4 types of email personalisation

email personalisation

Email personalisation can take many forms, and in the most basic sense, it contains the name of the recipient in the subject line or body of the email. It could also take the form of recommended products based on browsing history or browsing websites tailored to customer interests.

As customer’s expectation of relevance continues to rise, 74% of consumers now report becoming frustrated when they receive irrelevant email content from a retailer.

For example, when you visit a retailer’s website, you want an experience that speaks to you as if the retailer was a loyal and reliable shopping advisor. And the reward for retailers who achieve this goal? 48% of consumers spend more when the experience is personalised.

There are many types of personalisation that email marketers can leverage to drive deeper engagement with shoppers. To help determine which type is best for your marketing organisation, here we have broken down four personalisation strategies and the benefits and challenges of each approach.

1. Segmentation

This is the most popular type of email personalisation and possibly the easiest. You can segment your database by age, gender, loyalty, geography, behaviour and much more. To scale your effort effectively, write newsletter content where 80% of the content is the same, but then create a custom version for each segment you have.

While segmentation can help you increase the relevance of your campaigns and increase overall engagement, the disadvantage is that this type of personalisation is limited to ensure your content is suited to the segment.

In addition, segmentation is resource-intensive and the number of segments that a marketer can use is limited, this limits your ability to optimise. And because email content is based a campaign production schedule, this limits your ability to deliver content in real time, limiting its relevance.

2. Predictive personalisation

Many retailers rely on predictive personalisation or predictive analytics. It is a static process of anticipating your customer’s wishes, for example predicting a next-best-offer. It can be a disjointed process because as trends, favourites, requirements, tastes change continuously, the same applies to your customer profiles. Engaging your shoppers with unique content based on their ever-changing needs calls for a method that embraces their ongoing tastes and interests.

3. Collaborative Filtering

Another way to personalise content is to align product recommendations with products the shopper already likes. For example, people who buy black shoes often also buy black socks. 

This technique, known as collaborative filtering, is an approach to product recommendations in which recommendations are made based on a user’s product interaction history combined with the interaction history of all other users on a site. Famously, Amazon.com uses Collaborative Filtering extensively.

A distinct advantage of collaborative filtering is its broad applicability; collaborative filtering algorithms don’t need to understand the customer’s tastes and interests, but simply the correlation between products.  Collaborative filtering is a ‘bottom of the marketing funnel’ technique that allows retailers to cross-sell and link-sell products to shoppers by presenting them with items they would not have necessarily sought to purchase.

4. Understand, Rank, & Learn (-URL) Personalisation

For ‘top of the marketing funnel’ marketing activities such as email marketing, retailers need a way to truly customise the customer experience that recognises the tastes and interests of each specific shopper. To be truly personalised, the content must be clearly directed to each individual’s unique interests through a personalisation strategy.


  • The practice of dividing a customer base into groups of individuals that are similar in specific ways relevant to marketing
  • Can be based on a variety of factors such as age, gender, interests and spending habits.


Best-practice examples:

Most brands already use some form of segmentation

  • Good at predicting something like the next best offer, churn probability.
  • Provides ‘accurate’ predictions.


Best-practice examples:

Many brands already use some form of Predictive Analytics.

  • Also known as ‘people who did X also did Y’.
  • Based on a product-to-product correlation.
  • The same for all customers who browse the same product
  • Good for cross-selling.


Best-practice examples:


  • Also known as Learning-To-Rank.
  • Provides fresh, personalised content that will delight customers every time they open a promotional email.
  • This kind of iterative style invigorates people who have ‘tuned-out’ repetitive messages and generates a steady baseline of higher-margin purchases.


Best-practice examples:

Spotify, Mercanto


To deliver genuine 1:1 personalisation that recognises the interests of each shopper, retailers must deliver content to a segment of one, optimised in real time, based on behaviours and preferences that are specific to each shopper.

With machine learning and artificial intelligence (AI) tools, you can instantly analyse all available content, as well as historical and real-time signals, delivering the best content for every shopper.

Mercanto uses machine learning to instantly analyse all available content, as well as live and historical signals, to deliver the best email experience for every shopper. Click here to learn more about personalisation based on machine learning.

Comments are closed.