The Problem With Personalization Today

A recent study from Salesforce found that 72% of consumers would stop buying from a company because of privacy concerns. The problem is, the concerns keep mounting. From news of hacks and data breaches, to Google or Apple allowing users to blanket opt out, to the flipside, TikTok forcing users outside of the EU to opt-in to personalized ads. And the fact is, companies whose revenue relies heavily on user data don’t want to make it easy for those users to refuse to provide it. Users are having to weigh up experience with privacy. Now, that doesn’t seem very customer-centric, does it?

Retailers with mature personalization strategies are likely to have built complex engines for tracking a customer journey across multiple touchpoints. These strategies often include a combination of IP or MAC address tracking, persona-based personalization, demographic personalization, and stored cookies. The problem with this is that if any of these data points become obsolete by law or public opinion, personalization stops and the customer experience suffers. 

For retailers without mature personalization strategies, there is so much noise from vendors all promising similar outcomes. Most personalization providers utilize methods to capture and retain data indefinitely, which doesn’t benefit shoppers.

Here are some ways retail technologies personalize content:

Demographic or Profile-Based Personalization

This type of personalization assigns shoppers to segments based on things such as location, job, etc, and even some behavioral data such as typical purchase size and frequency. The benefit of this method is that it seems like a full picture is being created, but in fact this is a risk. The problem with segmentation of this nature, is that it doesn’t take into account what the customer wants, and instead makes assumptions based on bias and potentially outdated behavior. 

For the luxury sector, for example, aspirational shoppers are just as important as those that purchase, so it’s more about clicks on a site in real-time, rather than purchase history, to make engaging recommendations.

IP and/or MAC Address tracking

A Media Access Control (MAC) Address is a number that every electronic device has. Whilst it can’t be traced back to an owner, as there is no centralised database, it can be used in conjunction with an IP address to connect experiences across devices without shoppers knowing it. The benefit here is that it does provide a seamless experience for users with multiple devices on the same IP address. The risk is that consumers don’t get the choice, which can feel intrusive. 

Cookie-based personalization

Different types of cookies keep track of different activities online. Tracking cookies are used to link multiple visits to the same site. Session cookies are used only when a person is actively navigating a website, and once the shopper leaves, the session cookie disappears. The benefits of cookies are widely known — if cookies didn’t exist in online retail, your basket would empty every time you clicked a new link. Using cookies alone as a personalization strategy is risky because people can clear their cookies anytime and legislation about their use is in flux. 

Will it ever be possible to deliver engaging, personalized experiences using sustainable methods? 

Introducing Personalization with Privacy

Klevu offers shoppers fully private, fully personalized experiences by letting shoppers lead the way. But how?

With Klevu, personalization starts from a search. When a search happens, Klevu identifies the shopper’s intent and matches their behavior to the behavior of others using the Klevu platform. This is known as collaborative filtering, a method used by Spotify to deliver hyper-relevant results (some say eerily so!). Klevu then uses the information gathered to optimize the shopper journey. 

With Klevu, the AI is customer-centric. It allows shoppers to drive the machine. Allowing shopper behavior to influence the merchandising is a more genuine approach to win shoppers’ trust and confidence. 

For search results, Klevu AI personalizes all of the results below pinned products on search results pages. Search terms influence search results as well as product recommendations.

For category pages, personalization resets after a short period of time, in line with Google’s personalization policy. This ensures relevancy, and helps shoppers not get stuck in a filter or personalization bubble when browsing.

For product recommendations, products in recommendations banners keep changing dynamically based on shopper behavior and the behavior of others.