AI for dynamic pricing in retail

Using AI and ML for Dynamic Pricing

A new look at retail & e-commerce strategies

Old-fashioned corporate espionage is almost passé compared to what we can do nowadays with simple information processing of data acquired from the Internet. Why go to the trouble of doing anything illegal when virtually everything you need to know is released to the public as a matter of course?
Artificial Intelligence (AI) and all its vital subcomponents, such as Machine Learning (ML) and Neural Net (NN) processing have allowed us to gather and sort data the instant it becomes available. And it’s just the readily available everyday information of communicating with investors, sustaining clients, or seeking new business that reveals all this information for those wise enough to see it. AIs allow us to develop insights that just a century ago would have been thought of as divination or prognostication of the first order.

Seeing the Future

When you have enough information, understanding what’s going to happen in the future is no longer a matter for fortunetellers or mystics. It enters the realm of statistical certainty. Of course, this requires a “trained” machine with knowledge relevant to the problem.
For example, in its simplest form, ground coffee plus boiling water equals coffee. However watching thousands of different instances of this process occurring, and the consequences therefrom would allow an AI to see that Robusta beans receive a poorer reception than Arabica beans, for example.
It might note that Middle Eastern techniques of putting fresh coffee grounds in a pot, boiling and cooling it three times, and then adding copious amount of sugar before serving in tiny cups, is a very popular method of preparation. Yet elsewhere in the world, this might be unacceptable to many, instead preferring percolated, espresso or the extra treatment and ingredients to make cappuccino.
Even the origin of the coffee beans, such as Jamaica, Guyana, or Hawaii will influence the enthusiasm of the end user. Once you’ve gotten to that stage, a light, medium, or dark roast of the same beans can vastly alter the flavor.
Just the facts, Ma’am
Why is this minutiae important? AI is extremely dependent on having information, especially of the asymmetric variety, where you know something that your competitor doesn’t or can’t know.
Expert humans take decades to acquire the knowledge essential to their expertise. A surgeon knows what to do if an artery is accidentally nicked, but that was learned by watching many operations beforehand as a resident, not by figuring it out when it happened. Your unique knowledge database differentiates you from your competitor.
A fully established AI can acquire “knowledge” as quickly as you can upload a relational database. Ah, but there’s the problem!
You have a vast amount of information about your product or service, knowing everything from your best sales representative to your biggest client; from the name of the suppliers that allow you to stay in business, to the list of the most common and cheapest providers of everyday materials; even the annual weather predictions in production areas that determine how much of a given resource you will have available to purchase. That all needs to be organized in a way a computer can use it. There are programs designed to do exactly that.
But is there a way to know if insurrectionists or “freedom fighters” are going to jeopardize your international supply chain? Can fashion houses know if “pastels” will be the new “white”? Is there a solution for rapidly changing production techniques and decreasing prices killing off 25% of your business in the next five years because people can order a customized version of your “standard product” (running shoes come to mind) delivered in just one week? The answer to these questions and millions more is an unqualified “yes”.

Knowing What We Know

Possessing knowledge vs. actually using it is a very important distinction. A properly programmed AI would scan daily news, and historical trends, recognizing changes that humans wouldn’t see. It can suggest sales strategies that wouldn’t occur to us.
We don’t stock swimsuits when it is snowing outside. Yet an AI would recognize when a very influential travel agency was gearing up to offer massive price discounts on all-inclusive Jamaican Resort vacations during January. We can get onboard that sales-train by hanging a new sign saying “Resort Wear”, and adding a “Heading for a hot destination this winter?” flash to our website and ordinary circulars.
Customer behavior
Are humans aware of these campaigns? No. Shoppers come in, looking for beachwear, find nothing and leave without ever notifying a clerk. Cluing into these actions puts us head and shoulder above our competitors. Customers start to think of us as the smart, progressive store that always has what they need.
It’s not confined to retail, of course. B2B ventures love partners or suppliers that always have a ready supply of what they need.
Typically, customers that visit any site frequently have pretty well decided to buy. Creating a slight increase in prices may trigger the purchase, the customer fearing that the price is going to rise further. This is particularly useful in the pre-holiday season to capture early sales, or mid-holiday season, as sales begin to peak.

Competing retailers

Knowing what your competition is charging for your similar products is essential for success. In the old days, competitors would send undercover shoppers to record retail prices of the competition. Those days are long past because we can simply refer to the retailer’s websites for the latest information.

Humans doing that would be labor intensive. It would also be out-of-date the instant it was complete. AIs, on the other hand, thrive on constantly updated data, are exceedingly accurate, and never get tired when they are collecting it. And as we all know, consumers respond well to pennies of difference. What retailer do you know who always beats the competition by 2¢? $19.99 vs. $19.97—you know who it is!
Cost vs. Price
As you can see there are a lot of factors contributing to setting a price point. Your competitor’s price is certainly important, but the AI is also going to be aware of your warehousing costs, your labor costs, and most importantly, your purchase cost.
In modern retail, the days of loss-leader pricing are essentially over. Consumers are savvy enough buy dozens of nonperishable $2 items that are being sold for 99¢, as are your competitors, who will be happy to stock up at your expense. The AI will see to it that no such blunders are made.


Paralleling consumer behavior is Trend Theory. Predicting what will be in demand is one of the greatest strengths of AI. Let’s say (for example) that a new material called carbyne which is twice as strong as graphene (the strongest material previously known) is now available in commercial quantities. Harder, stiffer, and stronger than diamond, would the steel industry simply “go away”?
Would every new smartphone be encased in carbyne? Would cars, boats, planes, and trains suddenly become completely indestructible? Would the world be dotted with space elevators that extended up into Low Earth Orbit, and would all of the modern spaceship launching companies all go out of business?
Even without previous knowledge of such obscure possibilities, AIs would be in the best position to extrapolate from such events. For that reason alone you need to make sure that you are taking full advantage of an AI’s ability to accurately predict the future—because it’s your future—the future of your company.

The Road to Success

You must meet the needs of your customers to become their go-to source for whatever product you sell. Those needs include price, practicality, and functionality, as well as the intended application. AIs can identify micro-segmentation, in products and customer desires, so as to determine the ideal price for an item.
If we look at video cards for computers, the majority of them are used for their intended purpose—making a computer display work—but an appreciable percentage of them are now being used to mine crypto-currencies (like Ƀitcoin). The superior mathematical abilities of a video card’s GPU (Graphical Processing Unit) compared to a CPU make them ideal for this sort of work.
Regular video cards are still sold at regular prices, but top end cards are making a fortune for their manufacturers. Understanding this micro-segmentation of the market allows them to get the highest unit price possible for each product while still out-competing “the other guys”.

The Takeaway

Artificial Intelligence, Machine Learning, and the ability to emulate human thought with remarkably effective Neural Networks will change the world—already has changed the world—and will continue to drive change in the future. This is not the time to be sitting back and let others gain experience.
The people and companies that get in fast and early will have a tremendous lead over latecomers—possibly an insurmountable one—so don’t be that person. All of these challenges and hundreds more can be solved with a properly deployed and trained AI. You now need AI just to stay competitive, let alone to excel.
Please feel free to start a conversation with us here. Our plug and play platform is the ideal tool for end-to-end implementation of AI in your business. Give us a call today!