In online retail, brands today are overloaded with information, underserved by legacy, manual retail processes unfit for Amazon, and immersed in an ever-changing battlefront where algorithms are tuned to serve consumers and maximize Amazon’s profits. This happens even at the expense of the world’s largest brands.
If consumer brands can’t move fast enough, shoppers will abandon them, Amazon will punish them and more nimble competitors will feast on them. This is the new reality of today’s era of Real-Time Commerce.
As a platform designed from day one for speed and efficiency, Amazon updates prices, product listings and even search results every few seconds. There is not one Amazon, either. Each shopper gets a highly customized view of Amazon built on-the-fly designed to appeal to their needs and reflecting their past behavior. Imagine if every store in the mall could rearrange its inventory displays and make special price offers and bundles for each customer, and you get the idea of how massive a shift that is.
In the same vein, the Amazon platform is also designed as a massive feedback loop, constantly measuring customer behavior and the impact of minute changes to pages, prices, pictures, and text to spot trends that could help drive more sales.
There are more than 20 variables that Amazon’s search algorithms consider for each product page to determine whether that page lands on the first page of search results. Products that don’t land on the first page are 80% less likely to be purchased.
Note that I said product page - because the same products may land in very different parts of the search results depending on how the seller has described the product, images posted on the product page, how the product is priced, and other factors. Amazon processes tens of thousands of searches per second. This is the largest aggregation of shopping intent not only online today but in the history of retail.
On the back end, Amazon is a pricing and inventory engine for roughly 500 million products - a scale never seen before and one that dwarfs all online competitors. Across this massive constellation, Amazon automatically changes product prices multiple times per hour - something that’s impossible to do in a physical store. Third-party vendors selling on the Amazon platform also change their prices regularly. When inventories run low, Amazon sends an email to a brand or vendor. There is no phone call or hand-holding. If sales of a product or pack size jump and inventories run too low, Amazon might drop the listing unilaterally, with no warning.
That’s one risk. Here’s another. Amidst this constant shifting of prices and assessment of inventory, Amazon is constantly calculating whether to remove a product listing from the site if that product’s unit-level contribution to profits is too slim. This is called getting “CRaP-ed” (Can’t Realize a Profit) out on Amazon. Amazon will remove a CRaPed listing without notice. It’s up to the brands to figure out what happened and how to fix the problem. If a brand’s product is CRaP-ed out at the wrong time - in the middle of a high demand period - then the brand could lose massive amounts of sales in a very brief window. Sometimes brands even get CraP-ed out due to a price war between Amazon and a rival like Walmart.
That is not to say smart brands and sellers can’t win on Amazon. Savvy brands can piggyback on the popularity of rivals by buying sponsored listings on crucial keywords or bundling products smartly with competitor’s products. But this is harder for large consumer brands with hundreds of products to manage. In contrast, digital teams for smaller brands with fewer products can more easily monitor prices and spot opportunities where rivals are overpricing or are out of stock. This helps these smaller brands earn higher placement in Amazon search results, often at no extra cost and without additional marketing spend.
Smaller still are third-party sellers of products. These are the online equivalent of “fell off the back of a truck” sellers. They generally focus on only a handful of products which they manage very closely. These small third-party seller pursue super aggressive tactics like pricing a few pennies below the next closest competing seller (including for product pages of brands with a 1P (i.e., wholesale) relationship with Amazon) in order to attain the coveted top spot in the Amazon search box. These third-party competitors create different sizes of packs, as well, to draft off the visibility of a popular brand. The resellers are scrappy and digitally savvy, having grown up in Amazon and evolved their entire business structure to feed off the scraps left by Amazon’s algorithms at the expense of larger, less-nimble brands.
Laid out end to end, the new era of Real-Time Commerce is a massive problem for larger brands. Managing hundreds or thousands of product pages while keeping tabs on hundreds of competitor brands and thousands of resellers across dozens of variables that Amazon considers in search results equals billions and billions of potential combinations to consider.
The upshot? Large consumer brands simply can’t keep up with this velocity or scope of variables without the help of specialized tooling and technology. Building a tool to do this in-house would likely dwarf all other R&D costs they incur. Efforts to monitor and make changes manually are a poor use of a digital or online marketing teams’ time; the tasks are minute, grindingly manual and enumerable. Digital agencies can’t help, either. They may be ready to buy ads programmatically but are not equipped to manage hundreds of product pages and monitor prices, inventory and competition in real-time.
This isn’t a data, problem either. Data collection alone does not offer solutions to very complex problems like:
The brands that can leverage technology to quickly spot, then automatically act on and exploit these opportunities minute-by-minute, every day, will be the winners in the Era of Real-Time Commerce.
To win, they will need data aggregation and monitoring tools to capture all real-time information. They will require machine learning tools to analyze the massive piles of data, glean insights from the noise, then build self-learning algorithms and strategies. Those systems can take real-time actions based on what they see and learn. Riding on top of this Real-Time Commerce stack will be an intelligent prioritization engine that surfaces and acts on only the most impactful information and suggestions rather than drowning a brands’ team in a sea of unimportant data and marginal observations. Brands will need to know when a competitor undercuts their price by one cent to win Page 1 placement, when a slight change to a product description causes a slowdown in clicks, and whether to buy sponsored placements for a handful of products or all of their products in order to maximize profits.
That’s the irony of Real-Time Commerce. To win, brands must have technology that sees everything, but only calls out the handful of actions that can move the needle. And do that at real-time speeds, without fail. Sounds hard? There is no other way.