The Amazon anticipation is about more than cheap batteries

After the endless lineup of sombre Amazon opinion pieces, Omnicom's Tom Fryett offers something a little different - a warm welcome our new robot overlords.

At time of writing the rumour mill is in whatever comes after overdrive, but if the Amazon Australia launch is indeed finally here, then this article will join a few hundred more opinion pieces written in the coming days and weeks on what it means for the way we shop in Australia.

Most those written will hopefully be sombre reflections on the various predicted percentage points lost by different retailers, backed up by a great deal of careful analysis, not hyperbole.

This is not that article. I am an unabashed Amazon acolyte and for me their coming has been rightly heralded. By day as a media planner, I of course remain objective and fact-based, helping to prepare ecommerce strategies for clients. But as a consumer, I welcome our new robot overlords. Come and enjoy this view of the future, as we walk together to the peak of inflated expectations.


Amazon “employs” 45,000 robots in its warehouses, having bought Kiva Systems, a robotics company in 2012. This has become an oft-cited example of how automation is set to replace human labour.

However, a recent article in the New York Times quoted employee Nissa Scott who works in an Amazon warehouse in southern New Jersey. Her experience runs contrary to the popular line that robots are reducing the number of roles available. Instead, they ease the physical burdens of warehouse work and create new, more interesting roles for human employees. So, nothing to worry about then.


Amazon trialled its first successful drone delivery back in December of last year, in sunny Cambridge, England – by delivering a TV streaming stick and a bag of popcorn directly to the garden of a nearby customer. Is this the answer to Amazon’s Australian distribution challenge? Could drones help them cut down the last mile in rural areas? Many think aerial delivery could become a viable business sooner than thought.

Unfortunately, the current list of prerequisites is: having a huge garden, living close to the delivery depot and wanting an item that weighs less than 2.6kg. Drones care not for weekends, so the good news is it’s available seven days a week – but they can only fly during the day, if its sunny and there’s no wind. Sounds like physical popcorn stores are safer than other sectors for now.

Machine learning!

In 2014, Amazon patented “anticipatory shipping” – essentially predictive logistics and delivery based on their humongous purchase data pool. Since then, Amazon appear to be pursuing roadmap where they will speculatively send you products based on what you have bought previously, only then paying for what you decide to keep.

Amazon Wardrobe, their move into the fashion sector, appears to be the outrider to this model. Imagine a future of fashion recommendations from an algorithm – essentially Alexa sending you a new t-shirt because it thinks it will go great with the jeans you bought last week. In the background, in their relentless pursuit of faster delivery times, they will stock items in distribution warehouses based on what they think consumers in that area will order in the future. Data could prove to be Amazon’s most powerful potential weapon in a country as vast as Australia.

Shops without checkouts!

PHD’s recent publication, MERGE, discusses a future where technology enhances our lives by streamlining routine and mundane processes like waiting in line to check-out of a store, allowing us to start enjoying our new purchases right away. Amazon has done exactly that – leveraging technology that scans smartphones upon entry into their stores and monitors what’s removed from shelves.

The prototype Amazon Go store in Seattle was met with curiosity last December, being open only to Amazon employees adding to its mystery. Then when Amazon bought Whole Foods for $13.7bn in June this year it started to make more sense. Amazon Go requires shoppers to scan their smartphone upon entering the store, and the company’s “just walk out” technology will keep track of what is in your virtual cart through your smartphone.

When a shopper is done and they leave the store, the company charges their Amazon account. Amazon hasn’t revealed much about this detection technology, but said it was using a mixture of computer vision, artificial intelligence, deep-learning algorithms and other “sensor fusion” technologies that were pioneered in self-driving cars.

Award winning content!

Amazon Prime is a formidable proposition. In return for $120 dollars a year you get free delivery as well as a range of other services like Amazon Prime Video. It’s a strategy that’s working well – the average non-Prime member spends $625 a year in the US. But the average Prime customer spends $1,500 – as convenience and ease give rise to purchases like booking your cleaner via Amazon, let alone your groceries.

It’s a strategy that will see Amazon spend $4.5bn dollars on original content this year and may even lead it to buy cable channels or compete for more live sports rights around the world.

Netflix will spend around $6bn this year on original content and that’s its core proposition. Content is Amazon’s $4.5bn loss leader to become the first and last store you ever need.

Virtual assistants!

As Alexa and other virtual assistants get more and more lifelike, it’s easy to see why futurists and trend spotters like Mary Meeker are highlighting that voice is often the more natural way to interact with a computer. With a growing user base, Amazon Echo devices will fill our homes with Alexa – be that via the original Echo, the Echo Show for video calls and not forgetting the Echo Look giving you those fashion recommendations.

Presenting then, our new target audience, Alexa. Or maybe more specifically the A9 algorithm. As marketers, we’re going to need to know how to optimise our media buys to not just attract the attention of human eyeballs on a physical shelf, but an algorithm crawling the virtual shelf of ecommerce.

Currently the A9 algorithm incorporates not just sales but conversion rate, click-thru-rate, keywords and the listing more broadly to decide relevance to the user and ultimately to predict purchase likelihood by a human customer. But what will an A.I. driven virtual customer look for before putting it in our Amazon Wardrobe maybe box. Where are we most likely to be at a given time – at home or work? And how does that affect delivery time? What have we watched recently on Amazon Prime Video? What was our response to our Echo Look when we tried on the last t-shirt?

The arrival of Amazon in Australia has generated a great deal of commentary. Reading some, you would be forgiven for thinking the main reason for the “Amazon is coming” mantra being repeated up and down the corridors of the agency where I work since April is the prospect of slightly cheaper batteries.

But deep down Amazon’s arrival in Australia is exciting because of what it represents – disruption. Amazon will be a catalyst for a fundamental change in the next five to ten years of how we buy just about everything, even sectors where it doesn’t (yet) compete. Amazon or equally another retailer prompted by its arrival will now look at every friction point that exists within a sector and begin to make data-driven decisions about how to make it a better, more seamless experience for the consumer.

With any misstep, Amazon or said other retailer makes in the coming decade, commentators will point and say “see, Australia is too large to overcome the distribution problem”. But eventually the drones, the data and the delivery times will come together until ecommerce is simply commerce again. Eventually feeling as redundant a differentiator as separating out digital media has in recent times.

The danger as ever is not paying attention to the almost imperceptible, incremental changes that will now occur. Not moving at the speed of the consumer and ensuring your 1st party data is clean and actionable. Not making sure your product is optimised for the digital shelf. Not ensuring your site load speed on mobile is measured in milliseconds, not seconds.

When Amazon finally launches, will you be ready?

Tom Fryett is head of programmatic Sydney for OMG Programmatic.


Get the latest media and marketing industry news (and views) direct to your inbox.

Sign up to the free Mumbrella newsletter now.



Sign up to our free daily update to get the latest in media and marketing.