Attribution-ready marketing data
Improvado prepares the data attribution runs on. It unions your ad, web, CRM and eCommerce sources, resolves them to one user and account key, and loads the result into your warehouse. It does not sell a standardised multi-touch model: the models are configured to your goals, your lookback window and whether you attribute at user or account level.
GDPR CCPA Trusted by enterprise marketing teams
What does Improvado actually do for attribution?
It prepares and joins the inputs, so the model you run reads one clean table instead of a pile of separate exports.
Union every source into one schema
Ad, web analytics, CRM and eCommerce data land in one flat structure rather than side by side.
Deduplicate and normalise
Duplicate records are removed and values, units and formats are standardised across sources.
Flatten and extract
Nested structures become flat tables and details such as UTM parameters are pulled out of the URL.
How do you get spend and pipeline onto the same row?
With resolved identity keys. An ID graph matches records across your sources and gives every person and every account one id to join on.
One id per user and per account
Identity resolution unifies records from each source into a single global user id and account id.
Keys your systems already carry
Email, client ids, contact, account, opportunity and deal ids are the relations the graph matches on.
Spend joined to pipeline
With the keys in place, campaign cost and CRM outcomes can be read from one joined table.
Account-level attribution
Touchpoints roll up to the account when the buyer is a group rather than a single person.
Do we have to install another tracking pixel?
No. The core runs on top of your data warehouse, which is why there is no dedicated pixel tracker to deploy on the site.
No dedicated pixel tracker
The attribution core reads warehouse data instead of a tracking script on your pages.
Built for complex stacks
Several CRMs, payment platforms and storefronts at once is the case it was designed for.
Hit-level input is required
Activity streams are the input, so report-level exports are too coarse to feed it.
“Today, we can finally build any report we want in minutes due to the vast number of data connectors and rich granularity provided by Improvado.”
Which attribution model will we end up running?
The one you define. The standard models are available and a model specific to your business can be implemented instead.
The standard models, configured
First touch, last touch, linear, position based and time decay, set to your goals and lookback window.
User level or account level
Attribute a single person's journey or the whole buying group, depending on how you sell.
Or a model of your own
A model built for your business can be implemented with the professional services team.
Where does the output go once it is joined?
Into your own warehouse and the tool you already model in. The output is a table you control, not a report you have to accept as it comes.
A table you control
Choose the metrics and dimensions the output carries instead of taking a fixed report.
Your storage, your BI tool
Load the result into the storage you already run and visualise it in your own dashboards.
Weights down to the keyword
Each touch carries its credit and its channel, tied back to the user or account id.
Want to see this on your own data?
Bring your sources and we will show you the joined, keyed dataset your model would read.