Small Business Recovery Report

Data sources and methodology

1. Samples

This report does not use survey data. Instead it uses anonymized, aggregate data from around 1 million small businesses that use QuickBooks to manage their business. The exact number of businesses in the overall sample fluctuates over time on a monthly basis from a low of 800,000 small businesses to a high of around 1.1 million because it excludes businesses that do not have corresponding monthly data year over year. The annual revenue analysis also excludes any businesses with less than 10 months of corresponding data available year over year. QuickBooks Online customers who have not connected their business bank accounts were also excluded from the analysis. Consequently, this report only includes businesses that had deposits in their accounts during the given timeframes.

2. Net bank deposits

Throughout this report, the financial impact of COVID-19 is measured by comparing small businesses’ net bank deposits during the pandemic with the same data from the same businesses before the pandemic to reveal how things have changed for them year-over-year. Net bank deposits are incoming deposit transactions into small business bank accounts. This excludes non-operating deposits such as business loans and government assistance (for example, Paycheck Protection Program loans, also known as PPP). All of this data is anonymized and aggregated.

The net bank deposit data was used to calculate a monthly median year over year net deposit ratio (YOY revenue ratio) — referred to as “monthly revenue benchmarked to pre-pandemic monthly revenue” in the charts used in this report — which provides directional insights into small business revenues and recovery. The YOY revenue ratio is calculated as follows:

1. For each available business, calculate the YOY net deposit ratio by dividing the current month’s net deposit dollar amount by the corresponding month before the pandemic.
2. For each month, aggregate by taking the median YOY net deposit ratio among all available businesses for the relevant segment.

For example, a YOY revenue ratio of 85% for a given month means that:

• 50% of companies’ net deposits are less than 85% of the corresponding month before the pandemic.
• 50% of companies’ net deposits are greater than 85% of the corresponding month before the pandemic.

YOY revenue ratio was then overlaid with industry and geographic data to reveal the patterns of small business recovery between April 1, 2020 and March 31, 2021.

3. Industry classifications

For this analysis, businesses were grouped using SIC ( Standard Industrial Classification) categories, with 10 sectors (agriculture; construction; finance/insurance; manufacturing; mining; public administration; retail; services; transportation, communication, energy, & sanitary; and wholesale), comprising smaller industry categories (known as SIC2) and sub-industry categories (known as SIC4).

4. Total revenue decrease in April 2020

The $4.6 billion revenue decrease nationwide during April 2020 only applies to the roughly 1 million small businesses that are included in this report. It also excludes outliers beyond the 99.9th percentile of the data. Given that there are around 4.6 million small businesses in the U.S. with 20 employees or fewer, the total cost may be greater.

5. Per business calculations

Median values are used throughout this report in place of mean (average) values to eliminate outliers that can skew the data.

6. Urban vs suburban vs rural

The data sources for this classification are the U.S. Department of Housing and Urban Development and U.S. Census Bureau’s new measure based on research that asked people whether their neighborhoods felt urban, suburban, or rural — compiled by Census tract. For this analysis we used a zip code correspondence developed by Jed Kolko, chief economist at Indeed.com.

7. Annual revenue data

As described above (see footnote 2, “Net bank deposits”), annual revenues are estimated from anonymized, aggregate net bank deposit data. Annual revenues during the pandemic are benchmarked against the corresponding 12 months before the pandemic. In other words, annual revenue data from April 1, 2020 to March 31, 2021 is compared against annual revenue data from April 1, 2019 to March 31, 2020 from the same group of businesses.

8. Monthly revenue data

As described above (see footnote 2, “Net bank deposits”), monthly revenues are estimated from anonymized, aggregate net bank deposit data. Monthly business performance during the pandemic is then benchmarked against data from the corresponding months before the pandemic (for example, March 2021 vs. March 2019) from the same group of businesses.

Data limitations and considerations

The calculation of year-over-year revenue change required that any business included in the analysis have bank data for 2021, 2020, and the same months in the pre-pandemic years. Businesses were excluded if they had only recently connected bank data to QuickBooks or lost their bank connection within the one-year period.

A consideration of survival bias caused by businesses who discontinued operation or/and disconnected bank accounts during COVID-19 is also necessary. For example, the percentage of businesses that lost bank connections in 2020 (8%) is close to that in 2019 (7%), which means COVID-19 didn’t significantly increase the number of disconnected bank accounts in our data sample. Thus, the data is still comparable year-over-year without significant bias.

Net deposit data is only a proxy for business revenue and may sometimes over or under estimate the true business revenues because this may include non-revenue deposits, such as equity or loan transfers from other bank accounts not linked to QuickBooks. The majority of these were identified and removed by a transaction tagging algorithm. Another potential limitation is that businesses may have only connected one of their business bank accounts in QuickBooks, which would only provide partial insights of their net deposit activity.

Up next: About QuickBooks and Intuit

Source