Data connect session
We require a view of your Data Warehouse. You will find all the information you will need here: documentation. If you are using BigQuery, we have a designated data connector app. When the view is ready on your end, our engineers will sync the data.
To make sure the data is understandable for our models, we provide you with a data tagger tool to map and tag events in the data.
How to prepare for this step: decide on what data you should share and how. We will send you a document with key things to consider.
Send pre-onboarding checklist to customer
Preliminary analysis
Data Science team validate the data and let you know if there’s any mismatches or concerns
We will need you to mark event as a key event using a simple guide
Inital config and deploy standard pipeline
Data science: initial config and indentifiers validatation
Data engineering: data pipeline,MMP/Ad platform integrations with access tokens
Experiment setup
We are now sending our predictions to your ad platform directly or via the MMP.
Final config and data monitoring
Data science: final config
Data engineering: data monitoring, anomaly detection, and infoviz and platform access
Opportunity sizing presentation
We present our initial predictions for you, agree on the key metrics and decide on next steps.
Signal check / Attribution loss measurement
Data science: Monitor post conversion sending, account for losses of signal
Campaign start
Here we monitor the performance of the campaign, observing among others the initial conversion volume, learning phase, spend and volume
Monitors events and readjusts thresholds
Data science: event monitoring in the data platform and readjust thresholds