If the algorithm is the customer, how do Australia’s retailers need to change?

Business people riding electric scooter and bicycle on promenade against sky
  • Insight
  • 3 minute read
  • 17 Dec 2025

As our retail sector reconfigures around new domains of growth, agentic AI is already changing how Australians shop. Retailers who recognise this fundamental shift – and move decisively – will hold the advantage in seasons ahead.

John Higgins

John Higgins

Partner, Advisory, PwC Australia

We all know people who love the thrill of the hunt, who think gift-buying is a fabulous part of the Christmas tradition. They're the planners who enjoy browsing and comparing months ahead. For the rest of us, it's more like a seasonal sprint that kicks into high gear only when time is running dangerously short.  

Then comes New Year sales shopping, where the chaos flips but the urgency remains. While some of us are nursing last night’s revelry or making resolutions for the year ahead, others are diving into crowded shops and online deals to secure the best bargains before stock runs out. But this hectic holiday shopping season might be about to change for last-minute buyers and deal-hunters alike. AI in the form of agentic commerce is ready to step in.

Already, a growing number of Australians are using AI for retail decision-making. Our 2025 Voice of the Consumer survey found 56% of Gen Z Australians are comfortable using GenAI for meal planning and menu suggestions. For older generations, it is one in four (24%). On the horizon – and coming up fast – is agentic (AI) commerce, which autonomously completes shopping tasks – searching products, comparing options and making purchases based on personal preferences – without human intervention.

When value shifts, new ecosystems spring up

Such leaps in technology and changes in human behaviour were the inspiration behind PwC’s Value in Motion research. This shows how traditional industries like retail are already starting to reconfigure around new domains of growth – each representing a system of value creation. Leading businesses are redesigning how value moves, how it’s created, and, increasingly, how it’s consumed.

In fact, shoppers are already using agentic commerce this holiday season. Amazon’s Buy For Me pilot sends AI agents on missions to purchase gifts from external brand websites without customers leaving the Amazon app. The algorithm choreographs everything across the retail ecosystem (not just Amazon’s inventory). Meanwhile, Perplexity has teamed up with PayPal to enable instant checkout directly within its AI search engine – fast-tracking consumers from opening question to purchase to delivery.

Our Value in Motion research shows that AI will increasingly be used to create ecosystems and partnerships like this. In fact, consumers will expect it. 

The question for those of us who work in Australia’s retail sector is: will we be ready? 

Algorithm as the customer

On future Christmas Eves, I’ll have friends whose AI agents will have become so attuned to their shopping patterns, family preferences, and gift-giving history that Christmas shopping transforms from frantic hunt to effortless execution – with every purchase tailored to perfection. For some Australian retailers, this switch in consumer behaviour could force a rewrite of their marketing playbook. If AI agents make purchasing decisions – weighing up price, delivery speed, service terms, sustainability credentials etc – then the algorithm effectively becomes the customer.

Just as retailers once optimised for Google’s search algorithm, they must now master generative engine optimisation. But this is about more than marketing and efficiency – it’s about rethinking the entire value creation model.

For retail leaders, agentic AI provokes some pressing questions: 

  • How will we win when our customer is no longer a person but an algorithm optimising for value, friction, and relevance? 
  • What will differentiate us when agentic buyers negotiate, subscribe, bulk buy, or switch suppliers in real time? 
  • Do we have the data, systems, and governance for machines to trade with us safely, profitably and at scale?

On the one hand, I can think of a few quick wins: Retailers could expand product descriptions from 30 words to more than 1,000 words per product, making products, pricing, and availability more ‘machine-readable’. This rich detail allows agentic AI to make nuanced decisions – recommending a Pinot Noir that cuts through the richness of a roast turkey Christmas lunch, for example.

On the other hand, retailers need to consider some long-term issues: 

  • Who is our customer of the future and what is our value proposition for them? 
  • Price aside, what aspects of service will influence our future customers' purchasing decisions (delivery time, return policy, packaging)?
  • How do we avoid margin erosion when algorithms can detect – even predict – pricing behaviours?
  • Where in our business do we champion human experiences, and where should we optimise for algorithmic efficiency? 
  • Ultimately, will our bricks-and-mortar stores become irreplaceable, or irrelevant?

Obviously, the answers to those questions will vary for every retailer. High-frequency, low-emotional attachment products like groceries could face the greatest disruption (although some consumers will still want to surreptitiously squeeze their avocado or sniff their pineapple before they buy). Premium, discovery-based products like furniture and luxury goods may always champion experiences that algorithms cannot replicate.

Opportunity knocks for CPGs

Ecosystem collaboration could reshape how value is created for consumer-packaged goods (CPG) brands and manufacturers, too. Agentic commerce presents the tantalising possibility of influencing purchasing decisions directly. This could potentially make CPGs less reliant on traditional channels to market (retailer partnerships, loyalty programs, above-the-line advertising, etc).

Creating trusted ecosystems

As the retail sector reconfigures, success will depend on having the right capabilities in the right places at the right times. Reinventing business models can be challenging, but retailers don’t have to build everything in-house. 

Collaboration can be a game changer – which is where ecosystems come in. Global PwC research shows high-performing businesses are nearly 50% more likely to have a clear ecosystem strategy in place and likely to generate at least 60% of their revenues from ecosystems.

Now is the time for retailers to explore potential GenAI partners and commercial models. Pertinent questions to ask include: 

  • Which AI providers align with our brand and consumer base? 
  • Which partnership structures make the most sense for our business?

Such questions also extend beyond AI and into the wider retail value chain. For example, retailers might consider which partners could help meet consumers’ (and their AI agents’) higher expectations of fulfilment. Flexible, hyper-local delivery networks could become essential here.

In short: for many retailers, the question isn’t whether to collaborate, but with whom and how quickly.

Shifting value requires a shift in mindset

Value in Motion research suggests the pressure for businesses to reinvent is at or near a 25-year high. Tomorrow’s winning retailers should be thinking today about how to innovate at scale.

The next 12-18 months present a window of time for Australia’s retailers to imagine a world where the algorithm is the customer – and to consider the strategy, capabilities and partnerships that will be required.

Agentic commerce appears set to grow and grow. The degree and timing of impact remain unknown. But those who move early will hold the advantage in festive seasons ahead.

Contacts

John Higgins
John Higgins

Partner, Advisory, PwC Australia

Brian Man
Brian Man

Partner, Customer Transformation and Retail and Consumer Industry Lead, PwC Australia

Samantha Jansen
Samantha Jansen

Director, PwC Australia

Follow PwC Australia