KDnuggets Home » News » 2018 » Jun » Opinions, Interviews » Analyzing Personalization Results ( 18:n26 )

Analyzing Personalization Results


The 4th part of this series will help answer the following questions: “Should I improve something or make changes to the system? Can it work more effectively? Can I squeeze the lion’s share of it?”



By Valentina Podmazina, IO technologies.

This is part 4 of this series, look here for parts 1, 2 and 3:
1. Introduction to Content Personalization
2. Audience Segmentation
3. Technical Content Personalization

We have come a long way from learning content personalization to exploring its backend mechanisms.

Creating a system and strategy is crucial for building an innovative business providing clients with a unique user experience.

But there is another important task left to do… Analyzing the results.

This analysis will answer the questions: "Should I improve something or make changes to the system? Can it work more effectively? Can I squeeze the lion’s share of it?".

Follow this brief guide and you will learn how to improve your business using data analytics tools.

  1. First, ask the right questions.

    • How do you find out whether people enjoyed your content?
    • Have readers visited other articles on your website?
    • How long did it take for visitors to read one article?
    • Haven’t readers just looked through it?
    • What links have the highest click rate?
    • What can be improved to make your readers more addicted?
    • What are the visitors’ reasons for leaving the page without any actions?

    One universal answer to all of them is: "Use data and data analytics tools". It may sound complex, maybe even ambiguous, but data got from editorial analytics will lead you towards your final goal step by step.

  2. IO Technologies analysing questions about content personalisation

  3. Watch Charts

    To organize several metrics in one place check charts in editorial analytics that will give you a visual understanding of how articles perform throughout the day.

    Every analytics system (Google Analytics, IO dashboard, etc) shows the number of visitors sorted by dates. Therefore, charts and common data analytics tools give us a general view of visitors’ behavior and statistics. For deeper analysis, special metrics are needed.

    I want to tell you about which special metrics we use at IO. However, if you work with any other analytics system, the majority of metrics will be the same as they are crucial for any media business.

  4. IO Technologies watching charts and using analytical metrics

  5. Use Metrics

    • Pageviews. The total number of pages viewed during the selected time period.
    • Unique visitors. The number of uniques who visited your site during the selected period.
    • Click through rate (CTR). This is the number of clicks divided by the number of page views.
    • Depth. The number of pages users go through on average. This number indicates how interesting the website’s content is to users, making them stay and click on links.
    • Finished reading. The percentage of visitors who scrolled the text to the end.
    • Recirculation rate (opposite to Google Analytics’ Bounce rate). The percentage of users who read one article and then proceeded to another website page, without leaving the site.
    • Social actions. The total number of shares on social networks, i.e. reactions, comments and shares on Facebook, Twitter, etc.

    Every newsroom has its primary metric in their editorial analytics. For example, our clients CN Group focuses on their recirculation rate, and Express.co.uk is more focused on CTR. A detailed analysis of separate factors like CTR or recirculation helps you make the content personalized, as data allows businesses to analyze people's behavior deeper.

  6. Data-driven desicion making on audience segmentation using data

  7. Make decisions

    This step requires a smart attitude and the ability to make quick decisions. You don’t need to be an expert analyst and use professional data analytics tools to make high-quality professional conclusions.

    You can always find out what’s wrong with poorly performing articles and fix them. Here are several examples:

    • CTR too low? Try to experiment with headlines, maybe they are unclear and ambiguous, people don’t get what your material is about.
    • Depth stats leave you frustrated? Add tips & tricks to your articles. Share practical advice and experience.
    • Could recirculation rate be better? Insert more links into articles, try doing this in different parts of the text.
    • No one finishes reading? Make your texts juicier and leave the most interesting things for the end. In the first couple of paragraphs let readers know that there is a bonus (a useful tip) waiting for them in the end.

Wrapping Up Content Personalization Topic

Now, it’s time to pick out the essence and see what was vital for every data-driven client-oriented business.

Personalized content is what people want to see on web pages and what makes them engaged. You have to follow this trend or you will be overtaken by a huge number of competitors.

Original. Reposted with permission.

Related:


Sign Up