When was the last time you did not see Doritos chips, Coca Cola or Brawny paper towels during your weekly run at the grocery store? Probably never because these brands spend big money with these stores to be visible. Now when was the last time you saw these brands on Amazon? Probably never. I call that the dictatorship of the Amazon search engine. Unless you know the rules of the Amazon game, you are likely to be stuck on page 73 of the search results.
Because these brands are primarily using humans to work with Amazon. Those poor humans are attempting to understand the machinations of arguably the most complex intelligent system ever created in history.
Consider this. Right now Amazon is managing 600 million products in its online catalog. This would fill 4,000 Walmart superstores built next to each other. That is roughly 3 square miles, larger than the entire island of Manhattan. A typical Top 100 consumer packaged goods (CPG) company sells 1,500 products, which would easily fit in a few aisles in the back corner of a single Walmart store.
These are smart brands led by competent teams who have built considerable customer loyalty. But they are dealing with a machine. Machines and algorithms control the way Amazon adjusts prices and search page results, in real-time. Machines manage Amazon’s relationship with millions of suppliers. Amazon’s warehouses are now teeming with robots, all powered by Artificial Intelligence (AI). You get the picture.
While Bill Gates and Elon Musk are warning us against AI, Amazon’s CEO and founder Jeff Bezos loves AI and embraces it. Bezos says “…much of what we do with machine learning happens beneath the surface. Machine learning drives our algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising placements, fraud detection, translations, and much more. Though less visible, much of the impact of machine learning will be of this type — quietly but meaningfully improving core operations.”
So what to do? Give up and let your product’s success be dependent on an algorithm that you can’t control? Amazon gives you the option to drop a few million dollars into advertising campaigns and juice your search placement to gain prominence in front of Amazon shoppers. But as soon as your marketing dollars dry up, your product listing drops like a rock, making your product all but invisible again.
And even the ROI of an Amazon Ads campaign can be questionable if a brand isn’t using it to support a more concerted and long-term Amazon optimization effort. Prior to working with Boomerang, one of my newer CPG customers spent $150,000 in one month to drive $330,000 in sales, an ROI that will quickly bankrupt your business.
To operate in the universe of super-intelligent machines, you need to become a machine and leverage AI like a pro - like Amazon, in fact. This is still quite new for most consumer packaged goods (CPG) businesses.
Trying to unpack the Amazon search algorithm is what we have been working for 5+ years at Boomerang Commerce. We have been using our machine learning bots to test what works and what doesn’t. Based on our learnings from this adventure, I can say Amazon’s systems are designed primarily for Amazon’s interest across a broad range of measures - sales, conversions, margins, importance of the category and dozens of other factors. That makes sense to some degree, Amazon’s algorithms are designed primarily for the end customer, or the shopper. Amazon holds itself responsible for shopper experience. That said, most of the typical easy win-win scenarios that brands have come to expect in brick-and-mortar stores are out the window.
The only way to compete in this new arena is to build technology that can speak “Amazonian”. You need a machine, not a human, to do this. This reality is becoming even more urgent as we see Amazon’s growth and dominance accelerating.
We are now seeing a new generation of buyers that never go to any stores. They buy nearly everything online for convenience, and their world is 100% dictated by results from Amazon’s search engine. They don’t compare based on prices. They don’t even know other options exist, nor do they really care. This may create an existential threat for many brands who don’t act quickly to learn Amazonese. Will the brands you and I grew up with survive this massive change? That’s the $1 Trillion Question. Literally.