How to build an effective personalization strategy for publishers

Content personalization (or simply ‘personalization’) is an approach wherein a publisher tailors its content to the preferences of a specific audience. This makes content more relevant to a visitor, and matches a visitor’s personal interests, needs and behavior. The purpose of personalization is to help visitors find the desired content faster and provide companies a better chance at understanding their audience and converting visitors into loyal customers. With personalization it’s likely that your engagement score will increase, and the time visitors spend on your website will increase as well.

For instance, US publisher Business Insider was able to increase its website’s clickthrough rate by 60% based on personalization. The publisher used behavioral data to develop user profiles, based on visitor’s previous interactions with the site. The New York Times personalized content by using the visitor’s past behavior to make the most relevant content recommendations, including a highly relevant ‘Recommendations Page’ link that contained even more personalized content for specific audiences.  

When implementing a personalization strategy, it’s wise to first learn more about your possibilities. In content personalization, we distinguish three types of personalization: content targeting, recommendations, and notifications.  

Content targeting means adjusting the content on your website to meet the preferences of your (segmented) audiences. You can tailor headers, banners, CTAs, overlays: everything that we can label as content is suitable for targeting. As a publisher, content targeting includes tailoring your website’s header to the interest of your reader. For example, a reader that is interested in politics will see a landing page header that shows the most relevant political news. Whereas a reader that is interested in economics, sees the latest financial updates. Personalizing content such as headers, banners and CTAs increases engagement on your platform: this minor effort can already lead to a 5-10% increase in response rates. 

Recommendations are used within personalization strategies to show relevant content based on visitors' interest. There are several types of recommendations, such as popularity recommendations and related recommendations. Popularity recommendations is content labeled as popular due to the (high) number of impressions. For example, when you're scrolling on a news site, there is a category with ‘most viewed content’. Related recommendations are based on content that is labeled with the same category or topic (such as political news or pasta recipes). It’s like Netflix, but for your digital content. 

For example, when a user visits your website and watches a video, recommendations can guide them to watch another video based on their preferences. The power of recommendations is that it draws visitors in further. It’s also a way for visitors to efficiently deal with the information overload on the internet and helps them quickly find content they actually like. 

Personalized notifications are sent based on user behavior and notify your audience on relevant content. Notifications are a way to let your audience know that something new has happened, so they return to your platform. Examples are breaking news messages, in-app notifications and personalized newsletters. 

Publishers face various challenges when trying to successfully execute a personalization strategy. According to research conducted by Digiday, 68% of all publishers find it hard to execute a personalization strategy due to the lack of technical expertise. Aside from technical struggles, a lot of publishers lack personalization insights and believe the data that drives personalization is incomplete. 

But personalization can provide you with a competitive advantage: 42% of consumers are annoyed when content isn’t personalized and 72% of all customers even say they only engage with personalized messaging. Plainly put, visibly tailoring content to your audience's preferences is a big win for most publishers. 

So, where do you start when you want to actively personalize content for your audience? Here, we advise you to follow a few crucial steps, starting with implementing a proper technological solution: the core of your personalizations strategy.

In order to personalize the customer experience, you’ll need the right resources to do so. This includes a proper tech stack, one that allows you to capture data and process it for segmentation. There are multiple solutions when it comes to personalizing content for publishing goals, including a personalization engine to couple with your (headless) CMS or a headless CMS with built-in personalization. The latter combines the power of both content and data, and maximizes the impact of published content. Here, you can read more about the differences between these two approaches. Without the proper tools, even the best personalization strategy can fail to deliver results.

The next step is to always define personalization for your organization and think of your desired outcome. Your personalization strategy depends on premeditated goals, KPIs, and targets. If you want to increase time spent on your website, it’s wise to opt for recommendations. If you want to increase your conversion rates, personalized dynamic content is the way to go. Surely, you can combine these approaches but when starting out with personalization, less is more. 

All personalization starts with data. By collecting customer data, you can learn more about the preferences of your audience. A personalization engine and data-driven headless CMS both allow you to capture and evaluate user data. After collecting data you’re able to segment your audience around criteria that group their characteristics into having something in common. You can segment your audience based on demographics, age, gender, lifestyle, interest, educational level: you name it. This information is then used to provide a unique experience to your audience.

After segmenting your audience, design your first personalization experiment. This starts by describing the hypothesis. Informed by customer data focused on clear KPIs, a clear hypothesis helps you evaluate afterward. While there are several ways of writing down a hypothesis, it generally includes the cause and the effect. Make sure to define a threshold to validate or invalidate your hypothesis. This helps you assess the result of your efforts later. For example, your conversion must increase by at least 10 percent to succeed. Surely, your KPI or desired metric depends on your industry, your platform, and your initial personalization goal.

When you’ve formulated your hypothesis, it’s time to implement personalization and run your first experiment. Make sure to give your experiment enough time to produce relevant data and obtain a substantial sample size (the number of visitors included in your experiment). Without enough data, it’s hard to determine whether your efforts are successful or the changes in your performance are just random. The more traffic your platform gains, the faster your experiment goes.

As soon as your first experiment has gathered enough data, it’s time to measure and evaluate results. Did personalizing the header image to your segment indeed result in more conversions? By looking at your results, you’ll easily see if your personalized approach performed better – or not. Based on your results, it’s also possible to narrow down your segments even further.

Based on the steps mentioned above, you have a broad personalization strategy in place. Now, you can start implementing personalization. Keep in mind that this is an ongoing process, one where you continuously adjust segments to meet their specific needs. 

If you want to learn more about personalization, make sure to download our whitepaper ‘Personalization with a headless CMS’. Here, you’ll discover more about the technological solutions and best practices when it comes to personalization.