We have a very large site with a broad range of products to a large audience. Our customer segmentation; however, only has 5 main categories and we would like to further sub-categorize users on our site based on web activity.
Does it make sense to set up a smart lists that pulls data on web site visits to a page within a specific time frame and a constraint of having visited the site a certain amount of times to classify them as a particular type of customer? Basically we are trying to create a flow where if a customer visits a particular product or product family page more than 2 or so times within a specific time frame, we can then classify them as a particular type of customer and then funnel them into a nurture campaign for that product/product family.
I am trying to create the smart list with the lead database section, create a smart list where
Web Page is = XYZ
date of activity = in a time frame = This Year
Constraint of Min. Number of Times = 2
I want this to trigger a Role Change so I can classify the users as a particular type of customer - I'm having trouble with this step.
Is this the most effective /efficient way to sub-categorize customers for specific campaigns I am running?
Should I create sub segments and then trigger a segment change based on web activity parameters?
I would do a batch for this to run nightly. I suppose you could run a Segmentation too. But how do you determine which product stream gets priority?
That a good question, we are thinking the product with the most visits would have first priority. And from that point, if the customer chooses to view other content, we can begin to progressively profile to better target their more products. We also have RTP, so that might help us use the other product interest levels too... I'm thinking.
If you have a number of products and signals to track, I have found it much more robust to create separate scoring fields for each product you want to track.
You can then calculate a product score for each product or line of business. This provides a clean filter you can use in various places rather than replicating complex smart list filters. It also opens the door to look at a variety of signals beyond just web activity (email clicks, program success, etc.) for each product type.
I find it most sensible to consolidate your logic in a segmentation - one source of truth, easier to reference, and better performance than smart lists. Once you have a product level score you can then use these fields in your segmentations to determine the product-interest persona. Note that behavioral filters are not available in segment smart lists but the product-level score can become a proxy for this and it can be used in segment smart lists.
Now you can easily reference that segment everywhere and use it for nurturing, routing, content personalization, etc.
If you need to calculate which product-level score field is higher, you can do it in CRM and sync it back to Marketo or use a service like FlowBoost or the Excel webhook offered by Hoosh - check this thread for ideas:
Thank you Justin - this does sound like a good solution! I really appreciate your thorough feedback! Happy Friday.