Understanding human behavior is one of the key points for creating effective advertising strategies. This includes consumers likes, dislikes, interests, shopping patterns, etc. The technology of artificial intelligence (AI) systems is used to track information (shopping patterns) and create powerful insights into market behavior. As companies compete to put the right products in front of the right people at the right time, AI’s enhance the capability of webshops and online merchants to provide personalized product recommendations.
Personalized Product Recommendations
To understand what personalized product recommendations are, think of a sales assistant in any given brick-and-mortar store. Let’s say you are in a shoe store and are looking at a pair of running shoes. The sales assistant is there to make your shopping experience easier. Therefore, a store representative would see you looking at shoes and they would approach you. They will ask questions, such as what colors do you like, what fabric should your shoe be or what brand, and so on. Then, the sales assistant can make recommendations based on the information you provided.
This is the exact same system AI’s data analysis works like. Based on a user’s searches, clicked products or pages, what similar users viewed and so on, data is gathered and analyzed. This data help algorithms determine what products can be recommended to certain users.
The purpose of product recommendations is simple, boost the average order value (AOV) by making a sale that was not originally intended by the consumer. The personal recommendations are divided into cross-sales and up-sales.
Cross sales refers to suggestions of products that go together, like shoes and pants. Up-sales refer to either increasing the amount of the products one puts in a basket or the value of the product. For example, suggestions can be made to discounts. If one would buy 2 pairs of the same shoe, the second purchased pair would be half price. If one would look at a certain brand, a more expensive brand would also be recommended, so the user would buy the latter instead. Thus, increasing the AOV.
Who is using Personal Product Recommendations?
Today, every webshop uses some form of cross and up-sale. Market leaders, such as Netflix, Facebook, LinkedIn, Google, YouTube, and Waze all use data to offer recommendations that better suit their customers. However, the best example of this strategy is that of Amazon. The retail giant coined this through their famous headline, “Customers who bought this item also bought…”.
Cross-sale and Up-sale have been around for a while now, so, how did Amazon become so much better at it? Amazon created Deep Scalable Sparse Tensor Network Engine (DSSTNE), pronounced “Destiny”. The company released this open source framework in 2016 with the hope that it will “spur innovation in many more areas”. This was meant to make the DSSTNE AI framework open to any and all businesses to use for their own product recommendations.
According to McKinsey, in 2013 already, “35 percent of what consumers purchase on Amazon and 75 percent of what they watch on Netflix come from product recommendations based on such algorithms“.
Aside from personalized product recommendations, targeted content also refers to ads. Crawlers and scrapers have the capability to track users’ movements across the internet and store data. They review searches in search engines such as Google, ads clicked by the users, likes on products and so on. With this data, the algorithms powering an AI are able to make recommendations that promote products and services relatable to specific users. Moreover, they can track timing, and see what ads perform best in a certain step of the buying process. This helps boost cross-sales and up-sales by showing them at just the right moment.
All the data gathered creates a sort of cycle. A user is targeted with an ad, they click the ad, they view a product, decide to buy, receive a product recommendation, increase their AOV, and proceed to checkout. After the cycle is complete, it starts all over again using the same data constantly collecting more and getting feedback.
Chances are if you bought a plane ticket in the last decade the price of your ticket was not determined by a human on the other end, but an algorithm.
Dynamic pricing is the ability to adjust prices according to a markets’ supply and demand ratio, as opposed to offering fixed rates on products or services. Already prevalent in service industries such as, airlines, hotels and ride-sharing, this is becoming increasingly popular in online retail stores.
The goal of dynamic pricing is to adjust the price to the highest value a particular customer is willing to pay. The highest price they are willing to pay, or the price elasticity of demand, determined by a series of macro and micro environmental factors, such as competition on the market, new taxes, changes in oil prices, income of population, and so on.
A good example of dynamic pricing variation is the water-diamond dilemma, also known as the paradox of value. This dilemma asks a simple question: which is more valuable, a bottle of water, or a diamond? At first, you could say, well… a diamond is clearly more valuable as it is more expensive compared to a bottle of water. But imagine you were lost in the Sahara desert, and suddenly a person would approach you and tell you “You can buy from me, right now, either a 10 ct diamond, or a bottle of cold water for 1000 euros”. If you would have 1000 euros, which one would you buy? Which one is more valuable to you then?
The moral of the story is that some will pay more than others due to nothing other than circumstances, in other words, data.
With all things powerful, there is the opportunity for abuse of power. Having access to all this information about consumers, it is important for advertisers to maintain a system of trust between the data collected and the ways in which it is used. The more consumers are understood, the easier they can be converted.
Governments understand the risks implicated with this amount of user information we provide marketers every day. There are already movements to protect users from scammers or mal-intended retailers. One such example is EU’s GDPR, that will be effective starting May 2018. This reform is intended to ensure that the data collected on users is better protected and rightly used. Moreover, it gives the user the anonymity they have a right to, if they should choose so.
Advertising has always been a topic of controversy. Whether it be the portrayal of gender stereotypes, advertising to kids or advertising of harmful products. The idea of advertising to some may seem unethical in itself as tools of persuasion are used to entice potential customers.
However, many consumers enjoy the benefits AI has to offer when it comes to advertising. The truth is, advertising has always existed in some form or another, and is here to stay. With AI, the user experience is only expected to increase in quality and relevance. Consumers are expecting more personalized experiences and those who embrace this opportunity will be rewarded.
It is easy to see that the rewards of using AI technology will not only be a few extra sales, but recurring and loyal customers that will help drive the business upwards. Advertising and marketing are just a few of the examples where AI can be used. Read more about how to use AI for online customer support or even build your own website