eCommerce Personalization: Complete Guide for 2021
January 7, 2021
What is eCommerce personalization?
eCommerce Personalization is a group of techniques to deliver custom content on eCommerce sites based on the customer’s personal data or behaviors in order to increase the conversion rate, revenue, and long-term value.
eCommerce AI personalization combines eCommerce Personalization with tools that use artificial intelligence and machine learning to deliver personalized content.
The personalization of an eCommerce site has a significant effect on Conversion Rate Optimisation.
The techniques can be used to b2c or b2b eCommerce personalization. This article focuses mostly on the b2c, but most of what is written below can also be applied to b2c.
The three pillars of Personalization
When it comes to eCommerce Personalization, you have to introduce three pillars that support it:
The Customer Data Platform (CDP): where the data of all users and visitors are centralized
Personalization Techniques: how to perform Personalizations based on the user data
The Personalization Platform: the platform that uses CDP data to implement Personalization Techniques
The Customer Data Platform (CDP)
Data collected on the CDP about users can have the following sources:
Touchpoints made on Advertising channels
Past purchase data: enter past data and set the connection for future sales.
Other data released by the user: can be quizzes, Net Promoter Score.
Enrichment from third party data providers: For example, financial, income, or preference data
A CDP is complete when it allows you to do the following things:
Create user segments
Synchronize segments with other platforms: for example, with Google Ads or Facebook Ads
Use segments for personalization techniques
Create a multi-channel customer journey with effective "Orchestration" based on the user’s specific actions or reactions. Orchestration is a separate topic that requires a particular article.
Many of the personalization tools already have within them, some more, some less, the concept of Customer Data Platform. Without data, the personalization loses its meaning.
The 9 Personalization Techniques
Once you have organized the Data, you can start customizing the user experience. Below are all the techniques that a personalization platform can offer.
Here are the nine eCommerce Personalization Techniques:
Products Recommendation Widgets
Each of the techniques can be used in multiple eCommerce personalization use cases. I will highlight some use cases for each method, but each vertical should consider their users and purchase process.
1. Product Recommendation Widgets
On each e-Commerce, there are some specific places where you can insert product recommendations:
Product Detail Page
Order List Page
Order Detail Page
A widget is a section on the page or app screen where the Personalization Platform shows the recommended products.
The products displayed can be chosen according to different algorithms:
Most Sold: The most sold products on the site
Recommended for you: use user data to recommend the most suitable products for you
Trending: Products that are accelerating sales.
New: The new products
Similar Products: products similar to the one displayed
Bought Together: products purchased together with the one displayed
People who viewed this also viewed: products that have seen people who have already visited this product
People who bought this also bought: products purchased by people who bought this product.
The objective of the recommendation is obviously to make users buy more.
Different product recommendation platforms use various techniques to make the best recommendations:
Use machine learning in a variety of ways to choose from thousands of products based on data collected in the CDP
Allow the shopkeeper to set some business rules on the algorithm: category limitations, price limitations, geolocation limitations, database limitations in the CDP
Use of other exogenous factors, such as the weather in the geographical area from which the user views the site/app.
Fallback: use of a secondary algorithm in case the first algorithm does not return enough recommendations
2. Triggered Emails
This type of personalization consists of sending an email after specific actions:
Cart Abandonment: when the user adds a product to the cart but does not proceed with the purchase. The mail contains the products and a Call To Action to complete the purchase.
After search: when a client searches for a specific word but does not complete the purchase, the email contains products displayed after the search.
After order: in this case, you send an email x days after the order.
Custom trigger: in this case, the email is sent after a specific trigger inside the system. For example, when the order has been successfully delivered.
3. Personalized Emails
eCommerce email personalization consists of sending specific emails with recommended products. Generally, on the personalization tool, it is possible to select the algorithm with its possible Personalizations.
It can be implemented in two ways:
Widget inside regular newsletter: insert the recommendations in a routine newsletter
Product recommendation email: send a specific email only with recommendations.
Depending on your industry, but as a general recommendation, it is always better to minimize unwanted emails to users. It is better to include recommendations at the bottom of all emails you send for other needs:
This type of personalization uses notifications to send customized products or messages.
The delivery channels of the notifications can be:
Android Native App Push Notification
Apple Native App Push Notification
Chrome Push Notification
Safari Push Notification
Also, if you have a live chat on your website, you could send these notifications via these channels, such as:
Any other messaging platform that your customers use based on geography
Again, as with Personalized Email, I would suggest caution with these notifications. Maybe test them with a Segment subset and if you see an interesting click rate or open rate, then expand to the rest of the segment.
This type of personalization consists of showing customized popups within the page or app during the visit.
Exit popups can have the following types of interaction:
Get a resource + collect data (including email)
Product or Service Offer
Quiz + collect date (including email)
The popup trigger can be generated under one or more of the following conditions:
Exit intent: when the user is about to exit the page. Applicable almost only to desktop users.
After a specific number of pages visited
Based on CDP Data or Segments
After a total amount of seconds or minutes on the site
After a total number of seconds or minutes on the page
On certain pages
After a specific user action: add the product to the shopping cart, add the product to the Wish List.
6. Category Order
This type of personalization orders the products shown within the categories to put those that might be of interest to the user higher.
With the continuous transition to e-commerce via mobile devices, this technique is becoming more critical as mobile device screens have less space in the viewport. If the user does not immediately find what they are looking for, they have more chances to leave the store.
7. Page changes
This type of personalization allows you to change static pages based on factors already illustrated in the examples above.
Some use cases include:
In a clothing store, men should see the male clothing if the CDP knows that a user is a man.
Show a different page based on the traffic source.
Show specific payment methods instead of others.
Show a different product picture—for example, plus-sized models or color choice for models.
Search engines on websites until a few years ago were useless. Over time, specialized companies were born that have created technologies to implement the search for content within their website. Nowadays, they have achieved a good level of UX.
The personalization of website search engines is the next step to make them useful to increase sales.
The search can be customized in the following three ways:
Search Box Autocomplete: each user sees a different search autocomplete according to their data.
Search Box Results Preview: each user sees a different search results preview according to their data.
Search Results: each user sees a results page sorted differently according to their data.
9. Social Proof
Social proof techniques are not a real personalization in themselves, but many platforms offer it within their services.
If used in combination with user data, they can be useful.
Here are the major types of Social Proof:
People are watching: showing how many people are watching a product to stimulate the effect of urgency.
People bought: show the people who bought that specific product. Or show in general the latest sales. It has more impact if you show the sales of users geographically close to the user who is viewing the site.
Stock number: show the number of pieces in stock of the object in question.
The Personalization Platform
Everything written above doesn't make sense if you don't have one (or more) eCommerce personalization platform to implement the described techniques.
The purpose of this article is not to analyze the various existing personalization platforms. I also would like to thank some fellow Redditors that helped me find some examples in this article.
I am currently analyzing 169 companies -yes, you read correctly- in the field of personalization. When I'm done, I may publish another article with the results of my research.