Microsoft

Sitecore Profile Decay Rate – How It Affects Personalization

Behavioral profiling, for most marketers, is the ultimate digital goal when it comes to their customer experience solution. Not surprisingly, it is nearly always the first question I am asked when working with an organization that is either using, upgrading, or considering Sitecore CXP. “How do we personalize?” or “How can we get started with behavioral personalization?” often lead the discussion, even though many important best-practice steps are necessary to get to that point.

For this post, however, I want to get specific and do what is often asked. I want to jump ahead and talk a little about a specific, important setting when it comes to personalization: decay rate.

Before I dive completely in, please let me cover off on the high-level steps in the form of assumptions:

  1. You are on or considering Sitecore 8.1 or later with xDB.
  2. It is or will be optimized for best-practice marketing use not just as a CMS (ie: built with a solid component architecture)
  3. You have some level of comfort setting up goals, campaigns, taxonomy, testing, and profiles.

If you want to gain more comfort with items from number three above, please consider my video series on YouTube to help guide you through this crawl, walk, run, fly process.
My focus for the rest of this post will live completely in the Profile space in Sitecore’s Marketing Control Panel. First off, if you are starting with the out of box Profiles from Sitecore that include Focus, Function, Score, etc. please delete them. While they may work for some, most organizations should and probably DO start with Personas and will build other more suitable profiles from there.
Sitecore Profile Decay Rate
My hope is that you and your marketing team have already created a core set of personas to help guide content on your website. With that, you will want to create a new parent Profile as “Marketing Personas” or “[Acme.com] Personas” – something that speaks to the core use of this group. Within that Profile you will define a set of keys that represent behaviors and/or emotional states. These keys will be shared by both the Profile Cards and the Pattern Cards to respectively 1. Define the user’s pattern across pages, assets, goals and campaigns AND 2. Use that pattern to drive content in a component based on the pattern rule that matches what has been picked up.
This is a process that takes getting used to, and if you consider the content multiplier that comes with multiple personas per pre-defined variable component – I highly suggest you start simple and grow from there.
All of the set up leads me to the focus of this post. Since Sitecore version 8.1 we have seen the powerful addition of being able to track Profile building across multiple sessions. Very powerful indeed, since each user profile is not starting from square one per visit. For marketers, this means, if it was determined in a previous session that your user identified as (say) an “Explorer” persona – when they return for a future session, they will be made to feel like you know they are an Explorer by a home page banner that speaks to them. (infinite variations on this example and Profile/Persona names, of course).
Hold on though…we are human beings filled with emotional states and a world of influencers that may make us an Explorer one moment but perhaps more of a “Home Body” the next time we visit. If Sitecore is going to carry the profile/pattern across visits, we may never see any of your other products or services. We could get caught in a maddening Explorer loop. This is where Decay Rate comes in.
At the parent Profile level (ie [Acme.com] Personas) you can apply a percentage-based decay rate that is factored in session to session, reducing the potency of the previous profile(s) session. As an example: A user that has earned an Explorer profile key score of 100, that has an applied decay rate of 60%, upon their return (new session), they will enter as an Explorer key score of 40. This could then allow Sitecore to move them to a Home Body quicker, based on profile activity in a new user behavioral state. The result is a customer that feels more connected to your content and website because your site is simply a more balanced “listener.”
It is also important to note where this Decay Rate is located in the hierarchy of the Profile setup. Being at the top level of each Profile type, the cards contained within will all share the same decay rate. Personas, for example, would not likely share the same decay rate as a profile group for Experience Level. To play that out, if I am an Explorer or a Home Body and those persona traits can evolve over time and session to session at one rate – it would be likely that my experience level with your product, based on media downloaded, purchases made, etc., would continue on a more stable track upward. Using my favorite sport as an example: As a cyclist I can find myself shopping for mountain bike or road bike components on a regular basis, but as I continue to ride, purchase and read blog posts from a preferred store, my expertise in cycling as a whole keeps improving. My decay rate for bike type, should not be at the same rate as my decay for my cycling expertise as a whole – expertise should decay slower and content should continue to encourage my reading habits for both bike types.
As I said before, start simple and expand, but once you start refining your decay rate, you will see more accurate contextual personalization and ultimately better conversion. Please reach out to me on Twitter @Sitecoredial with any questions or comments.

Thoughts on “Sitecore Profile Decay Rate – How It Affects Personalization”

  1. Hi Rick, nice post, very clear description of the decay rate! I\’m implementing it now on a client\’s acceptance environment and I suppose it should work, but do you have some pointers on how to test this? I\’m testing it by browsing the site, viewing some pages that increase profile points, then I close my browser tab and revisit the site, but the decay rate doesn\’t seem to be applied and my profile based personalization scenario stays active even after closing and opening new browser windows a couple of times. So if you would have any pointers on that, that would be great!

  2. You are on the right approach for testing. Just make sure you have set your decay rate high to something like 95 and that you are working with profiles with enough \”room\” in the scale 0-10 for example to be able to identify easily that the decay is taking place. If you feel it is not, work with your developer or dev partner to make sure it is functioning. They should be able to review the raw data in the logfile.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Rick Bauer, Senior Solutions Architect

As a 5-time Sitecore Strategy MVP and Senior Solutions Architect at Perficient, Rick provides insight, training, and a passion for a fully-realized potential of the Sitecore Customer Experience Platform and Content Hub. He utilizes his years of hands-on Sitecore know-how and certification to help deliver clear solutions and actionable results for marketers - both new to the platforms, or looking to take them to the next level of maturity.

More from this Author

Subscribe to the Weekly Blog Digest:

Sign Up
Categories
Follow Us
TwitterLinkedinFacebookYoutubeInstagram