# False Negative **Category:** metrics **Short Description:** An error where a test incorrectly indicates the absence of a condition when it is actually present. **Last Updated:** 2026-07-07T00:00:00Z ## Definition A false negative occurs when a test, algorithm, or detection system fails to identify a condition or event that is actually present. In digital advertising, false negatives represent missed opportunities where the system fails to recognize valuable signals, such as potential conversions, fraud instances, or relevant audience segments. These errors can lead to underreporting of performance, missed optimization opportunities, and inefficient resource allocation. ## Formula **Formula:** `FNR = FN / (FN + TP)` **Result Unit:** ratio Share of real winners or real problems your test failed to detect — the 'missed opportunity' error. ## Calculation **Formula:** `False Negative Rate = Missed Positives / Total Actual Positives` **Explanation:** Measures the proportion of actual positive cases that were incorrectly classified as negative. Lower values indicate better detection accuracy. ### Components - **Missed Positives**: Number of actual positive cases incorrectly classified as negative - **Total Actual Positives**: Total number of actual positive cases ## Examples - A fraud filter reviewing traffic that contains 1,000 actual bot clicks catches 850 and misses 150: false negative rate = 150 / 1,000 = 15% of invalid clicks passing as legitimate - Attribution model failing to credit touchpoints that influenced conversions - Audience targeting algorithm excluding users who would have converted - Conversion tracking missing valid conversions due to technical issues ## Best Practices - Balance false negative and false positive rates based on business impact - Implement multiple detection layers for critical systems - Regularly validate detection accuracy with known test cases - Adjust sensitivity thresholds based on performance requirements - Consider the cost of missed detections when configuring systems ## How AdSights Helps **Tracking False Negative:** False negatives in creative testing are the winning variant you killed too early, or the fatiguing ad you kept because aggregate metrics looked fine. AdSights tracks per-variant trajectories against sample targets so underpowered 'no difference' calls are flagged before losers are scaled and winners are cut. When holdout-measured lift disagrees with in-platform attribution, AdSights surfaces the false negative risk — the campaign that looked incremental but wasn't, or the creative that looked flat but was genuinely better. ## FAQs ### What is a false negative in A/B testing? A real difference exists but your test doesn't detect it — variant B genuinely converts better, yet the readout says 'no significant difference' and the winner gets discarded. This is a Type II error, with probability β; a test's power (1 − β) is exactly its ability to avoid false negatives. A test with 50% power misses a genuine winner half the times it's run — a coin flip dressed up as an experiment. ### What causes false negatives in ad experiments? Mostly insufficient statistical power: samples too small, tests ended too early, or true effects smaller than the test was sized to detect. High metric variance compounds it — a noisy purchase ROAS needs far more data to reveal the same lift than a stable CTR. A test sized to detect a 20% lift will systematically miss real 8% lifts. If your tests routinely conclude 'no difference,' suspect the tests before concluding your creative can't improve. ### How do I reduce false negatives without inflating false positives? Add power, not looser thresholds — widening the significance level just trades missed winners for fake ones. Increase sample size and duration (plan for at least 80% power at your minimum detectable effect), choose lower-variance primary metrics, and run fewer simultaneous variants so traffic concentrates. Deciding the minimum effect worth detecting before launch is the single highest-leverage step. ### Are false negatives or false positives worse in advertising? It depends on the cost asymmetry, but false negatives are usually the underrated one. A false positive scales a mediocre ad — visible, bounded, self-correcting as data accumulates. A false negative silently discards a genuine winner, an opportunity cost that is never audited: nobody re-tests the ad that 'lost.' Teams that only guard against false positives with strict significance and quick kills accumulate false negatives instead. ## Related Terms ### Opposite Terms - **[False Positive](/resources/glossary/metrics/false-positive)**: Incorrect identification of a condition when it's actually absent ### Component Terms - **[Statistical Significance](/resources/glossary/metrics/statistical-significance)**: Helps establish confidence levels that minimize false negatives - **[Anomaly Detection](/resources/glossary/metrics/anomaly-detection)**: Systems must balance false negative and false positive rates - **[A/B Testing](/resources/glossary/creative/ab-testing)**: False negatives can lead to rejecting effective variations ## Featured in topic hubs - [Experimentation & Statistics](/resources/topics/experimentation)