TLDR
Lendi Group's marketing mix model found free-to-air TV was underweighted in its $10 million-plus annual paid media budget, and Lendi moved money into it from less efficient channels. Ads on ChatGPT and Google's Gemini delivered CPMs and acquisition costs too high to justify material spend. AI-influenced home-loan shoppers arrive at Lendi further down the funnel and better informed, compressing the measurable discovery window.
KEY TAKEAWAYS
What the model found
Lendi Group, the Sydney business behind Aussie Home Loans, spends more than $10 million a year on paid media, and its marketing mix model told it one of the oldest channels on the plan was being short-changed.[1] Michael McCash, Lendi's general manager of growth, told AdExchanger the Mutinex modelling showed free-to-air TV was "underweighted" and efficient, so Lendi responded: "We [started] diverting money from elsewhere in less efficient channels."[1]
McCash joined Lendi in 2024 as its first head of customer growth, after marketing roles at DoorDash and as chief marketing officer of Providoor.[2] His starting point, he told AdExchanger, is that "you can't improve what you can't measure", and Lendi now runs its own multi-touch attribution model alongside the Mutinex MMM.[1]
| Job | What Lendi uses |
|---|---|
| Marketing mix modelling | Mutinex |
| Multi-touch attribution | Built in-house |
| AI measurement (pilot) | Mutinex AI Visibility |
| Brand tracking, including how AI platforms treat Lendi against competitors | Tracksuit |
| Customer and mass-market surveys | Askable |
Lendi Group's measurement stack. Source: AdExchanger, 19 October 2026.
Where AI chatbot ads sit for now
Ads on AI chatbots such as ChatGPT and Google's Gemini have not been a major driver of return for Lendi so far, because the CPMs and customer acquisition costs are too high to justify much spend.[1] That doesn't mean the chatbots aren't changing how people shop for a home loan. McCash described them as "the first port of call" for some people looking for a home or a loan, and said those customers arrive at Lendi better informed and further down the funnel.[1]
"All of that pre-work is being done early and in-platform now," he said, which leaves the "initial discovery phase" looking "somewhat hidden again".[1] That is why Lendi is a pilot customer of Mutinex's AI Visibility product and uses Tracksuit to see how AI platforms treat the brand against its competitors.[1]
What it means for Australian media buyers
The free-to-air finding is the kind of result the networks want in every pitch: a performance-led lender, measuring with a local modelling vendor, moving money back into broadcast TV because the model says it works. The AI finding cuts the other way for the companies now selling chatbot inventory, since a brand can see that customers are using the chatbots and still decide the ads aren't worth the price yet.[1]
Mutinex, which Henry Innis and Matt Farrugia founded in Sydney in 2018, raised $17.5 million in 2024 to expand in the US.[3] Lendi's pilot of its AI Visibility product is an early local test of whether a modelling vendor can measure the part of the journey that now happens inside a chatbot, before a lender decides whether to pay to be in it.
SOURCES & CITATIONS
Xaviery Malinao writes for Prompt the Market on how brands and agencies are adapting to answer engines, drawing on Bushnote's work with clients across search, AI search and content.
