We know your customer

It is our mission to recognize the needs of your customers and enabling online retailers to personalize every single shopping experience

Request a demo Learn more
We help your customers to find what they are looking for

We help your customers to find what they are searching for and give them personalized incentives to maximize your conversions and revenue

Request a demo Learn more
We personalize your customer communication

Maximize your revenue through automatically sent and personalized notifications and enrich your newsletters with personalized product recommendations

Request a demo Learn more
13 February 2015

Don’t Take it Personally!

Personalization is a topic with a lot of depth. It's something everybody speaks about, but many people have yet to really get it right. Personalization comes involves a variety of use cases, but it has a tendency to be overly simplified by customers and prospects. We often see requirements on RFPs that simply ask for a CMS that “needs to have the ability to provide a personalized user experience.” This approach is vague and essentially useless.  

As a potential CMS buyer, it’s important to understand your needs in order to chose a solution that aligns with your business goals.

Below are four of the most common personalization use cases:

Customizable user interface

Typically, we see this required for portal-type sites that provide information in different formats with the ability to enable the visitor to modify the content shown. For example, a site visitor may only wish see sports articles and videos on a newspapers front page versus all content. This type of personalization is rather simple to implement.

Permissions-based personalization

Permissions or membership-based personalization is quite common. The two main scenarios can be divided into:

  • Intranet or extranet (partner web) access to content based on levels of permission
  • Premium content or services for paid content in media, and access to premium services based on group membership

This type of personalization is usually directly connected to a business goal and can be clearly understood. This type of personalization can be used in conjunction with gamification. For example, users may receive extra access if they fill out more information in their user profile.

Contextual content delivery

Contextual content delivery is connected to editorially-controlled personalization. It can be used to make the frontpage campaign different depending on the user context. For instance, personalization can be driven by location, demographics, or even visits from a certain device. In this situation, you can focus on promoting a local event based on where visitors are located. Another case includes returning visitors from an email campaign. These visitors can experience personalized landing pages, whereas first time visitors will see something completely different.

The more you dive into contextual content delivery as a form of personalization the more you’ll realise that there is a lot of inherent complexity and several requirements for a CMS platform.

Note: Learn more about adaptive and contextually aware content.

Behavior-driven personalization

Editorially-driven personalization can be beneficial, but it’s hard to pull off due to its complexity. Yet, there remains an expectation of web visitors today to receive a personalized experience when visiting major sites.

For example:

  • YouTube, Netflix, Apple TV: Recommended videos
  • eBay, Amazon: Products you may like

Visitors have been trained to look for recommendations. It’s expected. If you deliver a service within e-commerce or media, you must be able to deliver targeted content.

Recommendations can be delivered in different ways:

  • Trending recommendations: Today’s top articles or best selling products
  • Semantically-connected recommendations: Related content within the same category, tags, topics, companies, products or people
  • User-behavior based recommendations: People who bought one product also bought a related product
  • KPI driven recommendations: A product, matching a defined KPI, that is most likely to convert given the users behavior

Most media sites use trending content, typically based on simple statistics, to show you what’s popular right now. More advanced engines also use predictive analysis for providing more relevant content that is most likely to trend in the coming hours.

Semantically connected recommendations are nothing new, but it requires good meta information and linking of content. For example, this is often used in combination with trending content to show the most popular accessories sold with a TV. At this point it starts to become relevant to the user.

User-behavior based personalization is based on tracking user behaviour. You can define events to track articles read, social media shares or product purchases. This information can be used to classify users with preferences. Visitors interested in sports as opposed to celebrity news would be placed in separate preferences. Based on the preferences, content recommendations can deliver more targeted content based on what similar visitors have consumed.

And here is where it becomes truly relevant to the business. Most sites are commercially driven and have clear KPIs (key performance indicators). Linking recommended content or products to KPIs is important to the bottom-line. KPI-driven recommendations are powerful because they are based on self-learning algorithms that connect user behavior with predefined KPIs. You can achieve content recommendations based on what your users are likely interested in -- and those which are most likely to convert.

A conversion could be:

  • Reading an article or watching a video
  • Sharing an article on social media
  • Purchasing a product
  • Signing up for a free trial or demo

E-commerce vendors can also recommend products based on:

  • Products that are in stock
  • Products with high margin

Given the depth of the topic, personalization is often taken too lightly by customers and prospects. If you are looking to implement personalization, make sure that you are clear on the user scenario you would like to have solved and look for that capability in the CMS.

Here are a few helpful tips in you quest for deeper personalization:

  • Don't stick with simplistic standard requirements when it comes to personalization. Instead dive into detail and figure out what you really mean by "personalization.” Of course, not all possible scenarios are listed above, so try to elaborate on them and be as specific as you can when searching for a CMS.
  • Illustrate and specify what you are looking for with examples. Using personas or user stories will help you better understand your requirements when it comes to personalization.
  • Think about your goal first and foremost -- before even mentioning your idea for how to achieve it through recommendation. There is a chance that others may have already achieved what you’ve been trying to do in a simpler way.
  • Always keep in mind the workload that personalization can generate for your team in the backend. If you’re not careful, it can quickly become costly.
Written by Bård Farstad, Co-founder and Chairman of Product Innovation Board at eZ Systems
blog comments powered by Disqus