#160 · The best AI agency services to offer in 2025

youtube ↗Agency Operations

What tools are you using in your own business? Only ClickUp or something like Superbase? No, only ClickUp. I don't use um a database like Superbase. And where are you saving all your data? What's your workflow?

I use ClickUp as my main tool to store everything; it's visual and sufficient for my needs, so I don't need a real database like Superbase. I'll leave it there for now. Moving on to brand stats: my main channel has about 159,960 subscribers, Instagram around 26,648 followers, and the daily updates channel (where I answer questions) has about 8,844 subscribers. Subscriber growth is down to roughly 0.36% per day, Instagram growth around 2.62% per day, while the daily updates channel growth remains strong. Recently I posted about winning the school games, cross‑posted it to the schooler subreddit, and asked Sam Ovens and others to pin it to gain visibility and raise the standard in the school community. The post is long and was pinned, receiving comments from notable figures like Sam Ovens, Nate Herk, Spencer (who previously held the top spot), and Mr. Hormosi. I plan to meet them in the next couple of days. Although there's no direct ROI, I expect future returns from the connections and positive clout. On the sales front, I had to take down a posted sales call for Leftclick after the other party raised liability concerns; the first call already exposed potential issues, making me doubtful about continuing such posts. I’ve considered alternatives like sales coaching, role‑play sessions, or a podcast‑style format where I work with others and document the process, but these versions lack the perceived authenticity of a live, unfiltered sales call. Lead generation is doing well: Instantly launched an AI inbox manager, and the podcast campaign now yields a 2% reply rate, with about 1% positive replies. At my recommended volume of 270 emails per day, that translates to roughly one positive reply per day, or 20‑30 per month; I’m sending about ten times that volume, so I could scale up to ten calls per day if market capacity allows. I’ve also thought about simply giving people leads, but that raises logistical questions about booking calls and ensuring attendance without double‑booking. I’m open to any sensible solutions you might have. Finally, my channel’s relative growth rate has been declining, and I was about to publish a video intended to reverse that trend, though it hasn’t uploaded yet due to timing.

ClickUptools workflowbrand stats

Related answers

youtube ↗Lead Generation

Your school data consulting model is solid, but the main lever is getting more eyes on your outreach. With cold DM, only about 20% of recipients actually see your message, and only a tiny fraction respond or convert. To dramatically improve results, you should multiply your touchpoints—send the same prospect a DM on LinkedIn, Instagram, X (Twitter), and email, for example. Because each platform gives roughly an independent 20% chance of being seen, using four channels raises the overall visibility to about 60%, effectively tripling your potential revenue without changing your offer. Beyond that, the real bottleneck is top‑of‑funnel visibility, so invest in content that attracts a broad audience (e.g., YouTube videos that aren’t just ‘how to be motivated’ but showcase specific processes like AI agent workflows or $20k/month side‑hustles). Continuously test different angles and formats to see what drives the most views, then double down on the winners. Remember, for most small‑to‑mid‑sized businesses, sales and marketing—the ability to get seen and start conversations—are the biggest constraints, so treat lead generation as your primary focus.

Your school data consulting model is solid, but the main lever is getting more eyes on your outreach. With cold DM, only about 20% of recipients actually see your message, and only a tiny fraction respond or convert. To dramatically improve results, you should multiply your touchpoints—send the same prospect a DM on LinkedIn, Instagram, X (Twitter), and email, for example. Because each platform gives roughly an independent 20% chance of being seen, using four channels raises the overall visibility to about 60%, effectively tripling your potential revenue without changing your offer. Beyond that, the real bottleneck is top‑of‑funnel visibility, so invest in content that attracts a broad audience (e.g., YouTube videos that aren’t just ‘how to be motivated’ but showcase specific processes like AI agent workflows or $20k/month side‑hustles). Continuously test different angles and formats to see what drives the most views, then double down on the winners. Remember, for most small‑to‑mid‑sized businesses, sales and marketing—the ability to get seen and start conversations—are the biggest constraints, so treat lead generation as your primary focus.

cold dmmulti platformtop of funnellead gen
youtube ↗Other

Why do you manually track metrics daily, and what benefit do you get from it?

It's a good question. I'm the automation guy, yet I manually update a janky Google Sheet each day because the value isn't just the metric—it's the awareness and reflection it brings. Spending about 30 seconds daily yields far more value: I notice trends and spikes in follower growth that make me investigate causes. If you had many metrics and worked with a company, you wouldn't do this manually every day—it would take too much time—but as a small portion it's fine. I had meant to mention Mr. Beast because we're collaborating. I recently hung out with school friends Marty, Ryan, and Michael; Marty has over 5 million YouTube subscribers, Ryan about 160 k Instagram followers, Michael many millions on Instagram. It was cool to watch the local cactus squad and discuss optimizing school as part of a big challenge. School is going well; I'm in the platinum group for educational creators with 100 k+ followers, where we help each other grow. I comment on posts for five minutes daily and have reached number two on the leaderboard, meaning I'm a valued resource; people constantly ask for my opinion. I achieved this with little effort—just voice‑transcribing posts and noting my thoughts (X, Y, Z) in 15–30 seconds per post—yet I now get massive exposure. I love giving value and seeing how communities make interaction easy and human‑focused. The free school community is thriving at 7,400 members, and we've celebrated two members hitting 10 k‑month milestones. Each day I review new posts, quickly scanning for obviously AI‑generated content (e.g., a YouTube‑video‑style post) and delete it, then ask if it's truly a YouTube video; if so, I delete it. I must be trigger‑happy for time, but the rest of the community is great: I spot AI‑generated posts, remove them, and spend about five minutes leaving a few comments. New members often feel uncertain; seeing that success is possible encourages them. The no‑AI‑post rule has raised discussion quality. A recent post from Rodrigo looked suspicious due to odd M‑dashes, but he noted it was a Google translation, so language may explain it. I appreciate Nick's style—no hype, straight talk that works. Regarding a journey‑maker school example, I noted it was obviously AI, less tinkering, more doing, and deleted it. I can give a reason for deletion, but my rule is two strikes: first AI‑post gets benefit of doubt, second results in removal. I've become familiar with AI writing patterns, so I ban them aggressively, keeping the community high‑quality. I also sprinkle in posts as needed, though the top‑level look lets people market low‑value offers in the intro tab (e.g., 'Hey, how's it going? Do you want to make money online?'), which I only see when reported; then I ban the offender. I enjoy the community and will defend it with my Warhammer 40K sword metaphor. Direct messages have become a burden—many users just say thanks—so I read them but don't reply to stay true to my promise. Reading 40–50 DMs daily makes me reflect on how far one person can go; I might be the only one sustaining this volume for over a year. I use AI everywhere except for the actual coaching deliverable and my personal presence, which I haven't replaced, preventing the perception of being fake. It's an interesting thought experiment: could I scale the community to a million members a month while still managing it alone? Possibly, but we'll see. My stats show slowed growth (0.11 %), then a rise to 0.66 % after a Maker Zero boost to 7,500 members. I plan to add an email‑address question to Maker Zero to capture all member emails for promotions, accepting some leakage as inevitable. I'm aiming to return to the top three rank again.

manual trackinggoogle sheetmetricscommunity
youtube ↗Cold Email

I’m offering a missed‑call recovery AI dispatcher for towing businesses and am hitting Facebook DM limits with cold outreach. Should I change niches or keep going?

If you're using Facebook for this, it's essentially game over--Facebook DMs are extremely difficult, limited to about 10 messages per day per account, and you'll quickly end up in message requests. By contrast, a cold email account lets you send 25-30 emails daily and costs only $2-$4 per month, whereas a Facebook account runs around $50 per month due to verification overhead. Therefore, you need an intermediate step to obtain their email addresses, then conduct cold email outreach. Calling is also effective, especially for towing businesses that are often on the road; you can't reliably email auto shops, so walking in or calling works. This approach isn't perfect, but it's workable. Regarding my own situation: I've realized I need to set up ads. Yesterday I hung out with Jackie Chu in Vancouver, who's now ranking for businessman and is a brilliant guy--I plan to send him a message. I also met Tom Wang, who's really cool, and Tony, founder of Bessie. Their advice was to run retargeting ads from Maker School; they said not doing so would be idiotic. I agree and will implement that, with help from my team and the school. My strategy is to list the problems my community faces, pick five or ten, and record a video addressing each. The videos will reiterate that Maker School is 100% free if you don't land your first client within 90 days, with a full refund. We also have $20,000 in discounts; most people use about 10%, giving roughly $2,000 off, which means a new member gets an immediate 10x ROI in the first month before even getting a client--so there's no reason not to join. I'll track video completion percentages to gather more data. My offer is strong and the product quality is high, so I'll put it in front of as many people as possible. I'm also excited that a friend is onboarding people for Claro, and I'll meet school guys later today for a coffee-shop flyover activity and other fun. Overall, I'm amped up about everything happening, and I believe ads could unlock insane growth if executed well. As for my channel stats: my main YouTube channel currently shows 471,066 subscribers, which is abysmal growth; other metrics indicate decent growth elsewhere--my paid offer has 7,234 sales, my Maker School rank is still two, and we're making good money, though subscriber growth isn't what I expected because I'm not creating viral head-type videos. I'll keep growing at a steady clip and hope to reach 500k subscribers by year-end.

cold emailFacebook limitstowing
youtube ↗Lead Generation

I want to start a service business for US mobile mechanics but have no capital—just a laptop and time. What direction should I take?

If you're using Facebook for this, it's essentially game over--Facebook DMs are extremely difficult, limited to about 10 messages per day per account, and you'll quickly end up in message requests. By contrast, a cold email account lets you send 25-30 emails daily and costs only $2-$4 per month, whereas a Facebook account runs around $50 per month due to verification overhead. Therefore, you need an intermediate step to obtain their email addresses, then conduct cold email outreach. Calling is also effective, especially for towing businesses that are often on the road; you can't reliably email auto shops, so walking in or calling works. This approach isn't perfect, but it's workable. Regarding my own situation: I've realized I need to set up ads. Yesterday I hung out with Jackie Chu in Vancouver, who's now ranking for businessman and is a brilliant guy--I plan to send him a message. I also met Tom Wang, who's really cool, and Tony, founder of Bessie. Their advice was to run retargeting ads from Maker School; they said not doing so would be idiotic. I agree and will implement that, with help from my team and the school. My strategy is to list the problems my community faces, pick five or ten, and record a video addressing each. The videos will reiterate that Maker School is 100% free if you don't land your first client within 90 days, with a full refund. We also have $20,000 in discounts; most people use about 10%, giving roughly $2,000 off, which means a new member gets an immediate 10x ROI in the first month before even getting a client--so there's no reason not to join. I'll track video completion percentages to gather more data. My offer is strong and the product quality is high, so I'll put it in front of as many people as possible. I'm also excited that a friend is onboarding people for Claro, and I'll meet school guys later today for a coffee-shop flyover activity and other fun. Overall, I'm amped up about everything happening, and I believe ads could unlock insane growth if executed well. As for my channel stats: my main YouTube channel currently shows 471,066 subscribers, which is abysmal growth; other metrics indicate decent growth elsewhere--my paid offer has 7,234 sales, my Maker School rank is still two, and we're making good money, though subscriber growth isn't what I expected because I'm not creating viral head-type videos. I'll keep growing at a steady clip and hope to reach 500k subscribers by year-end.

lead genoutreachservice business