A Restaurant Full of People Who Aren't Hungry: A Conversation with Ademola Adelakun

Ademola Adelakun started as a professional magician, and that's not a detour from his marketing career, it's the root of it. Both are about attention psychology, making people lean in and lower their guard.

Ademola runs A2 Media, a research-led video engine for B2B SaaS that's produced hundreds of videos for companies like Shopify and Okta, videos built to drive pipeline, not just views. He started doing magic as a bullied kid looking for a way to break the ice instantly, and 24 years later that instinct, get someone to lower their guard first, shapes how he thinks about content that closes deals.

Going viral is a restaurant full of people who aren't hungry

Ademola's central argument is that the wrong metric quietly wrecks video strategies: virality. It feels amazing to rack up a million views, but ask most people what that did for their business and the honest answer is nothing. His analogy is the one that names this piece: going viral is like filling a restaurant full of people who aren't hungry. They come in, take photos, and leave without spending a dollar. Target correctly and fill it with people who actually want to eat, and that's where the real money is. The trap is that marketers walk into boardrooms with "we got 50,000 views," the board says "do more," and when someone finally tries to tie revenue to it, the disaster begins.

Average content that's direct and niche beats polished

Ask whether average content can win on distribution alone and Ademola reframes it around knowing your customer. If you niche down and understand your ideal customer as well as you understand a best friend, you know exactly how to position something to move them to action. Most companies don't know their customers that well; they assume steady sales mean the content works, until customer interviews reveal that what buyers actually want isn't in the same universe as the marketing. Content that's average-looking but really direct and niche wins every time, whether it's polished, ordinary, or "filmed with a potato."

The signal is lagging, and it takes 12-13 videos

Anyone expecting content to convert the first time is misreading the game. In B2B SaaS, his research puts it at roughly 12 to 13 pieces of content to convert someone (versus two or three in B2C, where the price is lower and fewer stakeholders are involved). The early indicators aren't sales, they're watch-through rate, shares, comments, and the kind of engagement where people ask questions or say "this was really helpful." If the response is "nice" or silence, you've probably made viral content that won't earn a dollar, or something nobody finished.

Buyers need to feel understood

What a B2B buyer needs to feel before they trust you is understood, that you see them. His example is the eerie TikTok video that opened with "if you're watching this, you're a Black guy in your mid-twenties who loves anime and eats at these restaurants," and made him put his phone down wondering how they knew. Do that for your buyer, describe their exact situation, you've talked to six vendors, you've tried X, and they think "this video was made for me." Then you show them how to fix the problem, and trust is built. Do it consistently and people decide you're who they want to work with.

Production quality can actively hurt you

The thing teams obsess over that has almost zero impact, and can even backfire, is production quality. A highly produced video now reads as "this was made to convince me to do something I don't want to do," so people resist it. Someone on their phone telling a genuine story tied to a point makes people think "yeah, I kind of want that." He's watched agencies that know nothing about a client's space shoot a slick $100,000 video that, once ad spend is behind it, turns out to have diminished trust and done the opposite of its goal. Often the better move is hiring micro-influencers who actually like the product. (A2 Media, he notes, doesn't do the behind-camera work itself, it partners with production companies.)

AI slop and the return of effort

Ademola isn't an AI naysayer, but he's disappointed by how it's used: companies treat content as a box to tick, and let AI check it, producing what everyone now calls slop. He expects diminishing returns fast, because people are already tired of guessing whether something is AI, the way you swipe away the second you realize something's an ad. The lower barrier to entry actually proved there's an art to this, and throwing videos out for their own sake desecrates it. Give a creative person AI tools and they make far better content, far more efficiently; let AI do it alone and you get garbage people tire of.

Where to start with zero videos

His concrete playbook: take all your sales transcripts, run them through an LLM, and pull out the ten most common questions, themes, and stories. Then make two things for each, a basic two-to-three-minute video answering the question, and a story version (a client you helped, or how you'd get someone to that transformation). Keep a pulse on what formats are working, and remember the goal is never to sell, it's to help, so that when someone hits that problem they go to the person who taught them everything about it. The basic answers matter for a specific reason: YouTube is starting to surface above Reddit as a source inside LLMs, so if you say on video what people actually ask on calls and in chatbots, you're more likely to get cited. And if you build for an audience of one, specific enough, you end up with an audience of thousands you can speak to as one.

Key takeaways

Viral is people who aren't hungry. Views feel good and rarely convert. Target the buyers who actually want what you sell.

Direct and niche beats polished. Knowing your customer like a best friend matters more than production value. Potato-quality can win.

Conversion is a lagging signal. It takes ~12-13 pieces in B2B. Read watch-through, shares, and questions, not first-touch sales.

Make them feel understood. Describe the buyer's exact situation and they'll believe the video was made for them, which is where trust starts.

Production quality can backfire. Slick reads as "being sold to." Authentic, phone-shot storytelling builds more trust than a $100k shoot.

Frameworks worth stealing

Mine your sales transcripts

Run your sales-call transcripts through an LLM and extract the ten most common questions and themes. For each, publish a plain two-to-three-minute answer and a story version. It's the fastest way to make content that's actually useful, it gets you cited inside AI answers, and the goal is to help, not sell, so the buyer comes to you when the problem is real.

Build for an audience of one

Write and shoot as if you're talking to a single, specific person whose situation you can describe down to a T. Counterintuitively, the more specific you are, the larger the audience you build, because thousands of people recognize themselves in that one description, and you can speak to all of them as one.

Quotes worth keeping

The lines I wrote down.

Going viral is like filling a restaurant full of people who aren't hungry. They take photos and leave.

Average content that is really direct and niche always wins, whether it's polished or filmed with a potato.

The goal should never be to sell. It should just be to help.

Rapid fire round

Same questions every guest. Quick questions, quick answers.

Best advice you've ever received? Take "failure" out of your vocabulary, you're just pivoting. This is his 14th business; he failed 10, and kept going.

Advice you ignored and wish you'd listened to? If you're thinking about firing someone, do it the first time, don't wait. It would have saved him a lot of pain.

What would you tell your younger self? He went on Dragons' Den around that time and it didn't go well, leaving him in a dark place for months. He'd say: this is just the beginning and your lesson, and people don't really care about what you do that much.

Ongoing challenge that keeps you up at night? The education piece, spending so much time teaching clients the wrong way to think about video, and hoping they'll come knowing the value and just work with him on execution.

Favorite spot? Alo in Toronto, a French experimental course-menu restaurant where he and his partner celebrate anniversaries.

Tool you can't live without? Claude. "I'm so sad to say it. It's so good."

---

Ademola Adelakun is the founder of A2 Media, a research-led video engine for B2B SaaS that has produced hundreds of videos for companies like Shopify and Okta. A professional magician of 24 years, he brings attention psychology to why most B2B video fails and what makes content actually drive revenue. Find him on LinkedIn.