Salesforce Marketing Intelligence Services

MI supports standard attribution models: first touch, last touch, linear, time decay and U-shape for touch-based attribution, plus first and last touch for funnel-based attribution. You can view several models side by side to compare how each one credits your campaigns.

Averaging several models into one blended model isn’t a standard MI option. To maintain that approach, the blend would need to be calculated outside MI, for example, in Data 360 or your existing Snowflake setup, with the result then shown in MI. Alternatively, you can choose one standard model as your primary view and retain the others for comparison.

MI attribution can use custom engagement signals (touchpoints), allowing you to decide which interactions count as an attribution touch.

MI doesn’t have a built-in setting that merges several sign-ups within a short period into one touch. This cleanup would happen in Data 360 before the data reaches MI. For example, sign-ups within a defined window (such as seven days) could be grouped into one touchpoint. Clear lifecycle stage rules in the CRM can also prevent repeat sign-ups from pushing contacts back through funnel stages.

MI can take over attribution reporting and dashboards, becoming the main reporting view over time.

Connecting MI to Snowflake data goes through Data 360’s zero-copy integration, which reads data where it resides without requiring a new pipeline. This means Data 360 is also needed.

If the blended attribution logic remains in Snowflake, MI would sit alongside it rather than replace it. We’d recommend running both dashboards in parallel for about one reporting cycle, checking that the numbers match, and then deciding whether to retire the Snowflake dashboard.

MI has two relevant features, both currently in beta:

Goals: Set KPI targets (such as cost per acquisition or sign-up volume) and track which campaigns are falling behind their targets. This is the closest match to the functionality they currently like in ActiveCampaign.

Paid Media Optimization Agent: Reviews Google, Meta and LinkedIn ad performance, recommends which ads to pause and can pause them directly.

Neither feature is generally available yet, and both operate under Salesforce’s beta terms. The agent also works directly in live ad accounts and requires additional Einstein and Data 360 credits.

We’d recommend leading with Goals and cross-channel dashboards, positioning the agent as a later step once reporting is in place.

MI has dedicated Google Ads and Meta Ads connectors that feed into one shared data model. Both platforms can sit side by side in one dashboard with CPC, conversion rate and ROAS for each. MI also incorporates CRM data, allowing ROAS to be measured against closed revenue as well as each platform’s reported conversions.