Predictive content in Marketo Engage refers to a feature that uses machine learning algorithms to dynamically personalize and recommend content to individuals based on their behavior, preferences, and characteristics. The goal is to enhance engagement and improve the chances of conversion by delivering more relevant and timely content to each user.
Automatically create content recommendations for each person based on their past interactions and profile data.
Personalize the user experience by dynamically displaying content that aligns with each stage in the buyer’s journey.
A/B test to compare the performance of different content variations helping to optimize content effectiveness.
Insights and analytics on the performance of the content marketing efforts.
If you’re looking to drive higher engagement metrics, launching Predictive Content can help!
How does Predictive Content work?
Using artificial intelligence (AI) Marketo Engage will work on ongoing 3-step process on discovering, deploying and analysing the success of your content.
Auto-discover content
No need to dig through your content library. Marketo is able to auto-identify content types. You choose which content is “allowed” to be recommended to your audience.
Learn and engage
Leveraging machine learning and the power of Adobe’s Artificial Intelligence (AI) the algorithm will automatically show enabled top performing content, to the right people.
Analyse
You’ll gather insights to understand engagement or conversions. Helping you close the loop on content ROI and supporting your decision on next steps.
Example use cases
Email
Rich media
💡 Consider using predictive content on the confirmation email for webinars or events as these typically have a 60% open rate!
Setting up Marketo Engage Predictive Content
Ready to boost your email performance and content return on investment? Here is a topline guide to launching predictive content in Marketo Engage:
Access Predictive Content in Marketo: If this is available in your subscription you will find the Predictive Content tile on the Marketo home screen. If you don’t see it, you will need to contact your customer success manager to enable it for you.
Choose your Predictive Content goal: You will decide which algorithm you want to maximise:
Clicks: Display the content most likely to make the person viewing the content click on it
Conversions: Display the content most likely to make the person viewing the content submit a form
Configure your content patterns & categories: Predictive Content requires two key components to work, inclusion of specific URL structures (i.e. jtfmarketing.co/insights, or jtfmarketing.co/resources) and content categories (i.e. blog, case study, download, webinar). This allows the Predictive Content engine to differentiate the types and report correctly.
Deploy the predictive content javascript: Unfortunately this is a specific javascript for predictive content, we recommend only deploying it where you want to auto-discover content i.e. case studies, events, and blogs.
Discover and add content: There are two ways to add content either by waiting around for cookied visitors to click on your content types and “auto-discover” (yes sadly not a scanner) or by clicking all content you want to appear as a cookied user. Either way only discovered content can be enabled in Predictive content.
Optimising & approving content for Predictive Content: Once your content has been discovered you can quickly optimise the content look and feel, summary content and enable it for specific channels i.e. Rich media, web bar or email.
Deploy predictive content modules: You will then need to deploy the predictive modules to either your web property or email campaigns. Enabled content will appear at random based on the top performing content.
Review & optimise using analytics: As with everything automation, it’s not a one-time project, we must revisit, review results and optimise the content to ensure it’s performing as expected. With a limitation of 25 pieces of content included, you may want keep swapping content around until you find what works best for your business.
What are the limitations of Predictive Content in Marketo Engage?
You can only choose one algorithm for everything. It’s clicks or conversions. Our recommendation is use clicks, and track your conversions within program templates (link to program templates).
The standard licence includes only 25 pieces of content. If you need more you can add-on additional tracked content at an extra cost.
Deployment is limited to websites and Marketo Engage emails. If you’re running multi-channel campaigns you will only be able to maximise it on page, or within email.
Character limits apply everywhere. If you thought it was tough enough on social media, try Predictive Content, the limits are:
Content Name: 255 Chars
Bar Title: 110 Chars
Rich Media Title: 50 Chars
Rich Media Description: 215 Chars
Email Title: 50 Chars
Button Label: 17 Chars
Tokens cannot be combined with Predictive Content. If you’re looking to use tokens to personalise further, you’ll need to consider Velocity Scripting (and not using Predictive Content!)
Styling. If you were not already frustrated by the out of the box email templates in Marketo (link to em temps), then this will get you heated up. Styling is limited when using Predictive Content.
Auto-discovery of content. It’s slow and might require someone’s time to quickly launch a proof of concent.
What does the reporting look like?
Predictive content is limited to the content that has been auto-discovered, and enabled. You will be able to see the views, conversions and importantly where the content has been getting traction with your audience.
What next?
If you’re interested in exploring some use cases and launching a pilot using Predictive Content, get in touch – we’d love to connect you with one of our experts to provide some ideas!
