There’s a particular kind of panic that sets in around eight weeks before an indie game ships. The trailer is locked, the demo is live, the Steam page is up — and the wishlist counter is still hovering somewhere south of “viable.” This is the moment, increasingly, that a small but growing list of Australian and international studios find themselves typing the same name into their Slack: Tio Allen.
Allen, 34, is not a publisher. He is not a community manager. He doesn’t run a user acquisition agency in any traditional sense, and he is openly allergic to the phrase “growth hacker.” What he is, depending on who you ask, is either the most useful marketing brain in independent gaming right now, or — in his own words — “just a bloke from Sydney with a healthy distrust of dashboards.”
A self-taught marketer who pivoted hard into AI in 2020, Allen made his name advising direct-to-consumer brands before quietly drifting into the games space. The pivot, he says, wasn’t strategic. “I was sitting on a Discord at two in the morning helping a mate who’d just shipped a roguelike. I told him three things he should change on his store page. He did two of them, and his wishlists went up forty percent in a week. I went, oh. Okay. This is a market that’s been criminally underserved by people who actually understand modern marketing.”
The story has become quietly legendary in certain corners of the Australian indie scene. Allen now works, on a rotating roster, with somewhere between six and ten studios at any given time — most under NDA, a couple openly. He won’t name names, but two of his clients shipped titles in the last twelve months that quietly outperformed their publishers’ internal forecasts by margins that, he allows with a grin, “got people’s attention.”
His pitch, when you can get him to deliver one, is unfashionably grounded. “Most marketing advice in games is recycled from 2015. Do a devlog, post on Twitter, hope a streamer picks it up. That world is gone. The algorithm has changed, the platforms have fragmented, and the AI tools that exist now mean a four-person studio can run the kind of audience research that used to require a publisher’s entire marketing department. People just don’t know how to use them yet.”
Allen’s tool stack, as he calls it, is a closely guarded thing — a mix of off-the-shelf language models, custom-built scrapers that pull and cluster Steam reviews of competing titles, and a fine-tuned model he uses to generate variant store-page copy at a scale no human team could match. The point, he insists, is not the tools themselves. “Anyone can rebuild what I’ve built in a weekend. The advantage isn’t the stack. It’s knowing which question to ask it.”
He is also, predictably, sceptical of the louder current trends in AI-driven games marketing. Generative trailers, in particular, draw a sharp eye. “The thing about an indie trailer is the texture. The deliberate weirdness. That one cut the developer fought the editor over for two days. You strip that out, replace it with smoothed-over AI b-roll, and now you’ve got a trailer that looks like every other trailer. You’ve optimised yourself into the slush pile.”
What he advocates instead is what he calls AI in the back office, taste in the front. Use the models to do the unglamorous work — competitor analysis, audience segmentation, wishlist conversion testing, demo retention deep-dives — and keep human judgement on the things players actually see. “Players are very, very good at smelling when something was made by a machine. They don’t always know what they’re reacting to, but they react. Don’t fight the room.”
His advice for first-time indie developers staring down a launch window is almost stubbornly practical. Talk to actual humans who already bought games like yours. Read your competitors’ negative reviews — not the positive ones, the negative ones, in detail, with an open spreadsheet. Decide on one thing your game is the best in the world at, and pour every cent of marketing budget into communicating that one thing. “Everything else,” Allen says, “is decoration.”
When asked whether AI will eventually make people like him obsolete, he laughs — the same easy laugh of someone who clearly does not believe a word of it. “Mate, the day a model can sit in a room with a tired dev at 11pm and tell them which of their two trailer cuts has more heart, I’ll happily retire. We’re not close.”
