Send a text message to 72727 with the word "twrStats" and receive an unpublished wedding statistic every Monday
Number of weddings last updated 08/2013
Spending and demand data last updated 10/2014
The Wedding Statistics provided on this site are collected and aggregated from different sources. The primary source for number of weddings comes from Federal, State and local governments. The primary source of wedding cost and other fact data comes from proprietary surveys to pre-wedding couples and post-wedding couples. All survey data is collected from random anonymous participants in electronic form. In some surveys we may use certain wedding entities. In that case, it would be specifically mentioned in the methodology of the final report.
Over 200,000 survey samples have been collected since 2005. The most current cost calculations are derived from 3,700 surveys samples collected in 2013. We also use data from the CDC, Census Bureau, Department of Labor, Bureau of Economic Analysis (BEA), COLI, GeoLytics, Inc. and Easy Analytic Software, Inc., as part of the estimating and forecasting process.
A key principle in market research is sample diversity. Without a diverse sample you end up with biased or unbalanced results. This is the main reason we collect samples from pre-wedding couples, post-wedding couples, and wedding businesses. We also go to great effort to make sure our samples are geographically and demographically diverse. Geographical distribution is typically 1-3% from large population areas and .5-1% from medium to small population areas. Demographic distribution includes multiple ethnicities, education levels, occupation types, income levels, first, second, and third plus marriages, and ages 16 plus.
While 5,000 or even 25,000 samples sounds impressive, the reality is that a sample size over 400 for the wedding population doesn't bring much more than the ability to segment data. We collect as many samples as possible so we can segment and build estimating and forecasting models for geographical areas. SurveyGizmo has an excellent write up explaining sample size if you are not familiar with its true meaning.
Our estimating and forecasting models attempt to account for weddings that travel into a market and weddings that originate from a market. We also take into account the economic and social factors of each market. We do not take into account current natural disasters.
The foundation for number of weddings is derived from wedding licenses registered at the US and State level. US and State level numbers are actuals published by the CDC or State. All other markets use proprietary models to arrive at estimates and forecasts.
The reason we estimate below the state level is that no source has actual numbers at Metropolitan, County, or Zip Code levels. Even if a marriage license is issued at the County Recorder's Office or local Clerk the couple can marry anywhere within the State making any numbers provided by a County inaccurate. Only State level numbers are the accurate account of weddings in that State.
We currently collect data on 55 products and services in 10 categories. This data is collected through random anonymous surveys to pre-wedding couples and post-wedding couples all over the United States. We then use proprietary models that combine local demographic data, local econimic data, and local survey samples to estimate spending and demand for each item, for each market, except for US level numbers.
We take a bottom up approach to "Average Wedding Cost." The total "Average Cost" is calculated using "Weighted Demand Average (WDA)," which is; average spent times demand equals the weighted demand average. Sum of weighted demand average for all items equals the average wedding cost. Using WDA instead of a summed average gives a better calculation, because it takes into account all items that couples may or may not purchase.
All other items are calculated from survey data except where noted.
This research is a work in progress. New data is continuously collected, aggregating, re-evaluating, and field tested to improve its reliability and accuracy.