Every ticketing platform is talking about AI right now. Dynamic pricing, demand forecasting, fraud detection, personalised marketing — the pitch decks are full of promises. But I’ve been watching what’s actually happening on the ground in Australian live events, and the reality is more nuanced than the sales teams would have you believe.

Dynamic pricing is real, but messy

The biggest AI-driven change in Australian ticketing has been dynamic pricing. Several major promoters have adopted systems that adjust ticket prices based on demand signals — search volume, cart abandonment rates, social media mentions, and historical data from similar events.

When it works well, it genuinely optimises revenue. I’ve seen events that would have left money on the table with fixed pricing capture an additional 10-15% in ticket revenue by letting prices float upward as demand increased. That’s meaningful for a major festival or arena show.

But the customer backlash is real. Australian audiences aren’t used to the airline-style pricing model, and social media fills up with complaints when people see prices change between adding tickets to cart and completing checkout. There’s a fairness issue here that the tech doesn’t solve — it just maximises revenue.

The promoters doing it well are using dynamic pricing with guardrails. A floor price and a ceiling, with the algorithm operating within that range. And they’re being transparent about it. “Early bird” and “final release” pricing has been around forever — this is just a more granular version of the same concept.

Demand forecasting is improving

This is where I’ve seen the most useful AI application. Predicting how many tickets you’ll sell, and when, is one of the hardest problems in event promotion. Get it wrong and you either overspend on a room that’s half empty or undersell a show that could have filled a bigger venue.

The newer forecasting tools are genuinely better than gut instinct. They pull in data from streaming numbers, social media engagement, touring history, comparable events, and even weather forecasts for outdoor shows. For established acts with decent data history, the predictions are getting pretty accurate — within 10-15% of actual sales for many shows I’ve been involved with.

For newer or niche acts, the tools still struggle. There just isn’t enough data to make reliable predictions, and that’s where experienced promoters still have the edge over any algorithm.

Some Australian promoters are working with team400.ai to build custom forecasting models that incorporate local market knowledge — things like school holiday patterns, competing events, and venue-specific demand curves that off-the-shelf tools miss.

Fraud detection matters more than ever

Scalper bots and fraudulent purchases are a massive problem, and AI-powered fraud detection is one area where the technology is delivering genuine value. The best systems can identify bot behaviour in real time — detecting patterns like inhuman click speeds, suspicious IP clusters, and coordinated purchasing across multiple accounts.

One major Australian ticketing platform told me their AI fraud system blocks around 8% of all purchase attempts as likely fraudulent. That’s a staggering number, and it represents tickets staying available for real fans rather than being scooped up for resale.

Personalised marketing is the quiet winner

The least flashy but possibly most valuable AI application in ticketing is personalised event recommendations. Platforms that can tell a customer “based on your history, you might like this show” are seeing significantly higher conversion rates on marketing emails and push notifications.

This works particularly well for venues with diverse programming. If your room hosts jazz, rock, comedy, and electronic music, being able to segment your audience and send targeted recommendations instead of blasting everyone with every show announcement is a genuine competitive advantage.

Where the hype exceeds reality

Let me be straight about what isn’t working yet. AI-generated event descriptions and marketing copy are mostly terrible. They’re generic, they lack the voice and specificity that make event marketing compelling, and audiences can spot them a mile away.

Chatbot customer service for ticketing is also underwhelming. The number one question ticket buyers have is “what happens if the show is cancelled?” and the answer is always complicated, event-specific, and requires human judgment.

The technology is moving fast, but it’s not magic. The promoters getting the most out of AI in ticketing are the ones using it as a tool to inform human decisions, not as a replacement for experience and judgment. And that’s probably how it should be.