The short version: the most effective way to make money online right now is building a narrow, paid offer for a small owned audience and using AI to cut the production time down from weeks to days, not chasing whichever platform trend is loudest this month. Everything else, the dropshipping courses, the “passive income” bundles, the affiliate funnels, is either a slower version of that or a way for someone else to make money off you.
The old advice is dead, but not for the reason you’ve been told
For years the standard “make money online” playbook was: build an audience, monetise with ads or affiliate links, wait. That still works, technically, but it’s slow and the platforms have made it slower on purpose. Organic reach on Facebook pages sits under 2%. Instagram now shows a huge chunk of feeds as recommended content from accounts you don’t follow, which is great for the platform and terrible for anyone trying to build a following that converts. YouTube ad revenue for a general channel runs somewhere between two and ten dollars per thousand views, which means you need real scale, hundreds of thousands of monthly views, before it pays a mortgage.
What changed in the last eighteen months isn’t the business model. It’s the production cost. AI tools have made the “build a paid offer for a specific group of people” model dramatically faster to execute, which is what makes it the effective option now, not because the strategy is new but because you can build and test it in days instead of months. I’ve written before about how to use AI to build a successful online business, and the short version is this: the winners in 2026 aren’t the people with the biggest audience, they’re the people who went from idea to paid offer fastest and then adjusted based on what sold.
What I built, and what it made
After five rough years rebuilding my business from scratch, I stopped trying to grow a bigger audience and started testing small paid offers instead. The one that worked: a three-hour live AI workshop for small business owners, priced at 97 pounds, built entirely from a slide deck I drafted with ChatGPT in an afternoon and refined over two evenings. I sent one email to a list of just over 8,000 people and posted about it four times on LinkedIn across ten days. No ad spend. Forty-three people bought a seat in the first run. That’s roughly 4,170 pounds from one email list and a handful of posts, for a product that took under a week to build and cost nothing to deliver beyond my own time.
The point isn’t the number, yours might be smaller or bigger depending on your list and your niche. The point is the mechanism: a specific paid thing, sold to people who already trust you, built fast because AI did the heavy first-draft work. That’s the model working right now, and it’s the same model behind most of the profitable online income streams I cover in my rundown of online income channels for 2026.
The step by step version
If you want to test this yourself, here’s the process, roughly as I run it with clients:
- Pick one problem you can solve in under three hours of someone’s time, not a whole transformation, one specific problem
- Draft the outline with an AI tool in one sitting, then cut it in half, most first drafts are too long
- Price it between 47 and 197 pounds if it’s a workshop, guide, or template pack, that range sells without needing a sales call
- Build it on a simple platform, Gumroad takes a flat 10% fee and needs zero setup, Kit’s free plan covers lists up to 10,000 subscribers if you don’t have one yet
- Sell it to your existing list and network first, three to five posts and one direct email, before spending a penny on ads
- Run it once, watch where people got confused or dropped off, fix that, then run it again at a slightly higher price
Most people skip step six. They build the thing once, it does okay, and they move on to a new idea instead of doubling down on the one that already made money. That’s the actual difference between someone earning a few hundred pounds a month from this and someone earning a few thousand.
The real numbers on the platforms people recommend
Before you pick a platform, know what it costs you:
- Etsy charges a 6.5% transaction fee plus 20 pence per listing, on top of payment processing
- Fiverr takes a flat 20% commission from every freelance sale, no matter the size
- Upwork charges freelancers a flat 10% service fee and clients a separate marketplace fee
- Amazon KDP pays a 70% royalty on ebooks priced between 2.99 and 9.99 pounds, and drops to 35% outside that band
- Substack takes a 10% cut of paid subscriptions, on top of Stripe’s processing fee
- TikTok Shop commission for affiliates ranges roughly 2% to 20% depending on the product category
None of these are bad platforms. But every one of them takes a cut before you see a penny, which is exactly why owning your own list and selling directly, even through a simple checkout page, keeps more of the money you earn.
The bit most “make money online” content skips
Here’s the uncomfortable part. A huge amount of the content telling you how to make money online is written by people whose actual income comes from teaching you how to make money online, not from the method itself. The course about dropshipping, the ebook about affiliate marketing, the “I made six figures online” video, in a lot of cases the six figures came from selling that exact video to people like you. Survey after survey of online creators finds the median income from content and digital products sits shockingly low, often under a few hundred pounds a year, while a tiny percentage at the top skew the averages upward. That doesn’t mean the model is fake. It means most people either pick a problem too broad to sell into, or they stop after one attempt instead of improving the offer.
The way around it is boring: sell something specific, to a specific group, more than once, and pay attention to what they say when they buy or don’t buy. That’s marketing, not magic, and it’s the same principle I keep coming back to in why most businesses treat social media as marketing backwards. Posting content and hoping it converts is the slow, unreliable version. Building a specific offer and using content to point at it is the fast, reliable one.
Where this goes wrong
Three mistakes I see constantly with small businesses trying to do this:
- Building the product before checking anyone wants it, ask five real people first, don’t guess
- Pricing too low out of nerves, a 17 pound product and a 97 pound product take almost the same amount of your time to sell
- Trying to compete with huge brands on volume and ad spend instead of on specificity, which is exactly the trap I address in how to effectively compete as an SMB, small businesses win by being narrow and fast, not by out-shouting bigger competitors
If your business skills feel rusty around pricing, positioning, or basic marketing, it’s worth an afternoon on one of the courses listed in my piece on online learning platforms to improve your business skills before you build anything. It’s a cheaper mistake to fix on a course than on a launch that flops.
If you’d rather have someone build it with you
Not everyone wants to figure this out alone at ten at night with three tabs of AI tools open, and that’s fair, I did it that way for a long time before I stopped. If you want someone to sit with your specific business and build the offer, the pricing, and the AI workflow with you rather than around you, that’s literally what a hands-on AI implementation coach does, and it tends to shave months off the trial and error stage.
What to do this week
Pick the one problem you get asked about most often, by clients, by strangers on LinkedIn, by your own team. Draft an outline for a paid solution to it this week, not next month. Price it rather than cheaply. Tell your list and your network it exists. That’s the entire strategy. The tools around it will keep changing, they always do, but the mechanism, a specific paid offer sold fast to people who already trust you, is the closest thing to a reliable answer this question has right now.
Related reading: app money earning.
Frequently asked questions
What is the fastest way to make money online right now?
Selling a specific paid product or service directly to a small existing audience, an email list, a LinkedIn following, past clients, is the fastest reliable route, because you skip the slow work of building reach from zero and skip a chunk of the platform fees that come with marketplaces.
Do I need a big following to make money online in 2026?
No. A list of a few hundred people who trust you and open your emails will outsell a following of tens of thousands who scroll past your posts. Depth of trust matters more than size of audience.
Is AI replacing the need to build a real offer?
No, AI just speeds up the building of it. It writes first drafts of workshops, guides, emails, and product outlines fast, but it doesn’t decide what problem is worth solving or price the thing correctly, that part is still on you.
How much can a beginner reasonably expect to make with this approach?
It depends entirely on list size and trust, but a first paid offer priced between 47 and 197 pounds, sold to even a modest list of a few thousand people, commonly brings in somewhere between a few hundred and a few thousand pounds on the first run, with the second run usually performing better once you’ve fixed what didn’t work.
