The balance between buyers and sellers is an important driver of how long it takes to buy or sell a home, bargaining power between each side, and home prices.

The number of active sellers in the housing market is simply the number of homes listed for sale over a given period. There is no equivalent public data source for the number of active buyers. As such, we use a model to estimate the number of active buyers by combining publicly available MLS data with Redfin’s proprietary data on how long buyers spend searching for a home in each metro area. Our model estimates the number of buyers and sellers in the U.S. housing market as a whole — not just the number using Redfin as their brokerage or real estate platform.

Our Approach

We apply a standard economic framework used by economists to understand how homebuyers and sellers “match” with one another: When buyers and sellers are searching for each other, the ease with which each side finds a match depends on how many people are on the other side. If there are many buyers and few sellers, sellers match quickly and buyers have to search longer. If there are many sellers and few buyers, the reverse is true. With information on how quickly buyers find homes and how quickly listed homes sell, we can infer how many buyers on average were in the market over a given period.

What’s novel about the data: Our method uses Redfin data on thousands of actual buyers’ search behavior to estimate how long buyers search in each metro area, each month—capturing local differences in buyer behavior over time across the country.

What We Report

We report the following metrics on a monthly basis for the following regions: National, Census Regions, and Top 50 metro areas. We also report these numbers by several property types–single family, condo/co-op, and townhouse. All metrics are seasonally-adjusted.

We classify markets based on the balance of buyers to sellers: a seller’s market has roughly 90 or fewer sellers per 100 buyers, a buyer’s market has roughly 110 or more, and a balanced market falls in between.

How To Use These Data

Identifying buyer’s and seller’s markets. The ratio of buyers to sellers provides a direct read on market conditions at the national, regional, and metro level. Nationally, we estimate the long-run average number of buyers and sellers is roughly equal, but with buyer’s or seller’s markets occurring 30-40% of the time.

Forecasting home price growth and days on market. Shifts in the balance of buyers and sellers tend to foreshadow changes in home prices and marketing time. A rising number of buyers relative to sellers can signal upward pressure on price growth and lower time on market for sellers, while a falling number of buyers relative to sellers can signal the reverse.

Understanding local market structure. Some markets show a persistently elevated number of buyers for each seller, while others show the opposite. These patterns can indicate structural imbalances driven by population growth, housing development, or characteristics of the local housing stock that create frictions–making it more or less difficult for buyers to navigate the housing market.

An improvement over months of supply. Months of supply—a commonly used measure of market balance—is calculated as inventory divided by the pace of sales. It is useful, but only captures the seller’s side of the market. A decline in months of supply could mean buyers are surging or sellers are pulling back—two very different stories. Our measure of buyers and sellers separates these dynamics by estimating both sides independently.

How We Estimate the Number of Buyers

A Simple Coffee-Shop Analogy

Suppose you are standing outside a coffee shop and would like to know how many customers are currently inside. With only two pieces of information, you could estimate the answer without entering and manually counting them:

• How long each customer spends, on average, inside the store
• How many customers leave the store every hour

For example, suppose each customer spends an average of 15 minutes in the store, and 40 customers leave every hour. Then there must be about 10 customers inside at any given time. (15 minutes × 40 customers per hour ÷ 60 minutes = 10 customers.)

If a surge of customers enters the store, the average time each customer spends inside should increase as lines grow and open seats become scarce. Based on this increase in waiting time you could infer, without needing to count the buyers as they entered, how much the number of customers inside at any given time has grown.                                                                                                                                                                                                        

Our method relies on a similar logic. In the housing market, we observe how many homes sell each month and how long buyers spend searching, and from these we estimate the number of active buyers. The key difference from the coffee shop is that the housing market is two-sided: both buyers and sellers are searching for a match, and the ease with which each side matches depends on how many people are on the other side. This two-sided structure is what makes the balance of buyers and sellers relevant for bargaining power and market outcomes such as prices and time on market.

Calculating the Number of Buyers

Our estimate of the number of buyers rests on three inputs:

The first two are derived from public MLS data. The third—buyer search time—is where Redfin’s proprietary data provides a unique contribution. We measure the time from a buyer’s first home tour to closing, which corresponds to the period of most active search. We use a standard statistical model (an exponential distribution) to convert median buyer search time into a monthly buyer match rate.

Our model implies a simple relationship between these three inputs: 

Number of Buyers = (Seller Match Rate ÷ Buyer Match Rate) × Number of Sellers

Economic Framework and Assumptions

Our model is based on an economic framework commonly used by researchers to estimate the balance of buyers and sellers in the market. We provide detail below on the quality of our data and how our model and its assumptions fit into the body of research on this topic.

Redfin’s estimate of median buyer search time is broadly representative of the market: Redfin is one of the largest residential real estate brokerages in the United States, with tens of thousands of buyers purchasing homes with Redfin Agents each year. For each region, we use a rolling measure of buyer search time to smooth out noise and ensure sufficient sample size. As a source of external validation, our estimate of the national average buyer search time is similar to the national values reported in the NAR’s annual Profile of Home Buyers and Sellers. For example, the NAR reports the median buyer in 2024 searched for 10 weeks; we estimate a median buyer search time of approximately 9 weeks in 2024.

The speed at which buyers find homes can be used to infer the number of active buyers in the market: We use median buyer search time, or the speed at which the typical buyer moves through the market, to estimate the rate at which existing buyers are finding homes. To do so, we make the simplifying assumption that the amount of time a buyer spends searching follows an exponential distribution, which allows us to estimate the buyer match rate using median time on market. This assumption is intended to capture the average rate of exit of successful buyers through the market.
                                                                                                                                                                                                                                                              Observed buyer search times reflect market-wide congestion and other frictions, including from buyers who ultimately do not complete a purchase: Our estimate of median buyer search time is based on Redfin buyers who successfully closed a deal. However, even buyers who ultimately withdraw without purchasing a home can contribute meaningful frictions to the housing market, such as buyers who have an offer accepted but fail to close. A market with more active buyers increases the time it takes for any individual buyer to find a home. Because the amount of time successful buyers spend in the market reflects this congestion, median buyer time on market serves as a signal of overall market balance (or “tightness”). We therefore interpret our measure of active buyers as the total number of effective buyers contributing to tightness in the market.

Relationship to the academic literature: Our approach follows the standard method of inferring the unobserved stock of active buyers from observable market data using a search and matching framework. Studies such as Anenberg and Ringo (2024) and Gabrovski and Ortego-Marti (2024) estimate the number of buyers indirectly from the seller side, using the rate at which listed homes sell and an assumed matching function elasticity to infer how many buyers must be in the market. Because we observe buyer search duration directly from Redfin tour data – conceptually the same measure as the NAR survey-based buyer time on market used by Genesove and Han (2012) – we take a more direct approach: we compute the buyer match rate from median buyer time on market and estimate the buyer stock as sales divided by this rate. This avoids the need to specify a matching function or calibrate its elasticity. As in Anenberg and Ringo (2024) we assume that buyer search time follows an exponential distribution. Finally, our approach to calling buyers and sellers markets is motivated by the empirical relationships between market tightness and home sales and price growth documented by Anenberg and Ringo (2024) and Carrillo, de Wit, and Larson (2015).

References

Anenberg, Elliot and Daniel Ringo, “Volatility in Home Sales and Prices: Supply or Demand?” Journal of Urban Economics, January 2024, Vol. 139. 

Carrillo, Paul E., Eric R. de Wit, and William Larson, “Can Tightness in the Housing Market Help Predict Subsequent Home Price Appreciation? Evidence from the United States and the Netherlands,” Real Estate Economics, January 2015, Vol. 43, No. 3, pp. 609-651.

Gabrovski, Miroslav and Victor Ortego-Marti, “On the Slope of the Beveridge Curve in the Housing Market,” Economics Bulletin, AccessEcon, 2024, Vol. 44, No. 3, pp. 948-960.

Genesove, David and Lu Han, “Search and Matching in the Housing Market,” Journal of Urban Economics, July 2012, Vol. 72, No. 1, pp. 21-45.

National Association of Realtors, “Profile of Home Buyers and Sellers: 2024,” www.nar.realtor, November 2024, accessed April 2026, available at https://www.nar.realtor/sites/default/files/2024-11/2024-profile-of-home-buyers-and-sellers-highlights-11-04-2024_2.pdf?utm.