Klaviyo flows

[Ana]

Hello all,

Currently I am using Fivetran Built-in Transformations to get the Klaviyo models.

However, the transformed model klaviyo__flows only contains the overall values for all flows, not on date level (which is important for my use case).

As a workaround I created a model using klaviyo__events to extract the necessary metrics from each flow, per day. However, when using this transformed table I get much higher values/metrics compared to what we see in Klaviyo UI.

Does anyone had this problem before? Or have any idea on how to model Klaviyo data correctly?

Thanks!!
Ana

[avinash-kunnath]

Hi @Ana ! Thanks for bringing this to our attention.

I know I responded in dbt-slack, but will just reply here as well:

Would you be willing share the query that you are using to extract the metrics? We could potentially make a recommendation based off what you built and make a recommendation based off of what you built. 

One thing to call out is that klaviyo_events  model is at a different grain than klaviyo_flows.  and is using last touch attribution, which could lead to the overcounts. It might be worth using the stg_klaviyo_eventsmodel in our source package  and joining that in with klaviyo_flows data to get more accurate data. 

[Ana]

Hi Avinash,
Thanks!
I replied you in the slack channel 🙂