Is my sales drop my own or the whole market's?
Compare your sales change with a public monthly sales series for your kind of business, where one exists, over the same months, both in percent. If the series fell as much, the market may explain most of it. The gap is the part the series does not explain. In the example, a store fell 10 percent and its series 4.
Principle and explanation
Take the same months a year apart, for your sales and for a public series for your kind of business. In the US, the Census Bureau publishes monthly retail and food services sales by kind of store, and the St. Louis Fed's FRED site serves them. Change in percent = (this year − last year) ÷ last year × 100. Compare percents, not amounts: the series covers far more sales. Your figures keep the season, so read the series unadjusted. The gap between the two changes is what the series does not explain.
What the gap means, and when you cannot compare
A gap near zero means the series matches most of the drop. A large gap points to causes the series does not cover, such as your own business, your area, a mix it does not match, or chance. With six to fourteen months of sales, compare your last three months with the three before, for you and the series, only if its seasons follow yours. Under six months, you cannot compare yet. Without a series for your kind of business, there is no market figure; compare your own customer groups.
Worked example
Illustrative scenario. These are fictional inputs, not customer results.
Lena runs a shoe store. Her sales for June to August were $90,000 last year and $81,000 this year. A public monthly series for her kind of store, read unadjusted for the same three months, fell 4 percent.
| June to August | Change |
|---|---|
| Lena's store: ($81,000 − $90,000) ÷ $90,000 | −10 percent |
| Public series for her kind of store | −4 percent |
| Gap: not explained by the series | 6 percentage points |
The series matches about 4 of her 10 points. The other 6 it does not explain, so she first checks which customers bought less. Had the series fallen 9 percent, it would match nearly the whole drop.
What this does not show
- It shows how much of a drop the series matches, not why either part happened.
- A cause of your own, such as a lost key customer, can coincide with a national dip of the same size and hide inside the matched part.
- It sets no target and says nothing about what a normal change is.
First step
Write your sales for the last three months or more. Add the same months last year or, lacking those, the months before. A short run need not show your trend. Find the public series and work out both changes.
Sources
- Federal Reserve Bank of St. Louis (FRED): Retail Sales: Shoe Stores (U.S. Census Bureau, Monthly Retail Trade and Food Services) (opens in a new tab)The US Census Bureau's monthly retail sales by kind of store, served by FRED, including versions not seasonally adjusted. Shoe stores are one example; no value from the series is used.
- Belief in the Law of Small Numbers (Tversky & Kahneman) (opens in a new tab)People hold wrong intuitions about chance and treat a small random sample as highly representative of its whole population. Reading a few months of sales the same way is this page's application.