Methodology: How do we analyze our data

The Dwellsy Rental Housing Index is built on a rigorous process that turns raw listing data into insights about rental trends across the U.S.

Before we analyze it, we ensure that the data is clean and ready to be analyzed.

We begin by removing duplicate listings, which can occur when properties are posted more than once or pulled in through multiple sources. These duplicates can distort the data, so we make sure that each property only counts once.

Next, we filter out incorrect listings. These are entries that include things like implausible rent amounts, wrong property types, or incomplete addresses. If a listing doesn’t make sense or is clearly wrong, it doesn’t belong in the analysis. In cases where the same property appears multiple times with small variations, such as slightly different prices, we adjust them to reflect the median value. This keeps individual outliers from skewing the overall picture. We also proactively eliminate fraudulent listings. Although fake listings might be common in the rental market, our system is designed to catch and remove them so they don’t compromise our insights or put renters at risk.

Once the data is clean, we move on to identifying outliers. These are data points that stand out from the rest, either due to error or unusual market behavior. We review each case and decide whether to adjust, remove, or leave it in place, depending on the context. The goal is to make sure extreme values don’t mislead the analysis.

From there, we calculate the metrics that matter. The most important is the median rent, which reflects the typical price more accurately than a simple average. We also use a rolling median, which smooths out short-term spikes or dips and gives a clearer view of long-term trends. To understand how rents are shifting, we track overall price direction — whether it’s going up, down, or staying flat — and look for cycles, such as seasonal patterns that repeat every year.

To separate signal from noise, we use smoothing techniques. An exponential moving average (EMA) helps us see recent changes more clearly by giving more weight to newer data. For longer-term patterns, we apply Gaussian smoothing, which softens short-term volatility and highlights broader cycles.

Finally, we visualize the data. Clean charts helps us see what’s really happening in the market. Through these visuals, we can identify long-term price shifts, seasonal patterns, and meaningful insights that guide renters, landlords, and property managers in their decisions.

For consistency and easy comparison across different markets, all charts in the Index follow a standard scale: we set the minimum axis slightly below the first value (about 500 units lower) and the maximum axis a bit above the last (around 500 units higher). This ensures that the data remains readable and comparable, no matter the city or time period.

By sticking to this methodical approach, we ensure that the Dwellsy Rental Housing Index offers a trustworthy, up-to-date view of how the rental market is really behaving.