What was the biggest lesson you learned in your hardest AI project on your AI journey?

#24 · $139,000 in February 2025

What was the biggest lesson you learned in your hardest AI project on your AI journey?

The biggest lesson I learned is that the actual product or service you sell doesn’t matter much—AI automation is just a fulfillment mechanism. What really counts are the core business systems you build around it: lead handling, sales scripts, onboarding, admin, hiring, marketing, taxes, etc. Those pieces stay the same across industries; only the specific service you deliver changes. Once you have a solid business foundation, you can swap in any offering and still succeed. I’ve seen the same principles work when I sold videography, local SEO services, courses, and even events—so it’s less an ‘AI journey’ and more a general business journey.

aibusiness fundamentalslesson

Related answers

youtube ↗Niches

I'm an 18‑year‑old full‑stack web developer with API and marketing theory knowledge. I'm unsure whether to start a marketing agency or an AI automation agency, especially since I'm in Saudi Arabia where few talk about this. Should I focus on building my personal brand as an AI expert first, given I have four to five months before university starts? What would you do in my position?

If I were you and thinking long‑term, I’d ask where everything is heading. Everything is moving toward AI and automation over the next five years; what happens after that is uncertain, but my honest advice is to choose the AI automation business model. I’ve found the most success with it and have run it four times. The main reason is leverage: AI and automation offer far more leverage than traditional marketing agencies. Traditional marketing agencies are declining—this model peaked around 2017‑2019 and has been falling since, partly due to market saturation and the rise of SaaS tools that automate tasks like scheduled posting, brand strategy, and marketing strategy. In contrast, the AI automation model is still early; it really took off around 2021‑2022 with ChatGPT, and I predict it will continue growing until about 2027‑2028, when AI agents could double the market and demand. After that, demand may dip slightly as agents become more advanced. If you're in Saudi Arabia, you're essentially seeing a delayed version of this curve, giving you a few extra years to ride the upward trend. So, if I were in your place, I’d go all‑in on AI automation.

business modelai automationmarketing agency

youtube ↗AI & Automation

Nick, I'm considering offering or selling AI to businesses. I want to start to some small businesses before aiming for bigger clients. How would you tackle this given they have different needs and situations or problems? Do you build a setup per business or do we build a set of services that could work like a general AI service? Also, for approach, direct messaging or in person.

Don't overthink the exact method—pick a few approaches and test them. Choose three different ways of reaching people, set a minimum daily outreach for each, and stick with it for a month or two (30‑60 days). See what works, then double down.

ai serviceoutreachtesting

youtube ↗AI & Automation

Is selling broad AI implementation tailored to each business valuable, or should I niche down into specific solutions like marketing and use AI within those?

If you’re making $0‑$20 k per month, keep selling broad AI implementations and use AI to create the perception of a highly customized service. Once you want to go beyond that revenue level, you need to productize and template your offering; that part of the market hasn’t changed even as AI models have advanced.

ai implementationnichingbroad aispecific solutions

youtube ↗AI & Automation

What do you recommend for me to start? I feel lost and can't find a good automation to sell; I was thinking about an inbound calling agent for a real‑estate or dentist firm.

Yeah, listen—feeling lost is completely okay. It’s fine to feel lost; the reality is you just have to make a move anyway. In modern entrepreneurship you basically have to lower your bar for what counts as “confident enough to succeed” and then just go for it. Decades ago, when starting a business required a lot more capital, you needed to be far more confident about ROI. Nowadays you can launch an AI automation agency with a couple hundred bucks, as I’ve shown multiple times, so you don’t need to be all that confident. The most important thing is just to get going. The idea that you can’t find a good AI automation to sell isn’t true; it’s that your standard for what counts as a “good” automation is just too high. Use the blueprints I’ve shown you, figure out the simplest three‑ or four‑module system you can, and start selling. Your biggest obstacle is yourself, not the niche or the specific firms you’re targeting. An inbound calling agent for real‑estate or dental offices could be great—just put something together, start talking to real customers with it, and you’ll quickly see how to improve the system into the high‑value solution you’re envisioning.

inbound callingreal estatedentistautomation