Attribution

Attribution helps teams understand which campaigns, channels, partners, and touchpoints contributed to a conversion. In DriveMetaData, attribution should be treated as a decision framework, not just a dashboard number.

What attribution answers

Use attribution to answer questions such as:

• Which channels are driving qualified users, not just clicks?
• Which campaigns influence conversion across mobile, web, and CTV?
• How do first-touch, last-touch, linear, time decay, or data-driven views change the story?
• Which ad partners show suspicious conversion patterns?
• Where should budget be increased, paused, or reviewed?

Touchpoints and conversions

A touchpoint is a measurable interaction before a conversion. Examples include:

• Paid ad click.
• Impression where view-through measurement is supported.
• Email click.
• Landing page visit.
• App install.
• Deep link open.
• CTV campaign exposure.

A conversion is the business outcome being measured, such as signup, purchase, subscription, booking, lead qualification, or renewal.

Attribution windows

An attribution window defines how far back DriveMetaData should look for eligible touchpoints. Common windows include one day, seven days, thirty days, or a custom period aligned to the buyer journey.

Shorter windows can reduce noise for fast decisions. Longer windows can better represent considered purchases or multi-step onboarding flows.

Model selection

Model	Best for	Watch out for
First-touch	Understanding initial discovery	Can undervalue later nurture
Last-touch	Short-cycle conversion optimization	Can overvalue closing channels
Linear	Giving all eligible touchpoints equal credit	Can dilute meaningful influence
Time decay	Journeys where recent actions matter more	Requires careful window selection
Data-driven	Pattern-based insight when enough data exists	Needs stable, high-quality data

tip

Compare models before changing budget. If every model tells the same story, confidence is higher. If models disagree, review the journey before taking action.

Privacy-first measurement

Privacy-first attribution means measurement should respect user consent, platform policies, and regional requirements. In practice, that usually means:

• Prefer first-party and consent-aware identifiers.
• Avoid depending on third-party cookies as the only matching method.
• Keep attribution logic documented and reviewable.
• Apply suppression and eligibility rules before activation.
• Use aggregated or modeled reporting where user-level data is restricted.

Common discrepancies

Attribution numbers can differ from ad platform reports because of:

• Different attribution windows.
• Different conversion timestamps.
• Platform self-attribution.
• Duplicate or missing events.
• Refund, cancellation, or offline conversion timing.
• Timezone and currency differences.
• Fraud filtering that removes suspicious activity.

When numbers do not match, choose a source of truth for each KPI and document the reason.

Recommended review cadence

• Daily: campaign spend, major conversion movement, fraud alerts.
• Weekly: ROAS, CAC, funnel conversion, partner quality, model comparison.
• Monthly: attribution window review, budget shift analysis, LTV and retention impact.