What is rolling averages?

In this article, we will explain what rolling average is and how it works in Penetrace. 

What is rolling average?

A rolling average (sometimes called a moving average) is a statistical method used to smooth out short-term fluctuations in data and highlight longer-term trends. Instead of looking at each individual data point on its own, a rolling average combines data from a set number of recent periods to provide a more stable picture of performance over time.

For example, if we calculate a 3-wave rolling average, each reported result is based on the current wave and the two previous waves combined. When a new wave of data is added, the oldest wave drops out of the calculation, ensuring the results stay current while reducing the impact of any one wave’s volatility.

Why use rolling averages in Market research?

In market research—particularly in brand tracking studies—we collect data from samples of respondents to infer insights about a larger population. Since these samples can vary slightly in composition from wave to wave, results can sometimes look more volatile than the true underlying trend.

Using a rolling average helps to:

• Reduce random variation caused by sample differences.
• Reveal true underlying shifts in brand metrics (e.g., awareness, consideration, usage).
• Provide more reliable trends for decision-making.

This is especially important in tracking research, where the goal is to understand how a brand is evolving over time rather than focusing on short-term noise.

How Penetrace calculates rolling averages

At Penetrace, we apply rolling averages on the raw respondent-level data rather than on pre-calculated averages. This distinction is important:

• If one data point is based on 200 respondents and another on 300 respondents, the rolling average result is calculated from all 500 respondents combined.
• This ensures that waves with larger sample sizes contribute proportionally more weight to the rolling average.
• Other approaches (less precise) sometimes calculate the rolling average by simply averaging the wave-level results, which would give equal weight to both waves—regardless of how many respondents each contains.

By using the raw data approach, Penetrace ensures that rolling averages reflect the true sample distribution across waves and provide more accurate, weighted results.