Data sources
Connect the data behind your marketing decisions.
Connect ads, sales and your store in minutes. Nothing changes without your approval.
Mark the channels you spend on
Paid media
Organic social
Ecommerce and revenue
CRM and marketing automation
Storage and warehouse
Scheduled files, for any source with no API of its own.
Whatever your mix is, it models
of the eighteen connect themselves during the free trial.
You authorise the account, the spend history arrives, the model runs on it.
we set up with you, once you are inside.
Warehouse exports, retail media, CRM and ecommerce revenue.
TV, radio, direct mail and events go into the same model as paid search.
Custom data connectors, built for your model
A marketing mix model does not just need your data. It needs it at the right granularity. Spend by campaign and by day. The conversion window your business actually runs on. The splits that matter across your own accounts. A generic pipe moves a vendor's standard schema and leaves the model to cope with whatever arrives. A custom connector is built for what the model has to read, so the data is usable on the first run. That is why these are built rather than stocked. And you do not wait for one. The same data goes in as a scheduled file while the connector is being built.
Reads
18 sources
Google Ads
Meta Ads
TikTok Ads
LinkedIn Ads
Microsoft Ads
DV360
Criteo
Outbrain
Meta Organic
TikTok Organic
Shopify
WooCommerce
Klaviyo
ActiveCampaign
HubSpot
Salesforce
Azure Blob Storage
Changes, with your rules
Example
Monitor results
Automatic
Prepare new ads
Automatic
Move budget between campaigns
Your approval
Launch a campaign or a test
Your approval
Changes run in Meta, Google and TikTok.
Check the setup, data coverage and refresh schedule for each source.
Reading data and making changes are separate permissions. Confirm which actions are supported before enabling execution.
Data access informs the plan. Separate action permissions let Cassandra implement supported budget changes, with approval where required and a check of the result.
Reading data and making changes are separate. You decide which changes Cassandra can make.
Available actions depend on your connected workspace and permissions.
See how the result is checked →
Example execution
Illustrative
Budget plan prepared
Prepared
Approval received
Approved
Budget change applied
Implemented
Confirm the scope and permissions. Request approval where your workspace rules require it.
Compare the observed result with the saved prediction. Keep supported findings and open questions for the next decision.