# Moving Average

**Category:** metrics  
**Short Description:** Rolling calculation that smooths data fluctuations over time.  
**Last Updated:** 2026-07-07T00:00:00Z

## Definition

A moving average is a statistical calculation that creates a series of averages from different subsets of data over time. By recalculating the average over a sliding window — commonly 7 or 28 days for ad data — it separates trend from noise, smoothing out short-term fluctuations and random outliers in metrics like CPC, CTR, or ROAS. Daily platform numbers are often too volatile to act on directly; the moving average is the version you can actually make decisions on.

## Formula

**Formula:** `MA = (V₁ + V₂ + … + Vₙ) / n`

Average of the last n periods, recalculated as each new period arrives — smoothing daily noise into a trend you can act on, in the same units as the metric being smoothed.

## Calculation

**Formula:** `MA = (V₁ + V₂ + ... + Vₙ) / n`

**Explanation:** Sum of values in the time window divided by the number of periods, where V represents values and n is the time window. The result is in the same units as the metric being smoothed (a CPC moving average is in dollars).

### Components

- **Current Value**: The metric value for the current period
- **Previous Values**: The metric values for the previous n periods
- **Time Window**: Number of periods to include in the moving average

## Examples

- Daily conversions of 10, 14, 9, 12, 15, 11, 13 sum to 84, so the 7-day moving average is 12 per day — a stable trend line despite ±25% daily swings
- 7-day moving average of daily CPC to identify bidding trends
- 30-day moving average of ROAS to assess campaign health

## How AdSights Helps

**Tracking Moving Average:** Daily creative metrics are too noisy to read raw, but over-smoothed dashboards hide fatigue until weeks after it starts. AdSights tracks each variant on smoothed trend lines tuned to the metric's volatility, so a genuine CTR decline separates from daily noise early — a creative refresh at the first real inflection, not after a month of wasted spend.

## FAQs

### What window should I use for a moving average of ad metrics?

Match the window to the decision speed. A 7-day window is the workhorse for ad data because it cancels the weekday/weekend cycle exactly while staying responsive; 14 or 28 days suit strategic reads like monthly ROAS trends; 3-day windows track fast-moving launches but keep more noise. The trade-off is fixed: longer windows are smoother but lag reality by roughly half the window length.

### What is the difference between a simple and an exponential moving average?

A simple moving average weights every period in the window equally — a day from a week ago counts the same as yesterday. An exponential moving average (EMA) applies exponentially decaying weights so recent data dominates, reacting faster to genuine changes at the cost of slightly less smoothing. Use the simple version for stable reporting baselines; prefer an EMA when catching turning points quickly matters, like fatigue detection.

### Can a moving average hide creative fatigue?

Yes — smoothing trades responsiveness for stability, so a long window can mask the early inflection where a creative starts decaying. A 28-day average of a fatiguing ad's CTR keeps coasting on old, healthy data after the recent days have broken down. The fix is reading two horizons together: a short or exponential average for detection, a longer one for confirmation. A short line persistently below the long one means fatigue is underway.

## Related Terms

### Similar Terms

- **[Statistical Significance](/resources/glossary/metrics/statistical-significance)**: Both help identify true performance patterns vs random variation. While moving averages smooth fluctuations, statistical significance validates if changes are meaningful
- **[Exponential Moving Average (EMA)](/resources/glossary/metrics/exponential-moving-average-ema)**: EMA prioritizes recent data while standard moving average weights all periods equally
- **[Anomaly Detection](/resources/glossary/metrics/anomaly-detection)**: Moving averages supply the expected baseline that anomaly detection measures deviations against

### Component Terms

- **[Standard Deviation](/resources/glossary/metrics/standard-deviation)**: Helps establish normal ranges around moving averages for monitoring performance

## Related Resources

- [Moving Average Interactive Example in AdSights Creative Fatigue Article](/blog/topics/creative-strategy/creative-fatigue-in-meta-ads-detection-and-management-strategies) - Demonstrates how moving averages help identify creative fatigue patterns in advertising performance