# Dreamdata: attribution models, reporting history and exports

Assess Dreamdata attribution models, reporting changes, warehouse access and migration before choosing.

Canonical page: https://quillwell.com/guides/software/dreamdata/
Updated 2026-10-31
Publisher: Quillwell Guides (https://quillwell.com/guides/about/). Editorial standards: https://quillwell.com/guides/editorial/

## The short answer

- Dreamdata is worth evaluating when your team needs an explainable credit rule for recorded journeys.
- [Dreamdata’s model overview](https://docs.dreamdata.io/article/tycc6odb2v-attribution-model) describes six models and Enterprise custom models.
- Dreamdata deletes disconnected source data after 30 days; arrange an independent archive first.
- Compare HockeyStack if its Snowflake export route fits your warehouse workflow.

## Choose the question and the model

Choose a model only after deciding what the report must answer. Credit allocation describes recorded touches; it cannot by itself show the causal effect of a campaign.

[Dreamdata’s model overview](https://docs.dreamdata.io/article/tycc6odb2v-attribution-model) describes six models with different credit rules, and custom models for Enterprise. Choose a rule your team can explain and keep it consistent when comparing periods.

[Dreamdata’s influence guide](https://docs.dreamdata.io/article/hgzsb56i8k-influenced-vs-attributable-leads-and-value) says a lead can appear in the influenced totals of multiple campaigns. Attributed leads divide credit according to the selected model. For a lead touched by two campaigns, adding influence counts can count the same lead twice.

Keep influence and allocated credit separate in the report your team reviews. If the decision requires incremental lift, design a causal test rather than treating either measure as the answer.

## Inspect the journey behind the report

Trace a known journey from source events to the account and outcome. A missing touch cannot receive credit, even when the model is configured correctly.

[Dreamdata’s LinkedIn Ads guide](https://docs.dreamdata.io/article/dx494qbfte-linked-in-ads-integration) describes company engagement in aggregated rolling windows. Low-volume metrics may be suppressed, and historic event dates may be approximated. Treat this company-level engagement differently from a precisely observed person-level exposure.

For a trial, include an anonymous visit and a later identified contact at the same account. Record what each source actually supplied, then compare it with the journey shown in the report. Trace an offline call through the CRM as a separate handoff.

[Dreamdata’s modelling schedule](https://docs.dreamdata.io/article/rnkjup6a7v-data-modelling-schedule) says debugger events can appear within minutes while reports may take up to 24 hours. Warehouse updates follow modelling. Agree on a reporting cutoff before comparing snapshots.

## Preserve a closed-period record

Save a report snapshot with its settings when the period closes. That gives the team a reference if a later report changes.

[Dreamdata’s history guide](https://docs.dreamdata.io/article/22uclfrmon-understanding-changes-in-historic-reporting-of-attribution) says anonymous events that become identified can change historic attribution. Event configuration changes can also alter the result. Save the model, date range and CRM stage with the snapshot.

For example, a visit that was anonymous at close may later be associated with a known contact. Compare the later report with the saved version and investigate the identity change before interpreting a difference as a change in campaign performance.

Assign an owner for the archive and for explaining changes. Retain underlying events where your agreement permits, so a reviewer can investigate the cause rather than relying on a chart alone.

## Plan access before disconnecting

Treat the shared warehouse dataset and your retained archive as separate assets. Test access and copying before a source is disconnected.

[Dreamdata’s retention policy](https://docs.dreamdata.io/article/daz3uwwaiu-data-retention) says disconnected source data is deleted after 30 days. It also says deletion in a source platform is reflected in the app the next day. These rules do not promise access after cancellation.

[Dreamdata’s BigQuery V2 guide](https://docs.dreamdata.io/article/igio0e36f0-bigquery-nev) describes an add-on for Team, Business and Enterprise. It shares a dataset into a customer project and requires project permissions and named query identities. A shared dataset is not automatically an independently retained backup.

[Dreamdata’s migration guide](https://docs.dreamdata.io/article/8wspwy9h7o-migration-of-historical-web-tracking-data) describes importing HockeyStack history through exported CSV in storage. Reconcile fields and dates at the destination before retiring the old account. Obtain the plan, add-on and retention terms for your own setup.

## How to compare the alternatives

Consider HockeyStack when its configured Snowflake export of raw actions or stage data matches your warehouse workflow. Its [Snowflake guide](https://docs.hockeystack.com/integrations/analytics-and-data-warehouse/snowflake) describes a historical backfill followed typically by daily export; check whether the fields meet your archive requirements.

[HockeyStack’s tracking guide](https://docs.hockeystack.com/technical-details/tracking/how-does-tracking-work) says declining cookies disables website tracking, even though its default identifiers do not use cookies. Include both consent outcomes when comparing journeys.

Consider [Ruler Analytics](https://www.ruleranalytics.com/pricing) when pricing based on monthly visits fits your forecast. Its indicative pricing also depends on product, data and integration requirements, so request a quote for your workload.

The [Dreamdata vs HockeyStack comparison](https://quillwell.com/guides/compare/dreamdata-vs-hockeystack/) gives both products the same journey and exit test. Use the [marketing attribution buying guide](https://quillwell.com/guides/marketing-attribution-software/) to choose the measurement question first.

## Common questions

### Does Dreamdata attribution prove incremental lift?

No. A model allocates credit to recorded touches. Use a causal test to assess incremental lift.

### Why can a Dreamdata report change after close?

[Dreamdata’s history guide](https://docs.dreamdata.io/article/22uclfrmon-understanding-changes-in-historic-reporting-of-attribution) says newly identified anonymous events or event configuration changes can alter historical attribution. Save a snapshot with its settings.

### Does Dreamdata retain disconnected source data?

[Dreamdata’s retention policy](https://docs.dreamdata.io/article/daz3uwwaiu-data-retention) says source data is deleted 30 days after disconnection. Preserve the required history beforehand.

### Is Dreamdata the best attribution tool?

Consider Dreamdata if its documented credit models and BigQuery add-on fit your decision and archive requirements. Compare the same journey before committing.

## Sources and scope

We read the linked vendor documentation on October 29, 2026.

We have not run the products or measured attribution accuracy, support or migration.

## Sources

11 vendor pages, checked October 29, 2026

[Dreamdata: Overview of Attribution Models](https://docs.dreamdata.io/article/tycc6odb2v-attribution-model)
Checked 2026-10-29. Attribution model rules and Enterprise custom models.

[Dreamdata: Influenced vs attributed leads and value](https://docs.dreamdata.io/article/hgzsb56i8k-influenced-vs-attributable-leads-and-value)
Checked 2026-10-29. Influenced leads and attribution credit.

[Dreamdata: Setting up LinkedIn Ads](https://docs.dreamdata.io/article/dx494qbfte-linked-in-ads-integration)
Checked 2026-10-29. LinkedIn company engagement and its historical limits.

[Dreamdata: Data Modelling Schedule](https://docs.dreamdata.io/article/rnkjup6a7v-data-modelling-schedule)
Checked 2026-10-29. Event visibility and modelling schedule.

[Dreamdata: Understanding changes in historic reporting](https://docs.dreamdata.io/article/22uclfrmon-understanding-changes-in-historic-reporting-of-attribution)
Checked 2026-10-29. Reasons historical attribution changes.

[Dreamdata: Data retention and deletion](https://docs.dreamdata.io/article/daz3uwwaiu-data-retention)
Checked 2026-10-29. Source disconnection and deletion.

[Dreamdata: Google BigQuery V2](https://docs.dreamdata.io/article/igio0e36f0-bigquery-nev)
Checked 2026-10-29. BigQuery add-on and dataset access.

[Dreamdata: Import historical web tracking](https://docs.dreamdata.io/article/8wspwy9h7o-migration-of-historical-web-tracking-data)
Checked 2026-10-29. Routes for importing historical web tracking.

[HockeyStack: Snowflake integration](https://docs.hockeystack.com/integrations/analytics-and-data-warehouse/snowflake)
Checked 2026-10-29. Configured Snowflake exports and backfill.

[HockeyStack: How Website Tracking Works](https://docs.hockeystack.com/technical-details/tracking/how-does-tracking-work)
Checked 2026-10-29. Website tracking and the effect of declining cookies.

[Ruler Analytics: Pricing](https://www.ruleranalytics.com/pricing)
Checked 2026-10-29. Indicative pricing and visit tiers.

[Suggest a correction or send a vendor fact update](https://quillwell.com/guides/request/?type=correction&page=/software/dreamdata/)