It’s an accepted truth that Aussies love an underdog. And our industry does too.

Some of marketing’s best stories come from brands that had no business winning on paper. Smaller budgets. Less distribution. Smaller shelf space. And let’s not forget competitors capable of spending their annual marketing budget before morning tea.

And yet they still won.

They hustled. We hustled. A sharper strategy. A distinctive brand. Brilliant creative. Smarter media. Earned ideas that generated attention they could never afford to buy. They punched above their weight.

That’s part of the magic of our industry; you don’t always have to outspend the biggest player if you can outsmart them.

But then came the machines. More specifically, LLMs.

As consumers increasingly ask AI what to buy, I’ve started wondering whether the rules of that hustle are changing. Because when AI recommends a brand, it isn’t particularly interested or impressed by your new brand campaign, your hilarious TikTok, or the fact that Marketing press briefly lost its mind over your campaign.

It’s not to say your brand campaign is wrong (it’s so right in many ways), but AI is looking for trusted signals, at scale.

And if trust is easier to establish when you’re bigger, more widely distributed and more documented online, an already difficult fight for challengers may be about to get harder.

So, if Australians love the underdog, do machines?

When we stop searching and start asking.

For years, marketers mastered search. Type “ice cream my kids will love” into Google and you get ads, retailers, reviews, Reddit threads and enough SEO content to make you question whether you wanted ice cream in the first place.

You weigh it up. Then you choose.

LLMs collapse that journey. Ask “what’s the best-value family car?” and the machine researches, compares, and recommends in seconds.

Wonderful for consumers. More complicated for brands. Because the question is no longer just whether the consumer knows you. It’s whether the machine does.

So, maybe size does matter after all?

LLMs need confidence. And confidence comes from signals at scale; reliable product information, authoritative references, fresh and crawlable content, customer reviews, availability, digital PR and a consistent footprint across the web.

Now ask which businesses are best equipped to produce those signals at scale?

Spoiler. It’s probably not the craft hot-sauce company operating out of Marrickville.

Large corporations, major retailers and household brands have an inherent advantage. AI doesn’t need to deliberately favour them. It simply needs to know more about them.

For a machine making a recommendation, more evidence can mean more confidence. And more confidence can mean more recommendations.

Scale starts feeding scale.

FMCG makes the tension particularly obvious.

Imagine a challenger that has built a distinctive identity, loyal customers and a strong social following. It’s already fighting the retailer’s private label on shelf.

Now the customer asks AI before they even get there:

“What’s a good-value product in this category?”

The retailer’s brand has competitive pricing, ubiquitous availability, consistent product data and an enormous digital footprint.

That makes it a very safe recommendation.

Once? Who cares.

Across millions of purchase decisions? That starts to matter.

Retailers already wield enormous influence over what gets onto the shelf. What happens when machines start influencing what gets onto the shopping list?

From mental availability to machine availability.

For decades, marketers have rightly obsessed over mental availability; when a buying moment arises, does your brand come to mind?

In the AI era, mental availability's companion may be machine availability.

When someone asks a machine what to buy, does it know you exist? Does it understand what makes you different? Can it find credible evidence for your claims? Does it have enough confidence to recommend you?

This doesn’t make creativity less important. Nor does it mean we all abandon brand building and spend 2027 producing 14,000 “LLM-optimised” FAQ pages. God help us!

It means the job is expanding.

Growth brands may now need to punch above their weight twice: once with people, and once with machines.

So how does the underdog fight back?

Challengers can’t manufacture scale. But they can manufacture signals of trust.

Signals like earned media that builds authority. Paid media that creates demand and discovery. Websites structured so machines can actually understand them. Creative that gets talked about. Useful content that establishes expertise. Reviews, social conversation and third-party validation that give machines more evidence that a brand is credible.

None of this is a magic AI hack. Which is unfortunate, because “Seven Tricks to Dominate ChatGPT” would be considerably easier to sell.

It’s really the same hustle that has always helped challenger brands punch above their weight. The difference is that we’re now building two kinds of availability.

Mental availability with people. Machine availability with AI.

Aussies still love the underdog. Our next job is making sure the machines love them enough to give them a fighting chance too.