MEASUREMENT METHODOLOGY

From Contribution to Causation: A Rigorous Ladder

We're precise about what we can claim. Here's our measurement methodology, the data requirements for each level, and what we don't do.

Default mode: Level 1 Contribution Accounting. Levels 2-3 require additional data or experimental design.

THE LADDER

Contribution → Causation

Each level builds on the previous. Higher levels require more data and coordination but yield stronger claims.

STATISTICAL RIGOR

Confidence Thresholds

Every receipt includes confidence intervals and significance testing. Here's how we think about thresholds.

DEFAULT THRESHOLDS

Confidence Level95%

False Positive Rate5%

Industry standard for marketing measurement. Suitable for most reporting use cases.

ENTERPRISE CUSTOMIZATION

Enterprise clients can configure thresholds based on decision stakes. Higher thresholds for high-stakes decisions, lower for exploration.

Error quantification available as enterprise feature.

WHAT EVERY RECEIPT INCLUDES

Point estimates

Contribution scores with best estimate

Confidence intervals

Range reflecting uncertainty

Sample size

N for each segment analyzed

MEASUREMENT LANDSCAPE

How We Fit In

Different tools solve different problems. Here's where Precise fits in the measurement stack.

Marketing Mix Models (MMM)

WHAT IT DOES

Top-down budget allocation using aggregate spend + outcomes

STRENGTH

Works without user-level data. Good for channel allocation.

GAP

Cannot see segment-level decisions. Slow refresh (monthly/quarterly). No real-time governance.

PRECISE'S ROLE

Precise operates at segment-level in near-real-time. Complements MMM with tactical granularity.

Multi-Touch Attribution (MTA)

WHAT IT DOES

User journey analysis assigning credit to touchpoints

STRENGTH

Good for cross-channel path analysis with full identity graph.

GAP

Dependent on shrinking identity coverage. Cannot attribute segment-level costs within DSP.

PRECISE'S ROLE

Precise works at the segment level within DSP logs, independent of identity resolution.

Platform-Native Reporting

WHAT IT DOES

DSP dashboards showing performance metrics

STRENGTH

Built-in, no integration needed.

GAP

DSP reports on its own optimization. No independent verification. No audit trail.

PRECISE'S ROLE

Precise is the independent verification layer. We read DSP logs but don't optimize them.

INTELLECTUAL HONESTY

What We Claim vs. What We Don't

WHAT WE CLAIM

✓Segment-level contribution accounting from logs
✓Reproducible methodology with version control
✓Coverage transparency (what we did/didn't measure)
✓Independent verification (no DSP conflicts)
✓Incremental lift when holdouts exist (Level 3)

WHAT WE DON'T CLAIM

✗Causal incrementality without designed experiments
✗Cross-channel attribution (that's MTA territory)
✗Budget allocation recommendations (that's MMM territory)
✗Guaranteed results (we show what we found, not what will happen)
✗Perfect precision (uncertainty is quantified, not hidden)

CROSS-CAMPAIGN LEARNING

When a segment appears across multiple campaigns, we aggregate evidence to strengthen confidence. A segment that consistently underperforms across 10 campaigns has stronger "cut" signal than one that appeared once.

Available now for multi-campaign clients. Contact us for details.

Independence Guarantee

Precise does not accept rebates, rev-share, or compensation for steering spend to any DSP, SSP, or data provider. We provide policy dials + receipts; agencies and brands decide whether to execute recommendations.

DETAILED Q&A