A Live Q&A on Delivering Claude Automations to Clients
A daily-updates vlog where the host works down his YouTube comments live, answering questions on Claude client delivery, niche selection, cold email cutoffs, and lead sourcing, then turns the camera on his own subscriber growth curve.
Posted
yesterday
Duration
Format
Talking Head
educational
Views
3.6K
71 likes
57 · 43
Big Idea
The argument in one line.
Scalable AI-agency work comes down to picking niches that are digitally scrapable, high ticket, and low compliance, then backing every offer with a real money-back guarantee so the incentive to deliver falls on you, not the client.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
You run or want to run an AI automation/consulting agency and need a concrete way to filter which niches are worth targeting.
You're doing cold email outreach and want a real benchmark for when to cut a non-performing campaign.
You build automations in Claude and need to hand them off to a client without becoming their permanent auth/maintenance bottleneck.
You're building a YouTube channel around a business and want to see how a creator with 500K+ subscribers reasons about growth mechanics in real time.
SKIP IF…
You're not doing any client services or outreach work; the niche and cold-email sections won't apply.
You want a polished, scripted tutorial. This is an unscripted comment-answering vlog with tangents.
TL;DR
The full version, fast.
Nick Saraev works down his YouTube comment queue live, answering entrepreneur questions about his AI agency work. The throughline: deliver Claude automations to clients by building under their own routines/connector auth so you're never a 2FA bottleneck; pick niches that are digitally scrapable, high ticket, and low compliance rather than cash-poor local businesses; cap a cold email variant at roughly 1,000 sends and look for a ~1% positive reply rate before scaling it; and always attach a real money-back guarantee to an offer because the incentive structure, not comfort level, determines effort. He closes by reviewing his own channel analytics and admitting his subscriber growth has gone linear as a percentage of his base, and that he's rebalancing between long-form courses and higher-frequency videos to fix it.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
States the channel's premise: solve real problems, answer subscriber questions from the comments, and think out loud about what's next.
01:01 – 02:06
02 · Claude automations for clients
Answers how to deliver a Claude-built automation to a client without becoming their permanent auth bottleneck: build under the client's own Claude account using connector OAuth and the Routines feature.
02:06 – 02:18
03 · AI video editing
Quick aside: confirms he does not use AI tools like Remotion to edit his long-form or short-form videos.
02:18 – 05:28
04 · Choosing a niche
Lays out the three-filter test for a viable AI/automation niche: digitally scrapable, high ticket, low compliance. Explains why local businesses like dog walkers underperform versus MSPs.
05:28 – 07:25
05 · Cold email testing
Gives a concrete threshold for cutting a cold email variant: roughly 1,000 sends per copy, targeting about a 1% positive reply rate before scaling further.
07:25 – 09:30
06 · Lead filters
Argues a thin lead pool is usually a filtering problem, not a supply problem, and recommends filtering primarily on 'runs ads' rather than stacking website/email/ad filters.
09:30 – 11:07
07 · Offering guarantees
Makes the case for always attaching a money-back guarantee to an offer, framed as an incentive structure rather than a promise, using the 'X result in Y days or your money back' formula.
11:07 – 12:10
08 · Maker Zero approvals
Explains that Maker Zero has no waitlist; rejected applicants are usually caught by an auto-flag system, and manually reviews/approves people who message him by name.
12:10 – 13:39
09 · Crowded lead lists
Addresses a question about scraping dental leads: shared scraper pools (Apify, Apollo, LinkedIn Sales Navigator) are oversaturated; custom-built lists from niche directories perform better but don't scale as easily.
13:39 – 14:54
10 · Hiring for YouTube
Reviews applications for a packaging/clipping/thumbnails role, noting most applicants lacked a title/thumbnail portfolio, and that roughly 5 of ~60 applicants will move to a trial.
14:54 – 18:57
11 · Posting more videos
Reviews his own publishing cadence: only 4-5 videos in the last month due to a deliberate long-form experiment, and how that hurt YouTube's browse-feature distribution. Plans a mix of one long-form course a month plus 10-30 shorter videos.
18:57 – 21:01
12 · Channel growth
Walks through channel analytics across YouTube, Instagram, and the Stacked podcast, and explains why linear subscriber growth is actually a declining relative growth rate at 520,000 subscribers.
21:01 – 21:13
13 · Outro
Closes with a reflection on ad revenue earned over two years and a sign-off.
Atomic Insights
Lines worth screenshotting.
The simplest way to deliver a Claude automation to a client is to build it under their own account using Claude's connector OAuth, so they authenticate once and you never have to ask them for 2FA again.
A niche is worth targeting only if it clears three filters: it's digitally scrapable, it's high ticket, and it has low compliance requirements.
The effort required to market to a $200/month dog walker and a $10,000/month MSP is nearly identical, so target the one that pays more.
Beginners should avoid high-compliance niches like healthcare, finance, and insurance because the regulatory overhead eats the time savings automation is supposed to provide.
Cap each cold email copy variant at roughly 1,000 sends before judging it; a positive reply rate near 1% is usually strong enough to scale, since 30-50% of positive replies convert to booked meetings.
If five of six cold email campaigns sit around a 0.09% reply rate and one sits at 0.8%, kill the five and build new variants off the one that's working.
Complaints about a thin lead pool are usually a filtering problem, not a supply problem: there are roughly 100,000 marketing agencies in North America alone that have a website, run ads, and have a public email.
A 'has a website' or 'has a valid email' filter is often redundant: if a business runs ads, it already has a website to send that traffic to.
Offering a money-back guarantee changes your own effort level, not just the client's confidence, because your own money is now on the line.
A guarantee should be framed as a trade, not a promise: 'I will get you 20 booked meetings in 60 days, or I'll give you all your money back.'
Maker Zero has no waitlist; most rejected applicants were auto-flagged by the platform, and the creator manually reviews and approves anyone who reaches out by name.
Scraping leads from shared, widely-used sources like Apify's Google Maps scraper, Apollo, or LinkedIn Sales Navigator means competing with everyone else pulling from the same pool; the best results come from custom-built lists from niche directories and conferences, at the cost of scale.
A linear subscriber growth rate is actually a shrinking relative growth rate: one new subscriber a day on a 100-subscriber channel is 1% growth, but the same one subscriber a day on a 520,000-subscriber channel is roughly 0.0002%.
Publishing only four long-form videos in a month, even high-quality ones, tanks a channel's browse/suggested traffic because YouTube's browse features reward frequency, not just depth.
The creator's plan to fix flattening growth is a hybrid cadence: about one long-form course per month for depth, plus 10-30 shorter videos per month to rebuild browse-feature volume.
Takeaway
Three filters, one guarantee formula, one growth-rate reality check.
WHAT TO LEARN
Client delivery, niche selection, and outreach all reduce to a handful of concrete decision rules, and the same math that governs cold email cutoffs also explains why a bigger channel's subscriber growth feels like it's stalling.
02Claude automations for clients
Build a client's Claude automation under their own account using connector OAuth and the Routines feature, so you never become the ongoing 2FA bottleneck.
03AI video editing
Confirm the tools you use aren't AI-editing your video; the value in a build-in-public channel is in the narration, not automated post-production.
04Choosing a niche
Filter any potential niche through three tests before targeting it: is it digitally scrapable, is it high ticket, and is it low compliance.
Match your outreach effort to ticket size; a dog walker and an MSP cost the same effort to market to, but pay very differently.
05Cold email testing
Cap a cold email copy variant at roughly 1,000 sends before judging it, and treat a ~1% positive reply rate as the threshold worth scaling.
When several campaigns underperform and one outperforms, kill the underperformers and build new variants off the winner instead of spreading effort evenly.
06Lead filters
Before assuming your lead pool is too small, check whether you're stacking redundant filters; a 'runs ads' filter alone already implies a website exists.
07Offering guarantees
Attach a real money-back guarantee to high-ticket offers, framed as a trade ('X result in Y days, or your money back'), because it changes your own effort level.
08Maker Zero approvals
Expect most low-effort community or job applications to get auto-flagged by filtering systems; a short, direct message with your name can get a manual review.
09Crowded lead lists
When scraping from shared, widely-used sources, expect saturation; a custom list is higher quality but takes more manual effort to build and won't scale as fast.
11Posting more videos
Publishing fewer, longer videos in a month trades away YouTube's browse-feature distribution, which rewards frequency independently of depth.
12Channel growth
Track subscriber growth as a percentage of your existing base, not an absolute number, since the same daily gain represents a shrinking growth rate as a channel scales.
Glossary
Terms worth knowing.
PRR (positive reply rate)
The percentage of cold emails sent that get a genuinely interested reply, as opposed to any reply at all. Used as the main go/no-go metric for scaling a cold email campaign.
Maker Zero
A private community/cohort the creator runs, with an application-based approval process rather than a waitlist.
Claude Routines
A feature that lets an automation run under a client's own Claude account and connected services, so the client authenticates once instead of the agency holding ongoing access.
Browse features
YouTube's suggested/home-page video recommendations, which reward channels that publish frequently; distinct from search traffic, which rewards depth and keyword relevance.
MSP
Managed service provider — an IT services company, used here as an example of a high-ticket B2B niche worth targeting over lower-ticket local businesses.
Resources
Things they pointed at.
01:05toolClaude Code / Claude Cowork / Claude Routines
“If there are two universes, one where you offer a guarantee and another where you don't offer a guarantee, in which universe do you think you're going to try harder for the client?”
Tight rhetorical frame that reframes guarantees as an incentive tool, not a risk→ IG reel cold open↗ Tweet quote
03:10
“The amount of effort that is required to market to both the dog walker's clients and to some big enterprise MSP is actually very similar. So why not just target the guys that make a lot of money for the client?”
“I have 520,000 subs. And with every increase in subscriber count, my relative growth rate has gone down.”
Specific number plus a simple math insight about compounding growth curves→ newsletter pull-quote↗ Tweet quote
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
hey what's going on guys this is my daily updates channel where i run you guys through everything that i'm doing in order to grow a business that did 400k last month so everything that you need to know in order to, but how I am running this process is here in this channel. And you guys can ask me at any point in time, any question you have about any of that stuff using our YouTube comment Q and a feature.
That means if you look at the comments down below this video, you'll notice most of them are questions. That's because I will go through and I will answer them. That's sort of the meat and potatoes of my daily updates.
Then I'll build and strategize in public for finally showing you guys my growth stats across YouTube, Instagram, and my products. This is my main channel over here. We are about to take that number.
and add a couple of zeros. And then this is my daily updates channel. What I'm doing is I'm sorting by new.
I have a seven day cutoff and then I'm just going bottom up, answering as many questions as humanly possible. Got an awesome repurposing workflow and really excited to see it all work. Also, I think I'm playing Spotify right now, aren't I?
I'm going to have to turn that puppy off. Let's do that. And then we'll go back here.
Let's see anything super embarrassing here. No. Okay, good.
So question here from Alvar, who says, hey man, curious about the N8n versus having cloud do it internally. How would you deliver an automation for a client that you made in cloud?
Getting auth sounds like a massive pain, and if they want to make changes to prompts in the automations or so on, do you have any thoughts? The simplest and easiest way to build cloud -based automations for clients is to connect to their cloud account. Use the built -in connector feature to have them one -click OAuth into all the services that they need to be part of.
And then when you build your cloud automation, do so under their routines offering and then add the connectors as their auth. You only ever need to do this one time and now you have fully automated agentic workflows that run without you and without you constantly having to check in to the client to ask them, hey, can you get me 2FA and so on and so forth.
So they've solved this completely now, which is awesome. You could do this in a variety of ways. You could do this through...
Cloud Cowork. You can do this through Cloud Code, obviously, and then the Cloud Routines feature. But that is how I would recommend it as of this moment in time.
Nick says, Harsh, do you use AI to edit your long form and short form, stuff like Remotion or something else? No, I do not. Not at all.
Temp Husky says, Nick, I know you said local businesses aren't the best to sell to, and I've learned this too, is they're often cash poor. So what niches do you recommend targeting that have the capital to invest? And what types of front end offers should you sell to those businesses?
So what niches are working in 2026 for AI and automation? Typically niches that have three things in common. The first is you need to be able to scrape the decision makers in that niche digitally.
And so you can't put in general contractors, let's say, who aren't listed on any social media or any directory, because it is simply impossible to scrape their information in a digital way, meaning it's not very scalable to do so. The second is they need to be high ticket.
And so you could sell services to a dog walker that has an average ticket price of, let's say, $1 ,000 a month. probably less. Or you could sell the same service to an MSP that sells packages that range from $1 ,000 all the way up to $10 ,000 a month.
The reason why you're going to want to go with the latter is because if you can get that MSP just one more client, you're going to make that company tens of thousands, potentially hundreds of thousands of dollars over the course of the lifetime of their engagement with them. Versus if you were to do the same with a dog walker, clearly a premium dog walker charging a grand a month.
Let's say on average, it's like 200 a month. You'd make far less. Turns out, magically, the amount of effort that is required to market to both the dog walkers clients and to some like, you know, big enterprise MSP is actually very similar.
And so why not just target the guys that make a lot of money for the client, then you'll be able to say stuff like, hey, I, you know, help generate $45 ,000 in like my first 30 days in this company, you know, because you acquired one client, it'd be much harder to do the same level of social proof if you were working with like a dog walker.
niches that are very high compliance now there's two ways of thinking about this one is that compliance needs regulation needs think high finance think like health care okay think like you know sensitive stuff like insurance or uh or selling financial products one way to think about it is like but nick if i get a client in those niches they're going to be really sticky because i will have accomplished this big like compliance hurdle And that's true.
I think if you really know what you're doing and you have a solid stack to deal with things like PII, personal identifying information, anonymization of data, data warehousing and storage in, let's say, domiciled within your country, you can do really well in an issue. The issue is these regulations are going to change pretty quickly with artificial intelligence.
And most people that are asking questions like this are beginners. And beginners probably should not be working with high compliance industries to begin with because it's just going to take so much more time and energy to get it right. So what a beginner should do is they should focus on industries and niches that are digitally scrapable.
That is, the companies will always have like a website up, they'll have like a LinkedIn profile up, they'll have sources that you could scrape. High ticket, so companies will sell products for significantly more than let's say some dog walker or some small little monthly SaaS. And then ultimately low compliance needs.
And if you can do that, you will crush, you could sell them anything on the front end, aka client acquisition, and you'll do really well. Thank you for this video. No problem.
Nick, hope you're having a great month so far. Just wanted to ask a quick question. It's about figuring out when to cut out cold email campaigns.
I've watched your cold outbound course. Notice you said cut out variants that don't perform well after sending to 500 to 1000 leads, especially with me having six campaigns running all at once. What's the volume that I should have sent for analyzing that it's time to cut out that campaign?
Is it 1K, 2K or more? And what KPIs are the decider for you to cut that campaign? Thank you so much.
Love, Nick. Yeah, I mean, what? What reply rate should you look for before you cut a campaign?
This is going to depend on the specific niche that you're in. Obviously, certain niches have lower reply rates on average simply because the niche has been worked through more often or there's larger or smaller ticket amounts. And so it's harder to sell to a niche of people if the average ticket is super, super high because there's just very few buyers for that service.
However, you got some rules of thumb here that I always come back to. The first is, if you're creating For every individual piece of copy, I would not send more than maybe 1000 emails or so before verifying, you know, if I were to like take all of my other email mailboxes and pump them 100 % into this campaign, could I see results.
And so for me, that's usually like a PRR, a positive reply rate of about 1 % or so, because I can usually book somewhere between 30 to 50 % of those replies to a meeting. So if I'm getting a positive reply rate of you know, 0 .8 % or something on the very first iteration of a campaign and I've sent a thousand emails and that's what it looks like statistically, I'm going to think, hey, shit, that's pretty good.
I mean, I haven't really put too much work into it and I'm already at a 0 .8%. I think I could probably get that to like a one, maybe 1 .5%. Sure.
Why don't we give it a go? But let's say, you know, of your six campaigns, you know, five of them are at like 0 .09 % PRR. And then you have one that's like a 0 .8, like obviously drop the five, look at the one that has a 0 .8 and then just create a bunch of variants based off that.
So anyway, that's typically what I go for. And there's no real like perfect way to know, unfortunately, but yeah, that's how I do it. Tim Husky says, Nick, can you go over how you get such volume in your lead lists?
I'm doing outreach to marketing agencies, pitching speed to lead AI appointment centers. But when I try to scrape leads using Appify's Google Maps or LinkedIn scrapers, the pool of eligible leads, aka have a website to use for email personalization, runs ads for SMBs. and has a valid email is only a few thousand, not enough to sustain 100 cents a day over an extended period of time, can you go through how you are getting so many good leads for cold email?
If you can only find a few thousand marketing agencies that have websites, run ads, or have emails, you're playing the game wrong. There are probably 100 ,000 marketing agencies that have those qualifications. across just North America alone, not even taking into account the European Union.
There are a lot of marketing agencies out there. So you're probably applying additional filtering that you don't understand or know about. Alternatively, maybe your filter source is wrong or the source of truth that it is using to determine whether or not somebody has a website is incomplete or not filled out.
I'd also encourage you to, rather than have multiple of these, just have one. So for instance, if... You are sending to a pool of leads that you qualify based on whether they have a website, whether they run ads, and whether they have a valid email.
Just drop all that. First of all, how does a valid email, obviously, like how else are you going to send? So websites and run ads.
You know that if somebody run ads, they have a website, right? Where the hell else are they going to send the traffic to? So you don't need to have a website filter out a bunch of things because the data is incomplete.
Runs ads, filters out a bunch of things because the data is incomplete. Has a valid email, filters out a bunch of things because the data is incomplete. You just need to run it, you know, filter based off the running of ads.
And so just go on like Facebook, you know, ad library scrape or whatever. Just find a bunch of people that are running ads and then boom, that's like your primary core data source. I guess to the point that I'm...
making is just like use one filter, like have that be your primary filter. Of course, you can filter aside from that based off company size, maybe geography, but like just use that as your core determinant. I'm sure that could be systematized.
Pernith says, Nick, quick question. In your offers, cooperating tutorials, you always use the structure like I'm confident I can get you X in under Y days, or you get your full money back. While I built fairly complex things before, none of them were tied to monetary results.
So I'm not comfortable enough to offer guarantees. How would you approach this? should I offer guarantees even if I'm not comfortable being able to deliver it?
My opinion is absolutely. If there are two universes, one where you offer a guarantee and another where you don't offer a guarantee, in which universe do you think you're going to try harder for the client? In which universe do you think you're going to learn more to improve your own business?
Obviously, it's the one where you offer a guarantee. Why? Because your ass is on the line.
Generally speaking, as the great Jack Roberts once said, show me the incentive and I will show you the outcome. If the incentive is you've guaranteed a client a result and you have to deliver them that result, or you'll give them all their money back, in that universe, you will work far harder than if it's just like, can't guarantee stuff, but just pay me a bunch of money anyway.
It's not nice to hear this, right? Because you're like, well, I should only really guarantee stuff if I absolutely can. But understand that you're not necessarily just guaranteeing stuff.
You're creating an offer where it's like, I will get you this. And if I don't get you this, I will give you this.
And so I will get you 20 book meetings in 60 days. And if I don't get you 20 book meetings in 60 days, I will give you all your money back. That's sort of how it works.
I was misled for a very long time by one of the first companies I worked with that you should never, ever guarantee anything. And that kept me throttled for a very long time. So you should 100 % guarantee things.
I'd say it's like the minimum entry ticket in order to get in these days. Digi says, Nick, how long does it typically take to be approved to join Maker Zero? My brother, Maker Zero does not have any sort of waitlist or anything like that.
If you did not get in, it is almost certainly because school has auto -flagged you. I have no idea why the hell this keeps happening, but a lot of people are auto -flagged. So goodness was auto -flagged.
I'm going to approve them. Sorry for leaking some emails here, guys. This fella here, auto -flagged.
Approve, approve, approve. The really unfortunate thing is, With this offering, half the people that got auto flagged are actual spammers.
And so if I let them all in, then they will literally just like spam the whole community and ruin it. So despite there being a high false positive rate, yeah. And if you want me to approve you, I can do this.
You just need to like send me your name, like actually send me your name. you know, here. I mean, I know it's kind of crappy, but just send me your name and I will literally go through it and do it.
Presumably, if you're the sort of person that is willing to comment on one of my YouTube videos about this, you understand daily updates, you know about it, you know about me, and you know that I go through this whole process, you're probably not going to be a spammer, right? Chance says, hey, Nick. Hey, Chance.
Nick says, Arnab, what do you think is the best way to scrape leads? I want to scrape dental businesses and I'm pretty confused about which path I should take if I want to scrape them. I have a smart lead subscription.
I think I should go for Appify, but the problem with that is the leads on those are already being hit hundreds of times a day. So what should be my strategy for scraping leads? Well, if you are scraping leads using a system that millions of other people have used before, presumably.
there will be a lot of outreach sent to that same group of people. And so this occurs when you're sending to a pooled or shared group of leads like you would with like an Amplify scraper or Apollo or even LinkedIn sales navigator to some extent. The harsh reality is there's just no better or more scalable way to scrape.
And so your options are scrape this group of leads, get really good at standing out and then win or just don't do outreach entirely. The best results I've ever gotten have always been from totally custom lists that I have painstakingly put together using a hodgepodge of different scrapers and conferences and one -off directory listing websites that nobody really knows about except for industry professionals and so on and so forth.
That said, LinkedIn and Appify and Apollo and all these services there are far more scalable, so I just end up using those. My recommendation is, if you're going to be doing that, you've got to learn how to stand out. And that's really just a core of what makes Outreach so successful anyway.
All right, guys. God wants you to grow a beard. The stats are clear.
Thank you guys very much for these awesome questions. I'm going to be answering them all over the course of the next few days. And if you guys have any more, feel free to put them in.
So a lot of people have reached out to me to ask, hey, Abdulalil, how's it going? A lot of people have reached out to me to... offer their services for packaging and clipping and so on and so forth.
I would like to thank you, anybody that has submitted an email or an application or something like that. I have already reviewed something like 60, and I'm going to be selecting five people to carry forward to the next step. Right now, that means that any one individual person has about an 8 % chance of getting a trial with me.
And then I'm probably going to whittle down the five trials I do to like, you know, maybe one or two people and work with them side by side. Which means, you know, collectively, there's about a 2 % chance that you would work with me if you're okay at this. I've had a lot of people reach out to me that have no experience doing any sort of thumbnails or any sort of titles or any sort of YouTube packaging before.
Maybe I should have made it clear, but that was one of my core prerequisites. So if you don't have experience with that stuff, unfortunately, probably just doesn't make sense. There were some people that didn't have experience with it and then tried just giving me it up front.
that's the right approach for a position like this. But you know, I'm looking for a very specific thing. And I saw most of these titles and thumbnails from total novices just did not have those core foundations in place.
So yeah, anyway, I still really wanted to thank you guys for sending that out. And anybody that I've, you know, selected to move forward. It's gonna be awesome working together, like really, really awesome.
I'm in the process of building out a team and doing really well. So super excited about that. The goal really is to just bump these numbers up.
I've realized like, it is virtually impossible for me not to win. um everything in the universe is just oriented my channel as like like the best channel i think the most authoritative resource on the stuff my issue is really just like i have dropped the ball on volume recently and you know i did so knowingly as an experiment i've only published one two three four five videos in the last month and the last video here was Not really like a video in the same, this is basically when I stopped my experiment.
So it was really just four videos in the last month. These were long form courses. I had a cloud code course over here, codex course over here, how to think clearly in the era of AI over here.
And then, yeah, I mean, I had like four courses that were really long form and just showcased my ability to think deeply and then provide a lot of value. And I love doing that. And it's awesome, but I'm unfortunately suffering a lot from browse features, which are pretty important on YouTube.
And by not having browse features, I've sort of shot myself in the foot. So I'm going to go back to optimizing for a mixture of both browse features and search features. I'm going to have like maybe one long form course a month, and then I'm going to have like, you know, probably 10, 20 or maybe 30 videos, depending on how frequently I can perform and on how well the ideation works out.
and i'm just gonna make sure those videos are as systematized but also as high quality as humanly possible start the ball rolling again you know if i think back to lifetime on my channel well not lifetime i guess like last two years probably i think two years yeah so i started like just a little bit before this um i think two one oops yeah there you go i actually started my first youtube video like back here right um i had this channel for years before but i didn't do anything with it so first youtube video way back here right um you can actually see like a massive growth spurt right around here where i got like a tremendous number of subscribers i got to the point where i was making like 2k a day and i did that for a long time and after this period i i think i like gained the channel size at the time i think i gained like 50k subs or something like that in like a month and i think my channel was like 80k to begin with so it was like a meteoric rise and ultimately you know what yielded me that growth was just like consistency over a long period of time that did not happen all at once what you don't see here is i was publishing like a video every single day every day every day every day
So I was working with a couple of really great cracked people on that. And that's why I was able to grow so much. And yeah, you know, I replicated that kind of over here with a bunch of other really good videos.
But it was nowhere near as I mean, this was 3k subscribers gained in a day back when I had like 50k subs, right? That's like basically an 8 % channel growth in a day. I've been unable to reach that even with like a 3 .5k sub day because my channel is just so much bigger now.
I want to get back to that. I want to get to the point where I'm growing like three to 5k subs a day. And I cannot continue like this mediocre, mediocre.
critique or mediocre growth around like 500 to a thousand. Um, because you know, AGI is here, ASI is coming. And I, I do want a real, a genuine shot at being like a core place where people go to get information about that stuff.
So that's really what I'm doing. And I'm going to make it happen. Going to make it happen.
Don't know if I'll be able to do the one mil subs before end of year. I also am going on a two week trip, literally in five days where I'm going to Hawaii and I'm just going to be with my family hanging out. Um, but I've decided I'm going to bring video equipment and I'm still going to continue shooting regardless, just because I want to make sure this is sorted.
And yeah, you know, just some reflections on some guys that I've spent time with here that run really big, successful media businesses. I've realized like, you know, I'm really good at media and I've been splitting my attention a lot between a variety of different things. Like I, I should probably double down on this media stuff.
Anywho. Yeah, that's what I'm doing. I've earned over 200 K in advertising revenue.
That's crazy. You know, what's so crazy is like, that is, I haven't spent that much money. in this whole period of time, meaning just on ad revenue alone, that's about 300k CAD, just on ad revenue alone, I'm basically positive over the last two years.
That's not entirely accurate, I guess I've probably had like some break swanky flights now. But yeah. So yeah, anyway, to that end, how are things going?
I have my main YouTube channel here, which we've just been watching my daily updates channel is going pretty well. We also have the stacked podcast, which is growing quite consistently to my issue with that, if I go to 24 hours, and I think I'm going to have to rearrange this, because I'd like to see main YouTube daily updates, stacked podcasts, and then probably main IG.
So I will have to rearrange that.
My issue is, despite the fact that all of these are quite linear, and they're going up, I don't want linear anymore. Check this out. All of these are basically like this, right?
That's not good enough, guys. Even this isn't good enough. this definitely isn't good enough this is not good enough i mean you know we got somewhere and now we're kind of like here what i want okay is i want this and i'm not getting that and the reason why is because despite the fact that content is inherently extremely leveraged it looks like i am receiving fixed returns for my time what this means is if you have a linear growth rate as a percentage of an asset that is itself growing, that means that the relative growth rate is actually going down.
And so if I had a channel that had 100 subs, and then I was getting one sub a day, I would be increasing my total subscriber count at a rate of 1%, right? But eventually I have 200 subs. Well, if I'm at 200 subs, and I add a sub a day, I'm no longer increasing by 1 % a day, I'm increasing by 0 .5%.
And if I had a thousand subs, I'd be increasing by 0 .01. Well, anyway, I have 520 ,000 subs. And with every increase in subscriber count, my relative growth rate has gone down.
I would like to have the same growth rate at minimum and ideally more growth rate. So the question is, how do I do that? And, you know, I'm in the process of figuring that out right now and hiring a team to assist me with this again.
And I will keep you all posted. Anywho. That's it for me.
Hopefully you guys appreciate it. A peek behind the curtain. I will see everybody here tomorrow.
Have a great rest of the day, guys.
The Hook
The bait, then the rug-pull.
Nick Saraev opens his daily-updates channel by stating the premise up front: he answers subscriber questions live, on camera, sorted straight out of his YouTube comments section. What follows is 20 minutes of unscripted problem-solving, from client delivery mechanics to a hard look at his own subscriber growth curve.
Frameworks
Named ideas worth stealing.
02:30list
The Three-Filter Niche Test
Digitally scrapable (decision-makers findable via website/directory/social)
High ticket (client's offer price justifies the outreach effort)
Low compliance (no heavy regulatory overhead like finance or healthcare)
A filter for deciding which niches are worth targeting for AI/automation services, used to explain why an MSP beats a dog walker as a client.
Steal forQualifying any B2B service or agency niche before building outreach infrastructure around it
09:50model
Guarantee Offer Formula
Structure every offer as 'I will get you [result] in [timeframe], or I will give you all your money back' rather than an unguaranteed flat fee, on the theory that the incentive structure determines effort level.
Steal forAny high-ticket service offer where the seller wants to force their own accountability
05:40concept
Cold Email Cutoff Rule
Send each copy variant to roughly 1,000 leads before judging it; a positive reply rate near 1% is strong enough to keep scaling, since historically 30-50% of positive replies convert to booked meetings.
Steal forDeciding when to kill or scale a cold outbound campaign variant
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
A 30-minute unscripted Q&A where a $300K/month consultancy owner answers subscriber questions on cold email, pricing, and AI models, then makes the case for staying a generalist early and niching down only once growth stops compounding.
Nick Saraev answers five subscriber questions and shows off an AI animation tool he's building, arguing AI has erased the old line between technical and non-technical founders.
A daily-updates vlog where an AI automation consultant answers subscriber questions live, argues boring blue-collar service businesses beat a white-collar hourly job, and closes by pulling up the actual spreadsheet behind his growth numbers.
A daily-vlog walk through comment Q&A, an unscripted business-building session, and the once-a-year discount math that adds $50,000 to a $300K-a-month business in a single day.
A live YouTube-comment Q&A where a $300K-a-month automation builder lays out the RACE framework for deciding which AI system a business should build first.
00:30
00:39
00:55
01:27
01:43
01:59
02:47
03:02
03:18
03:34
03:50
04:06
04:22
04:38
04:54
05:10
05:42
05:58
06:13
06:29
06:45
07:01
07:17
07:49
08:05
08:21
08:37
08:53
09:08
09:40
09:56
10:12
10:28
10:44
11:00
11:32
11:48
12:04
12:35
12:51
13:07
13:23
13:55
14:11
14:27
14:48
15:14
15:30
15:46
16:02
16:18
16:34
16:50
17:06
17:22
17:38
17:54
18:03
18:27
18:41
19:13
19:29
19:45
20:01
20:17
20:33
20:49
29:58
17:26
33:46
32:15
22:22
14:58