Marketing analytics build
- Now Repository and data connections
- Open with you 4 items
A data analysis application, owned by Lingopie, that connects your marketing, analytics and subscription data and answers questions about what each channel actually produces in paying subscribers. Built in your GitHub, on your BigQuery, run on your own Claude accounts.
The plan
-
Access and setup
Accounts and data access in place. First queries verified against your BigQuery on 12 September, covering both reading your reporting data and writing to the operations dataset.
-
Repository and data connections
Repository skeleton and the BigQuery connection are written, and the connection is proven end to end. Warehouse reads for Google Ads, Meta and Customer.io, plus a Mixpanel connector, follow next.
-
First reports
Weekly and monthly scorecard on one basis, cost per trial beside cost per paying subscriber, blended CAC against the starting picture.
-
Analysis and first reallocation moves
Outlier and waste finders, Google Search read, platform truth check. First budget moves proposed for approval.
-
Opportunity list, budget allocation, handover
Ranked opportunities with measured effect, marginal cost curves, onboarding for your team.
Recently completed
-
First queries verified against your BigQuery, and the operations dataset proven writable
-
Survey of the warehouse completed, covering every dataset, to establish what the reports can be built from
-
Repository contents written and ready to land
-
Build plan and repository plan finalised
-
Access granted across BigQuery, GA4, Mixpanel, Customer.io and GitHub
-
Data dictionary received and adopted as the reference
What you will hold at the end of the build
-
The analysis application, live on your data, in your repository
In progress
-
Valuation report across campaigns, audiences, keywords, languages, landing pages and email
Not started
-
Ranked opportunity list, with what was executed and its measured effect
Not started
-
Written recommendation for what comes next
Not started
What we need from Lingopie
-
Confirm the three starting numbers (monthly paid media spend, monthly new paying subscribers, how CAC is calculated today) so the starting picture can be frozen
-
Name the approver for any live campaign or budget change
-
Confirm how Google Analytics data should reach the warehouse. The analytics export is not currently present in BigQuery, so the reports will use the Segment event data unless you would prefer the export switched on
-
Read access to any further datasets the cost methodology needs, once the exact list is confirmed
-
BigQuery job permission at project level
-
Mixpanel and Customer.io data region