Synthetic Control Causal Impact: Measure Change Without Guessing

A before-after chart is rarely enough. If traffic, demand or seasonality were already moving, the post-period change can look impressive while the true incremental effect is smaller or even negative.

The Synthetic Control Causal Impact Lab builds a counterfactual from unaffected control series. It fits weights on the pre-period only, then compares post-period actual results against the synthetic counterfactual.

Why Synthetic Control Helps

A synthetic control is a weighted blend of controls that tracks the treated series before the intervention. If the pre-period fit is credible, the post-period gap is a better diagnostic than a raw before-after comparison.

What To Watch

The biggest risk is using controls that were also affected by the change. The second risk is poor pre-period fit. The lab therefore reports pre-fit RMSE and exports weights so the model can be reviewed.

How To Use The Lab

1. Open the lab and load the sample.
2. Paste aggregate CSV with period, actual, two or more control_* columns and is_post.
3. Confirm controls are not affected by the intervention.
4. Review pre-fit RMSE before trusting post-period lift.
5. Export the effect CSV, weights CSV and receipt before sharing the claim.

What This Does Not Prove

This is diagnostic counterfactual analysis. It is not automatic causal proof, legal advice, financial advice, medical advice, scientific certification, traffic proof, ranking proof, revenue proof or AdSense approval proof.