:40today thanks for watching the video and
32:42if you are building a Dev tool don't
32:44forget you can apply to YC anytime
32:48[Music]

About this transcript

This page contains the full transcript of How To Start A Dev Tools Company | Startup School by Y Combinator, generated from the public captions YouTube serves with the video. The transcript has 5,473 words across 773 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.