We’re attempting to categorize our shoppers into buckets, such a discount shopper, high value customer, etc.

Our Klaviyio implementation receives events via Segment, which is embedded in our custom ecommerce platform. We want to use our product_viewed event, which has a supporting price attribute, to accomplish this task. We’d like to keep a running average of price across each of these product viewed events per profile and then associate these averages with various buckets, as mentioned above.

Are there any recommend ways of achieving this? We have about 2 million profiles we’re trying to keep up to date.

• We’ve considered running these calculations in our events database and running a script to update these averages as a custom attribute through the identify API, but the throughput is too low at 700 calls/min. By the time we update all our customers, that data is essentially out of date. We want more of a real-time view.
• Web feeds are too small to hold the data, it seems. So we have ruled that out as well.
• The logic for creating segments in Klaviyo does not appear to support calculations across events, so this does not seem to be an option either

Any help is appreciated.