YT YouTube2Text

YouTube2Text → Transcripts → If You Ask Claude 5 Questions, It'll Make You Build 10x Faster

If You Ask Claude 5 Questions, It'll Make You Build 10x Faster — Transcript

by Austin Marchese · 5,038 words · 728 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00I just interviewed Ryan Oyang, an AI
0:01engineer at WAP, a startup worth over $5
0:03billion. And he said something that
0:05changed how I work.
0:06>> We've helped millions of people use AI,
0:08and the fastest builders have all
0:10answered these five questions.
0:11>> So, I had Ryan ask me these five
0:13questions. And it turns out if you use
0:14Claw to answer these simple questions,
0:16you will build 10 times faster. So, in
0:18this video, I'll break down the exact
0:20questions he asked me, how they change
0:22what I do, and then the exact prompts
0:24you can run to answer these questions
0:25yourself. And at the end of the video,
0:27I'll share what Ryan said is the most
0:28common problem people hit after they
0:30answer these questions.
0:32>> I'm Ryan O Yang and this is the age of
0:34AI.
0:34>> So the first question, what's your
0:36context and what are your desired
0:37outcomes? Claude has PhD level general
0:39knowledge, but it knows nothing about
0:41your specific situation, your goals,
0:43your limitations, your restrictions, and
0:45what winning specifically looks like for
0:47you. So I asked Ryan how which is
0:48one of the fastest growing startups on
0:50the planet, answers these questions
0:51internally so that they can build
0:53faster. You can think of as a
0:55business that helps you run your
0:58business. If I give Claude the correct
1:01context about what is and where I
1:03want to expand, it's it helps me get
1:07crystal clear on the steps I need to
1:09like get there. So, for example, we
1:13have all these different products at
1:15And we saw that more and more
1:17agents were trying to use our products.
1:19And by giving Claude that context, it
1:21can suggest things like, oh, what it
1:23sounds like you need to do is build a
1:26really easy to use CLI.
1:28>> So, I want to break down what he just
1:29described. The context provided to
1:30Claude didn't just make the answer
1:32better. It pointed Claude at the right
1:34problem to look at. And in Ryan's case,
1:36it was about making something that's
1:38easy for AI agents to interact with. And
1:40getting this answer right sets the
1:41foundation for everything that we're
1:42going to cover in this video. In the age
1:44of AI, we need to build things that
1:45solve real problems because that's how
1:47you provide value. The fastest builders
1:49don't do the most things. They make the
1:51biggest impact. And Ryan, he helps
1:53millions of people use clawed plugins.
1:55So his context is different from mine
1:57and yours. And that's totally fine. The
1:59point isn't to copy his context. It's to
2:01write down your own. So to establish
2:03your context, to make sure you're
2:04actually building the right things, what
2:06exactly should go into it? The first
2:07thing that you want to do is tell Claude
2:10about your current situation as
2:11accurately as possible. So, what does
2:14your business do? Who does it target?
2:16Where are you located? How much effort
2:19are you putting into your business?
2:20>> This context informs the direction. And
2:22notice the one thing that seems
2:24unnecessary. Where are you located? It
2:26might feel irrelevant, right? But your
2:28location, your skills, and how much time
2:29you have, this all influences what the
2:32right answer is for your situation. And
2:34so the second half of this first
2:35question is your desired outcome. The
2:36goal itself is usually easy, right? I
2:38want more customers. I want to save
2:39time. But the hard part and the most
2:41effective builders, make it concrete.
2:43And the second you give Claude the
2:44correct context, it can better turn that
2:46abstract goal into one number and one
2:49deadline. On screen is an exact prompt
2:51that will force you to provide context
2:52and then drive you to a desired outcome
2:54based on these goals. And when
2:56describing your desired outcome, make
2:57sure to level it up. So in Ryan's case,
2:59it wasn't make an AI agent tool. it was
3:01help our customers interact with our
3:03tools using AI. If you go a level lower,
3:06a task gets optimized as an isolated
3:08job. But when you layer it up, Claude
3:10can make more strategic decisions. So
3:12when I was working with Ryan to answer
3:13this question myself and I provided the
3:15context around build partner.ai, my
3:17tool, we use the CLI, which you can
3:18install by running this command, but
3:20this effectively allows you to use WOP's
3:22tools directly in Clawude. And then once
3:24it's set up, because I have pulling
3:25my website analytics, it can see all of
3:27my website visitors. And because has
3:30industry data from top performing
3:31websites on its platform, it didn't just
3:33hand me a number. It was comparing my
3:35current product to real benchmarks. And
3:38so that's how I can pull the desired
3:39outcome, not out of thin air, but based
3:41on real numbers. And when doing this and
3:43trying to answer this question on what
3:44my context and desired outcome was, it
3:47became clear that even though I wanted
3:49more monthly recurring revenue, the
3:50problem I needed to solve was improving
3:52my conversion rate. So once we provided
3:54Claude with the right context to make a
3:56clear desired outcome, what should you
3:58actually build to get there? Brings us
4:00to the second question. What's the most
4:01important thing to build? And this is
4:03about defining the MVP. So this question
4:05actually has two parts. The first is
4:07find what only you can do. And then
4:09second, find the minimum steps to one
4:11real result. So to help us answer this,
4:13I actually asked Ryan how they think
4:14about it at And listen to how he
4:16describes your unique ability or your
4:18secret sauce. And then at the end of the
4:19clip, I'll describe how you can identify
4:21your own secret sauce. What we spend our
4:23time doing is ruthlessly prioritizing
4:27what is actually unique to our business.
4:30Like what can WAP do that no one else
4:34can do? And that's usually like the most
4:36most most most high lever thing to focus
4:39on. And for us it's really about our
4:41ecosystem. the millisecond that you get
4:43a payment on you can spend it on a
4:45W ad or you can spend it on your W card
4:48to maybe pay rent for your office or pay
4:51a contractor or you know buy something
4:53that helps you grow your business even
4:55faster. I think like the web of things
4:57that you can do after you make money on
4:59is incredibly important and is part
5:01of our secret sauce.
5:02>> So secret sauce is this ecosystem,
5:04but how do you find your secret sauce?
5:06So this is a bit abstract, but your
5:08sauce is your unique ability. That is
5:10the hard thing that you can do that
5:11others can't.
5:12>> Your location, your skills, how much
5:15time you have to spend on the business,
5:17like these are all things that make you
5:18unique. For any operator, for any
5:20executive, for any just person who wants
5:23to be more efficient, you should just
5:24ask yourself questions of like, what do
5:26I like to do? What am I good at? And
5:29that you usually get to your sauce
5:31pretty quickly. A good way to think this
5:32is actually there's a perfect example in
5:34Silicon Valley. In the pilot episode,
5:36Richard's Pied Piper is a forgettable
5:38music copyright app. Then Julie and
5:40Peter Gregory spot the compression
5:42algorithm hitting inside it and then a
5:45bidding war starts. The value was
5:46actually in the one thing that only
5:48Richard could build. That was his unique
5:50sauce. Now, once you've identified this
5:51unique sauce for your situation, there's
5:53actually two ways to implement this
5:55depending on what you're working on. And
5:57so, the first is streamline and automate
5:58what isn't your unique ability. And this
6:00allows you to focus your human energy on
6:02what makes you unique. In turn, helping
6:04you save time on the other things. And
6:06the second is double down on what makes
6:08you unique and figure out ways to
6:09enhance that unique ability to in turn
6:12increasing the value you provide. For
6:14either one, this unique ability can
6:15serve as a measuring stick. So, right,
6:17are you working on your unique ability?
6:19If no, how can I speed up these tasks?
6:21If yes, how can I double down on this?
6:24This is how you build faster, not by
6:26doing more, but by pointing your time at
6:27the things that deliver value faster.
6:29Now, whichever bucket you fall into,
6:31your focus needs to be on creating your
6:33minimum viable product. And so, I asked
6:34Ryan how he thinks about creating an
6:36MVP. And he told me about a car
6:37detailing business he started using
6:39WOP's own tools.
6:41>> Let's just start really dumb, right?
6:42Like, what do I need to do to get my
6:45first customer? Well, I have to talk to
6:47someone who has a dirty car. Simplest,
6:50simple, simplest. I'm gonna put some
6:52flyers on people's cars. And then how am
6:54I gonna deliver the service? Well, I
6:56guess I have to buy a pressure washer. I
6:58guess I have to buy soap. I guess I need
7:00to buy like a wash mitten, one of those
7:02sponge thingies.
7:04Boom. Now I have a to-do list of like
7:06things I need to buy from Home Depot.
7:09And what I have to print and what I have
7:11to do to get my first customer. And I
7:13would say that is literally the MVP of a
7:15car wash business. So with this in mind,
7:16you can ask yourself, what do I actually
7:18need at the bare minimum to get one
7:21result? That's the core of MVP thinking.
7:23Start with the result you want, work
7:25backwards, and cut everything that isn't
7:27necessary to get there. And I do want to
7:29call out that this isn't just for new
7:30businesses or new products. MVP thinking
7:32works on everything you're already
7:34doing.
7:34>> If your metric is saving time, well,
7:37then you ask yourself questions about
7:39like, where am I wasting time? This is
7:41just a general process that you can
7:43constantly do no matter what state
7:45you're in.
7:45>> Exactly. Like we help people run their
7:47business, but you could just as easily,
7:50you know, run this for um your job, your
7:54process, even like your health, you
7:56know? I think this applies for almost
7:58any goal that you have.
7:59>> So, that's Ryan describing it, but
8:00here's how you can put it into practice.
8:02First, take the metric you chose in
8:04question one and list the actions that
8:06could move it. Then narrow that list
8:08down to the minimum steps needed to test
8:10one high lever outcome. Now, I'll dive
8:12into this more in the next question Ryan
8:14asked me, but by cutting scope on
8:16anything you're building, you're
8:17focusing on the high signal components
8:19that make this valuable. And honestly,
8:21the hard part isn't doing this. It's the
8:23strategy of knowing which move is worth
8:25your time. So, how do I think about this
8:27strategically? Well, when looking at my
8:28tool, build partner.ai, I have my whole
8:30product hooked up in so it sees all
8:32the website traffic. I can then use
8:34WOP's economic intelligence tool to help
8:36me determine high leverage tasks I can
8:39complete for my business. And because
8:41has the data on what's working
8:43across businesses on its platform, it
8:44can point at the highest leverage move
8:46better than almost anybody because it
8:48has access to this unique data. So that
8:51way it'll recommend something concrete
8:52like fixing a landing page or starting
8:54ads. Then I'll ask to create a list of
8:56minimum steps to test it. Now, if you're
8:58starting a business or you're scaling a
8:59business, if you don't have WOW
9:01configured, you can use the first link
9:02below to sign up, which I am affiliated
9:04with transparently. But either way,
9:06here's a prompt to help you
9:07strategically answer this question about
9:09what is the most important thing to
9:11build. So, at this point, you've
9:12provided context and your desired
9:14outcome to Claude, and you know the one
9:15thing to build for your MVP. Now, those
9:17first two questions, a lot of people do,
9:19but this next question, 99% of people
9:21skip, but it is more important than the
9:23first two. Question three, what are you
9:25not building? So, this is the biggest
9:26problem with building with AI. It's not
9:28knowing what to build, it's knowing what
9:30not to build. So, if you found what you
9:32can do uniquely, Ryan's argument is that
9:34you should spend your time on nothing
9:36else.
9:36>> Once you have your goal and once you
9:39know what your sauce is, I would say the
9:41most important question to ask is what
9:43should you not build? And this when I
9:46say like ruthless focus on the things
9:49that are important, it it sort of
9:51automatically means you shouldn't be
9:53focusing on things that are not
9:54important. And he is right. Right. These
9:56other things can just be so distracting
9:58because building is so easy now that
10:00everything can feel like it's worth
10:02building. And here's him talking about
10:03this building addiction loop.
10:05>> Get addicted to building everything. You
10:08know, everything is shiny. You know, I
10:10as a programmer, I would love to build
10:12my own database and like ah yeah, dude.
10:14like look at my cool database console
10:16and it's got all these great colors and
10:18like I'm like benchmarking performance
10:20like oh I can save like you know a bunch
10:22of milliseconds here on this database
10:24query but like who cares like it it
10:27doesn't matter um and I should know that
10:30it doesn't matter honestly at the
10:32beginning because it stops me from going
10:34down these rabbit holes that um honestly
10:37cause me to lose sight of the original
10:39focus and the goal that I had. And the
10:41key word he says here that I want you to
10:42focus on is, "But who cares? It doesn't
10:45matter." Oh, you build a tool that's 80%
10:47as good as a tool we already use. Great.
10:49It doesn't matter. Oh, you build
10:50something that doesn't actually provide
10:51value. Great. Oh, you built something
10:53that's really cool looking, but it
10:54doesn't save me time or help me make
10:56more money. Great. It doesn't matter.
10:58And so, one real benefit here of what
11:00you are not building is it makes it
11:01super clear what you shouldn't do. Your
11:04job as the manager of Claude is to
11:06connect the pieces, not to build all of
11:08the pieces. And if you ignore that,
11:10you'll end up with an AI tool graveyard
11:11where you'll have 10, 20 half-used
11:13tools, none of them actually doing
11:15anything. And so for build partner,
11:16there is just a bunch of things that
11:17I've elected that I will not build. Do I
11:20want to vibe code a payment platform
11:21that has to figure out how to accept
11:23payments in 195 countries? No. Instead,
11:26I just tell Claude, "Use to set up
11:28payment for this product." And it just
11:30does it. And now I can get payment
11:31without building anything. Do I want to
11:33figure out how to run ads across Tik
11:35Tok, Instagram, YouTube, Facebook, and
11:36Google search? No, I'm not an ads
11:38expert. So, I'll just use CLI from
11:41Claude and tell it I want to configure
11:42ads for these platforms and it just
11:44doesn't. I don't have to think about
11:45setting up accounts on every channel,
11:47optimizing CPMs. I'm leaning on the
11:49tool to do it. And so, the point is, if
11:51you want to build faster, you can't
11:52focus on things that don't matter and
11:54that you shouldn't build yourself. Now,
11:56there is one word of caution here when
11:58exploring the use other tools framework.
12:01And here's Ryan addressing this exact
12:02issue that you may run into once you get
12:04tool trigger happy.
12:05>> I don't want to manage 10 tools. my
12:07agent doesn't even want to manage 10
12:08tools. It's about maximizing value per
12:10tool
12:12versus maximizing the number of tools.
12:14>> So, the goal is to look for tools with
12:16robust functionality to cut down on what
12:18you need to build. Obviously, here we're
12:20talking about W, but I chose them
12:22because I use them for payments, help me
12:23run ads, business intelligence, and my
12:25website analytics. I do not want to
12:27build any of those, and I wanted a
12:29single tool that does all of these
12:30things. So, to help you answer this
12:32question, run this prompt, which will
12:33look at what you're building and what
12:34you spent your time on in the past, and
12:36then help you identify what you should
12:37not be building. Once you've done that,
12:39run this prompt, which will help you
12:41identify any external tools that can
12:42help you with what you're trying to do.
12:44At this point, we've answered three of
12:46the five questions. Now, before we get
12:47to the last two, and then a word of
12:49caution from Ryan about what you need to
12:51do after answering these five questions,
12:54you guys all know the drill. This is our
12:56anti-slap agreement. Everything in this
12:57video is built by humans for humans. So,
12:59all that I ask as part of this agreement
13:01is you subscribe to help this content
13:02reach more people. And if you already
13:04subscribed, make sure to turn on post
13:05notifications. That really helps as
13:06well. And as a thank you to everyone, I
13:08appreciate all of you. I give away a
13:10Claude Max subscription every video. So,
13:12this video's winner is T. Carson5045,
13:14who's building agents for narrow AEC
13:17pain points. And to enter this video's
13:19giveaway, comment below with what you're
13:20building or what you're deciding not to
13:22build. And a fun fact here that I want
13:24to share with all of you, Ryan and I
13:26actually built our last startup
13:27together, IK, which raised over $30
13:28million. So for everyone building,
13:31double down on your human connections
13:32because you never know where your career
13:34is going to go. But anyway, we've gone
13:36through three of the five questions.
13:38Question four is, what are your building
13:39hot zones? Hot zones are critical parts
13:42of your work that you need to get right.
13:43For everything else, just let AI carry
13:45the torch and you monitor the outcome.
13:47For me, within build partner, we have a
13:49custom plugin that helps people build
13:51faster with Claude. And this goes out to
13:52over 5,000 people when I update it. So,
13:54this is a hot zone to say the least. So,
13:56I asked Ryan, "What do the most
13:58successful businesses on do when
14:00thinking about this?"
14:01>> All the successful business owners that
14:03we see on are really dialed on
14:06making sure that um their a they're
14:10really working with their AI to spend a
14:12lot of time on the things that like
14:14really are the most important to their
14:16business.
14:16>> But from there, he does add one really
14:18important caution to this. You can't
14:20just let AI run free. These hot zones
14:22are areas of what you're building to be
14:24extra focused on. On screen, you can see
14:26a prompt that will help you identify
14:27what your hot zones are. And then what I
14:29like to do is I I add in my claw.md
14:31anytime I'm working with a hot zone,
14:33flag it with me and make sure I approve.
14:35And so again, for build partner, these
14:36hot zones and how I think about it is
14:38the plugin as well as our WP payment
14:40flow. I want to make sure that the
14:41payments is working properly. But this
14:43is just part of it, right? How do you
14:45actually work in the hot zones
14:46themselves? A term that you've likely
14:48heard is eval, which is short for
14:50evaluations. And this is the process of
14:52testing your changes to make sure it's
14:55getting better, not worse. The fastest
14:56builders know when to measure twice and
14:58cut once. And in hot zones, it makes
15:00sense to establish evaluation criteria
15:02when needed. Evals are definitely a
15:04large topic, and I might have to make a
15:06full video on this, but it's important
15:07that you have a strong foundational
15:09understanding of this process. So, what
15:11should you do today to implement your
15:12first eval? This is what I call building
15:14a minimum viable eval. And here's Ryan
15:16talking about how to do this.
15:18>> Claude's really good at making eval
15:20nowadays. Um, everyone's is going to
15:23look different, but um, it's generally
15:26going to be your input, which is the
15:30thing that you're changing. In my case,
15:32it's the skill file. There's going to be
15:34some program that Claude helps you
15:36write, which like runs all of these,
15:39call it test that does the squats for
15:40you. And then at the end you get a
15:42report which is how your input did
15:46against the suite of tests that you ran
15:49against it.
15:50>> So from that think of it like a simple
15:51equation. You have your input and your
15:53output. This is the whole grading
15:54mechanism. And then you have what you're
15:56changing. This is the variable that
15:58you're adjusting to see what the output
16:00becomes. And you may be thinking this is
16:02just for technical work but it isn't. In
16:04our conversation we covered two kinds of
16:05evals. The first is a standard eval
16:08which uses a programmatic approach to
16:10evaluate quickly through testing cycles.
16:12This is typically when you hear about
16:13people talking about eval loops and they
16:16automatically improve themselves. In
16:17order for that to work it has to be
16:18extremely programmatic. The second slow
16:20eval are for non-technical work and this
16:22involves more human judgment for molding
16:24the evaluation criteria. So to put a
16:26real world example of this I did a slow
16:28eval for a skill that I built for an
16:30engineering firm that I worked with.
16:31Let's call the skill slashgenerate
16:33report. And so we would use a skill to
16:35generate the report. We would take that
16:37output. We would get feedback from the
16:38engineering firm that we worked with and
16:40then we would adjust the skill until the
16:42output was approved by them. And then we
16:43would store that approved version. Then
16:45the next time we would run the report
16:46and they would provide feedback. We
16:48would change the skill but we would
16:50rerun it on all the past approved
16:52reports to make sure that nothing breaks
16:53and it doesn't regress for what those
16:55previous edge cases were. And so you can
16:57imagine over time that we build a test
16:58suite for these reports and it can run
17:00this evaluation criteria over time to
17:02get better and better. And this process
17:04isn't instantaneously, but this slow
17:06eval process can really make a
17:07difference. Now, to identify any
17:09specific hot zones that are worth
17:10building a minimum viable eval for, use
17:13this prompt, which will help you get
17:14started. Now, the last question that
17:15Ryan asked me is what I think is the
17:17most underrated. Question five, what are
17:20your bottlenecks? And you may be
17:21wondering, I'm not sure where my
17:23bottlenecks are. And so, Ryan thinks
17:24about it as something that you identify
17:26as you run towards your goals. Let's say
17:29you're trying to scale your revenue and
17:32you get your conversion rate to a good
17:34place. Your bottleneck is how many
17:36people actually land on your website in
17:37the first place. You're converting all
17:39of them, but like no one ever lands
17:40there, right? Or you could have the
17:41opposite problem. Oh, so many people
17:43land on your website, but none of them
17:44check out. What? Like your bottleneck is
17:47your conversion rate.
17:48>> And what he discussed here is why you do
17:49questions one through four first. That's
17:51so that we know we're moving in the
17:52right direction. So that these
17:54bottlenecks are relevant to our desired
17:56outcome. But the biggest transformations
17:57are when you find the real bottleneck
17:59and then suddenly one fix can just
18:01change everything. So how do you
18:03actually spot your bottleneck? Often
18:04it's the things that you actually avoid.
18:06>> Everything that people don't enjoy doing
18:09eventually becomes a bottleneck. And
18:10I'll give you a good example. Like a lot
18:12of business owners, they start out being
18:14employees. Maybe you started off as a
18:17website designer and you've made this
18:20killer website and you just love making
18:21websites and you keep improving your
18:23website. Well, you're eventually going
18:25to be bottlenecked by the quality of
18:28your product or maybe how you drive
18:30people to your website because you're
18:32not a marketer, right? But you just
18:33because you love improving your website
18:35so much, it's really hard for you to
18:36switch and like focus on these hard
18:39problems that are uncomfortable. You're
18:41unfamiliar with them.
18:42>> So, a lot of the times you can identify
18:43bottlenecks through what you don't want
18:45to do or what you're not good at or
18:47looking at what you haven't spent any
18:48time on. That's the avoidance test.
18:50Procrastination is wisdom. Use this as a
18:53signal for where your bottlenecks may be
18:55located. Here's a prompt to help you
18:56identify bottlenecks and quantifiably
18:58identify what the CLI can solve for
19:01you. And then if you have the CLI
19:02installed, here's how you can use that
19:04to help streamline these bottlenecks.
19:06Now, at this point, you may have two
19:07questions. How do you know if you're
19:09working on the right bottleneck? That's
19:11where you want to think about effort
19:12versus return. If you feel like you're
19:14putting in way more energy than you're
19:16getting benefit out of, you likely
19:17aren't focused on a bottleneck. You're
19:19focused on something else. Then how do
19:21you know if you've solved a bottleneck?
19:23The best way to do this is use
19:24benchmarks. If you look at something
19:26like your conversion rate and it's some
19:27percentage below industry standard,
19:29that's a bottleneck you can still
19:30improve. But in order to do that, that
19:32requires knowing what the standard is,
19:33which is exactly where I'll use WOPS
19:35data to help me get that answer. So here
19:36are two prompts to help you. First, find
19:38your bottlenecks by what you avoid. And
19:40second, use the CLI to put an actual
19:42number to this bottleneck. So after
19:44these five questions, what's the part
19:46Ryan said almost everyone gets wrong?
19:48And this is the most common problem. You
19:49only ask these questions once. The
19:51answers that you provide to all five of
19:53these questions, the reality is it has a
19:55shelf life. What worked on day one isn't
19:57enough on day 100 and you couldn't have
19:59known the right answers when you started
20:01because things change. So you have to
20:03schedule periodic reveals. Here's how
20:04does it and the cadence I'd copy.
20:07>> Every quarter we find out what our
20:10biggest desired outcome is. What what
20:12are we trying to improve about the
20:14business of right? is a
20:16business. We happen to help people build
20:18businesses, but we are also a business
20:19and there's different things to focus
20:20on. And then week over week, we might
20:23think about, okay, what is the most
20:25important thing that's going to move us
20:26to that outcome?
20:27>> Whenever you're working on something,
20:28you can bucket your work into two
20:30categories. The first is execution and
20:32then strategy. Execution is for when you
20:34know exactly what's needed, and you want
20:35to revisit this weekly, if not daily.
20:37This is your list of to-dos. And for me,
20:39I literally just run this on an Apple
20:41Notes every day. I have my items that I
20:43have to do. Question four and five that
20:44we covered in this video. That's where
20:46this falls. What eval are you proving?
20:48What metric did it move? What are you
20:50working on today to push you toward the
20:51desired outcome? In my eyes, execution
20:53is actually really easy. [music] The
20:55second bucket is harder. Strategy. This
20:57is where you're planning what's needed.
20:59But if you do this every day, then you
21:00get stuck in strategy and you don't
21:02actually build anything. So
21:03realistically, you want to revisit this
21:05monthly or quarterly. And questions one
21:07and three are strategic. That's what's
21:09my sauce? What am I not building? The
21:11goal is to find a goldilock zone.
21:12execute quickly so you avoid analysis
21:15paralysis, but make sure it's actually
21:17in the right direction. And if you skip
21:19reasking yourself these questions,
21:21here's what can happen.
21:22>> If people don't stop and ask themselves
21:24questions about like what's happening,
21:26what they want to do, what's giving them
21:28joy, what they're good at, they would
21:29never have discovered these other
21:31opportunities and they'd be maybe like
21:34incorrectly focused on that one thing
21:36that they identified maybe like many
21:38months or even years ago. So, what you
21:40need to do today is put a recurring
21:42event on your calendar. Weekly for
21:43execution, monthly or quarterly for
21:45strategy, and don't overthink it. So,
21:47ask yourself these five questions to
21:49make sure you're building the right
21:50things. First, what is your context and
21:53desired outcome? Second, what's the most
21:55important thing to build? Third, what
21:56are you not building? Four, what are
21:58your hot zones? And five, what are your
22:00bottlenecks? And then, you need to ask
22:02them again on a specific schedule. So, W
22:04helped me put this all into practice,
22:06right? the data behind my numbers, the
22:07CLI that pulls them in, and the tools
22:09that mean I don't have to build
22:10payments, a website, or analytics
22:12myself. So, if you want to try
22:14click the first link in the description,
22:15and shout out my boy Ryan. I love having
22:17friends who are smarter than me building
22:19awesome things so that I can work
22:20together with them. Now, one thing that
22:22I did discuss here is knowing the
22:23importance of when to use the right
22:25tools. is one of them. But if you
22:27want a breakdown of eight free tools
22:29everyone watching this video should use,
22:31go check out this video. If you like
22:33this, I guarantee you'll love that. I'll
22:34see you over there. Pacing.

About this transcript

This page contains the full transcript of If You Ask Claude 5 Questions, It'll Make You Build 10x Faster by Austin Marchese, generated from the public captions YouTube serves with the video. The transcript has 5,038 words across 728 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 choose a caption track available on the video.

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, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.