The fear of falling behind is their best-selling product

Wednesday morning. A wrought-iron table outside on Corso Aldo Moro in Santa Maria Capua Vetere. The air smells of exhaust fumes mixed with that sharp, slightly sweet stench of burnt coffee from the bar on the corner. Berto's phone is resting next to his empty espresso cup, smeared with dried foam. On the screen, lit up by the pale nine o'clock sun, three sponsored ads scroll past one after another.

Some bloke with an overly groomed beard and a backdrop of fake palm trees explains how he's "automated his entire business using artificial intelligence working two hours a week". Scroll. A sped-up video with a synthetic voiceover promising to turn you into a "certified Prompt Engineer in 28 days". Scroll. A platform with a minimal logo promising access to "all the secrets of AI for business" for just ninety-nine cents.

Berto doesn't say a word. No comment, no cursing. But he stops. He sits perfectly still, staring at the screen for exactly twelve seconds. Then he puts the phone face-down on the table — but not too far away. He keeps it within reach, as if letting it go might mean missing the last train before everything falls apart.

Same morning, different time zone. Bobby is sitting on the worn, damp velvet seat of a Metrolink tram, suspended on the tracks between Salford and Manchester Piccadilly. Outside the window, the usual red bricks are wet from the fine drizzle, and the sky is the colour of concrete. Bobby opens LinkedIn. Three sponsored posts in a row hit him square between the eyes. The terminology here is different — cleaner, more corporate: upskill, AI-powered career, future-proof your workflow. The surface is polished, but the emotional lever underneath the bodywork is the exact same war machine. The exact same hook, buried in the exact same wound.

I see their expressions every time we meet up for a project. What Berto and Bobby are looking at isn't a training offer. They're not weighing up a curriculum, credits, or learning modules. They're looking at an advert for a cure to a deep, visceral anxiety — one the very people selling the course helped manufacture from scratch.

The moment everything changed — and why Berto feels guilty about it

Until a few years ago — let's say up to 2021 or the early months of 2022 — marketing for a small business owner was a game with fixed rules, almost reassuring in their predictability. Berto would put two hundred euros a month into Meta Ads for his window and door factory in Caserta, post three Instagram posts curated by his nephew or the local agency, and word of mouth did the rest. Bobby, at his sports goods shop in Salford, ran a couple of remarketing campaigns on Google and watched people come in with a discount code on their phone. It was an imperfect system, but a comprehensible one. Manageable.

Then the toy broke. Organic reach on the platforms collapsed — anyone who's managed campaigns between 2019 and today felt it in their own bones, no studies or percentages required. Cost-per-click tripled, algorithms stopped rewarding simply "showing up", big brands raised the barriers to entry with enormous budgets, and user attention became a scarcer resource than lithium.

Berto didn't feel guilty about that. That was just the market.

Then November 2022. ChatGPT.

Not AI itself — that had been around for years, in labs, academic papers, and big tech companies. But the moment AI became front-page news, pub conversation, the subject of your neighbour's comments. Every day, on every channel: "AI will replace your job within twelve months", "anyone who doesn't learn prompting today will be unemployed tomorrow", "the future belongs to whoever automates". A relentless bombardment, wall-to-wall, with no dissenting voices.

This created an emotional equation in Berto's head — and not just his. My commercial results are declining + everyone says AI is the definitive revolution = if I don't buy that ninety-nine-cent course, it means I'm old, lazy, and heading for the shutters.

But Berto isn't alone. Sitting at the same table, metaphorically speaking, is Sara. Sara is thirty-four, works as an administrative assistant in Naples, earns fourteen hundred euros a month, and for the past two years has been told her role is among the first at risk of automation. Sara sees the same ads as Berto, but with a different lever — not the fear of losing clients, but the fear of losing her job. Both of them are fertile ground. For different reasons, but equally real ones.

This feeling of inadequacy isn't stupidity. It's a rational response to concrete market signals, artificially amplified by those who have every interest in keeping it alive.

Who inhabits this ocean — and what they're really selling

If you decide to look for a solution in this market, you quickly realise the offering isn't all the same. There's a specific ecosystem that populates the sponsored posts in your feed. To navigate it, you need to learn to recognise the species first.

The keynote guru

You spot them immediately: they turn up at every industry event, they have a YouTube channel with fifty thousand views stuffed with videos bearing titles written in all-caps, and they've self-published a book on Amazon that they use as a business card with weight. This category doesn't sell practical tools — they sell the vision. They explain how the world will change in ten years' time, drop impressive names three times a minute, and use charts with arrows that always point upwards.

The entry-level course is just the hook. The real business model is continuous upselling: the exclusive community at eighty pounds a month, the masterclass at two thousand euros, the private consultation at five hundred euros an hour. They're not necessarily charlatans — some of them genuinely know their stuff. The problem is the enormous distance between their slides and Berto's reality, when what he needs is to figure out how to answer fifty furious client emails by this evening, not how the job market will look in 2035.

The guru sells inspiration packaged as methodology. Inspiration has its own value. But it's not worth two thousand euros a masterclass.

The microlearning platform with dark patterns

This is the industrial version of training. It presents itself as a super-polished mobile app, with five-minute lessons you can follow while waiting for the Tube at Southwark or queuing at the post office in Portici, colourful progress bars, and digital badges rewarding you every time you complete a module. The entry price is laughable: a three-day trial at €0.99.

But training, for these platforms, is merely an incidental cost. Their real product is the architecture of the payment funnel, designed to make you forget you've signed up to a recurring annual subscription that kicks in the moment that third day expires.

We'll get into the detail of this in a moment.

The certified digital trainer

Their natural habitat is LinkedIn. They have a headshot taken by a professional photographer, a polished company background, and a string of badges in the summary — often self-issued or awarded by bodies no one has ever heard of. They'll sell you the specific course at €497, "discounted this week only" from a completely made-up original price. To give you a sense of value, they pile a series of bonuses into the basket: ready-made prompt templates, lists of a hundred tools you'll never open, access to webinars recorded two years ago.

In the majority of cases, this professional takes the free official documentation from OpenAI or Anthropic, translates it into Italian, puts their logo on it, and resells you the convenience of not having to find it yourself. It's not theft — it's packaging. The problem is when the packaging costs five times what's inside.

The question that dismantles this model in ten seconds: does this professional actually use, day to day, in their real work with real clients, what they're teaching you? If the answer isn't verifiable, you already have what you need to know.

The scalable dream seller

The most aggressive variant. "Here's how I generated £8,247 in three weeks with this method — and now I'm going to teach it to you." Video featuring an expensive car, an apparently free lifestyle, precise numbers that sound credible precisely because they're so precise.

There's just one question that dismantles this category, and I've never heard anyone answer it satisfactorily: if the method really works that well, why do you need to sell it to me?

An entrepreneur with a genuinely replicable and profitable system has no financial incentive to teach you that system — they have every incentive to scale it themselves. The fact that they're selling it is already the answer. And it's the same answer that applies to Sara, who's looking for a way out of her situation: whoever promises her three thousand euros a month through a €97 course has found their way to make three thousand euros a month — by selling that course to her and five thousand other people in the same position.

How the business model you can't see in the ad actually works

Here we get to the mechanism. Not the intentions — the mechanics. Understanding how a system is built makes you immune to that system, regardless of who's running it.

The emotional funnel comes before the commercial one

Nobody buys an AI course cold. The purchase is the final act of a progressive exposure strategy that works on your weak points for days, sometimes weeks.

The first step is the awareness ad: it presents a real problem using the precise language of someone who's studied it closely. "Are your campaign costs going up while your results go down?" Berto recognises that line because it describes exactly what he's living through. It's not magic — it's research.

The second step is anxiety amplification: it introduces the external factor that transforms your problem into urgency. "While you wait, your competitors in Gragnano are using AI to halve their costs." For Bobby in Salford, it works the same way — with Manchester businesses as the bogeyman.

The third step is lowering the barrier: "No engineering degree required. Try our method for €0.99." By this point the psychological resistance has already been worn down by the previous steps. The decision to click feels almost logical.

Every step is tested through A/B testing on millions of users. You're not up against a creative copywriter. You're up against statistical data on what works on people with your demographic profile, your age, your area, your sector.

The trial that doesn't look like a subscription — and my test from the inside

In January 2026 I did something I'd recommend to anyone who genuinely wants to understand these mechanisms: I bought the trial of one of these microlearning platforms. Not to learn — to see the funnel from the inside, the way a forensic examiner walks a crime scene.

I spent €0.99. The interface was polished, the first lessons were quick and well put together. Then the upselling screens arrived — add-on packages, premium tools, extended access. In one of those steps, a button labelled "continue" — not "purchase", not "add to basket": continue — triggered a charge on PayPal. Three days later I had an active annual subscription at €79 that I had no memory of explicitly confirming.

It took me twenty-three minutes and four confirmation screens to cancel it. I kept the screenshots. They're still in a folder in my Dropbox, labelled "documented dark patterns — January 2026".

This is the mechanism known as a dark pattern — a deliberate design approach that uses user psychology against the user's own interests. The checkout page shows the trial price in large text, with a generous font and high contrast. The terms of automatic renewal — usually €59–97 annually — are written in light grey on a white background, at the bottom of the page, beneath three lines of legal small print and a dropdown menu you have to click to open.

Each of these elements, taken individually, is technically legal. The combination isn't — or at least, it shouldn't be. The EU Consumer Rights Directive requires pre-contractual transparency. But enforcing that right takes time, energy, and a bank willing to open a chargeback. The calculation these platforms run is precise: most users won't have the time or the inclination. They pay, grumble on social media, and then forget about it.

How to read reviews — and why the average score tells you nothing

The first red flag appears before you've read a single review. It's the number.

A hundred thousand reviews on Trustpilot. A hundred and twenty thousand. In some cases, more. Numbers that belong to the global retail giants — platforms with tens of millions of active customers, years of operation, a presence across dozens of countries. Not to a three-year-old edtech startup with a €79 annual subscription.

When a niche service accumulates a review volume that belongs to a multinational retailer, that disproportion is already a question worth asking. Not an accusation — a question. How is that possible? The answer almost always runs through mechanisms of active, systematic collection, not an organically enormous customer base.

"But that platform has 4.6 stars on Trustpilot with a hundred thousand reviews. How can it be a trap?"

Let me explain precisely.

It's called invitation flooding. The platform asks you to leave a review via an in-app notification exactly fifteen minutes after sign-up — at the precise moment when you've just entered your details, the interface has welcomed you with a confetti animation, and you've completed the first two-minute introductory lesson. At that moment your dopamine is high, you feel engaged, you think you've made the right call. You leave five stars.

The real problems arrive thirty days later, when you discover the recurring charge. At that point you go and leave one star, furious. But your negative review is literally swamped by the daily flood of five-star reviews generated from new trial sign-ups. The average score stays high. The average is manufactured, not measured.

The correct way to read a platform is this: filter directly for one-star reviews. Read the first ten. If five hundred people across Manchester and Caserta are all complaining about the same thing — "they charged me without warning and customer service doesn't respond" — we're not looking at unlucky customers. We're looking at a business model that has built that failure mode into its revenue projections.

The third signal is how the company responds to negative reviews. "We're sorry to hear about your experience, please contact support" on every one-star review isn't customer care — it's reputation management. A response that actually addresses the specific problem is an entirely different signal.

For documented complaints in Italy, Consumatori.it and the AGCM — the Autorità Garante della Concorrenza e del Mercato — are the correct channels for unfair commercial practices. Not Facebook groups, not Reddit threads: those official channels, with your full name and documentation.

The gym membership metaphor — and why it's not Berto's fault

The AI training market runs on a placebo for the conscience. Buying the course serves the same psychological function as joining a gym on 2nd January after you've eaten everything in sight over the Christmas period.

The simple act of swiping the card instantly softens the guilt and quiets the anxiety. It makes you feel as though you've already solved the problem, even if you never open the first lesson. The platforms know this perfectly well. They know that a large proportion of subscribers won't get beyond the first modules. Their system is optimised for people who pay, don't bother support, and keep financing the acquisition campaigns.

For Sara, the mechanics are even sharper. A €0.99 subscription that becomes €79 a year on a salary of fourteen hundred euros is not a trivial sum. And the guilt — I should have read the terms more carefully — is the second trap, the emotional one, that stops her from going ahead with the chargeback.

It's not their fault. It's architecture.

How to find your feet: what you actually need

Here's the part I care about most, because an article that dismantles the problem without giving you the tools to move forward is incomplete.

Four questions to ask before you buy

Not a checklist to pin on the wall. A way of thinking that takes less than a minute to apply.

Who is this person or platform, and what have they actually built with what they're teaching?

Not what they claim — what they've built. If they're selling a course on scaling an e-commerce business, do they have an e-commerce business they've scaled? If they teach Meta Ads, are they managing campaigns for real clients with real budgets? Search their name on Google: not the testimonials on their own site — those are curated. Look for external traces, verifiable profiles, work documented by third parties.

How do you cancel, and how many clicks does it take?

This question, asked before you pay for anything, is worth more than any other. If you can't find the answer in under a minute on the site, you already have your answer about the business model. A serious platform has no incentive to hide the exit procedure.

Do the negative reviews all point to the same problem?

Filter for one star on Trustpilot, read the first ten. If they're all different — someone disappointed by the quality, someone by the support, someone by the timing — you're within the normal range for any service. If they all describe the same problem, that problem is structural.

Is the promise measurable or just atmospheric?

"Transform your career" is atmosphere. "Learn to reduce client response times by 40%" is a promise that requires evidence. Anyone making measurable promises without backing them up with data is already less credible than someone who doesn't make them at all. Anyone making only atmospheric promises isn't selling training — they're selling an emotion.

AI for your business: what actually works today, no courses required

Berto doesn't need a €500 course to start using AI in a genuinely useful way. Neither does Bobby. Neither does Sara.

For copy, emails, social posts, quotes

ChatGPT on the free plan, Anthropic's Claude, Google's Gemini. All free on their base plans, all with official documentation in Italian, all accessible from a browser without installing anything. An afternoon spent experimenting on a real problem from your actual working day teaches you more than three hours of video course. This isn't an opinion: it's how adult learning works. It's called learning by doing, and it doesn't require a subscription.

For images and visuals

Canva has integrated AI tools directly into the interface Berto is probably already using. Adobe Firefly is included in Creative Cloud plans. Freepik has become Magnific — complete rebrand, if you visit the site now you're redirected straight there — and has integrated AI generation, upscaling, and editing into the same platform people already used for downloading stock images. Three tools you're probably already paying for, or have always used for free, that now do things which two years ago required professional software costing hundreds of pounds. You don't need to learn something new — you need to open what you already have.

For automating repetitive processes

Zapier on the free plan connects the apps you already use — CRM, email, spreadsheets, notifications — without any coding knowledge. n8n is more powerful and flexible: it has a cloud version starting at €20/month, or it can be installed for free on your own server if you have a minimum of technical know-how. For Berto, Zapier is the right starting point. n8n is the next step, when — and if — the need grows.

For data analysis and campaigns

AI functions are already built into Google Analytics 4, Meta Business Suite, and even Excel and Google Sheets. Before buying a new tool, explore what you've already paid for.

The real competitive advantage in AI isn't technical expertise — that can be picked up in a matter of weeks, and the tools change so quickly that what you learn today is partly obsolete in six months. The real advantage is knowing which problem in your business you're trying to solve. That answer doesn't come from a course. It comes from the experience built up over years of real work. AI is an amplifier — apply it to a clear process and it makes it faster; apply it to a confused process and it generates chaos at industrial speed.

Time — the variable nobody mentions

There's a hidden cost that trainers always leave off their sales pages: your time.

A thirty-hour course that Berto will never finish because he has deliveries to manage in Portici has a return on investment of precisely zero, regardless of what the entry ticket cost. An hour spent getting things wrong on your own — testing a free tool on a tiny problem from your daily routine — is worth far more than twenty hours of video of someone else demoing features using made-up data in a controlled lab setting.

The time you dedicate to your business is your scarcest resource. Don't waste it watching other people do demos.

If you've already paid

It happens. It's not your fault — these funnels are built by professional teams with years of testing behind them and significant acquisition budgets.

First thing: check all your active subscriptions. On PayPal, go to Settings → Payments → Manage Automatic Payments. On your credit card, check the last three months of statements looking for recurring charges of a fixed amount. Plenty of forgotten subscriptions turn up this way.

If you're entitled to a refund under the terms of service and you're being refused, the route is a chargeback through your bank. It's not a complicated process — ask your bank directly. For documented unfair commercial practices in Italy, the AGCM accepts online complaints at agcm.it. Consumatori.it collects case records and provides support with procedures. In the UK, the Competition and Markets Authority performs the same function.

A note on GDPR that very few people make use of: you have the right to request the deletion of all your personal data from the platform's servers. Many of these platforms are registered outside the European Union, but if they operate with European customers they're still subject to the regulation. A formal deletion request, sent by email with read receipt, creates a legal obligation that isn't dependent on where the company is incorporated. Simply mentioning the Data Protection Authority often unblocks customer service teams that seemed otherwise impossible to reach.

Conclusion

Berto gets up from the table on Corso Aldo Moro. He pays for his coffee at the till, nods goodbye to the barista, and gets into his car to drive to the factory. His phone is in his pocket, silent. The ad with the bloke and the fake palm trees is still out there in the digital ether, ready to hit someone else — but today, Berto has stopped looking at it. The difference between the Berto of this morning and the one from a week ago isn't that he now has a degree in computer science. It's that he understands how the cage is built. And when you can see the bolts on the trap, the promise instantly loses every last bit of its power.

In Salford, Bobby steps off the tram at Manchester Piccadilly. There's a cold wind that cuts right through you as you leave the station. He closes LinkedIn, puts his earphones back in his jacket pocket, and walks towards the shop. He hasn't stopped wanting to improve his work, hasn't stopped wanting to understand how technology can help him. But he's understood that serious learning never arrives inside a box with a red countdown, a crossed-out price, and a "buy now" button expiring in two minutes.

The artificial intelligence market is real. The skills to use it are real. The urgency not to sleep through a changing world is legitimate. But there's an enormous difference between deciding to learn a tool in order to apply it to your work, and buying the instant gratification of a one-euro subscription to make yourself feel better. The first path takes time, effort, mistakes, and patience in front of a blank screen. The second costs ninety-nine cents. And then seventy-nine euros a year, until you notice it on a rainy day when you're going through your bank statement.

Proper training doesn't need to make you feel guilty in order to sell you a solution.