Peak-to-trough ratio

How it works

The ratio compares the strongest part of the year with the weakest. At 1.0 the year is even; at 3× the best months bring three times the revenue of the worst. It sums up seasonality in one figure, so you can compare stores and categories. A version built on the single best and worst month swings with one sale or one outage.

Two things distort it. A store that grows through the year shows its later months as a peak, so detrending removes that growth first. An incomplete month looks like a trough. In the first run of our own analysis, 51 of 100 stores showed a June 2026 trough, because for 72 of the 100 the data ended on 10 June.

To measure yours, see how seasonal is your store.

Formula

Revenue of the 3 best months ÷ revenue of the 3 worst months, within one 12-month cycle

Example

Example store, not client data.

The tableware shop’s three best months brought 250,000 + 280,000 + 320,000 = 850,000. Its three worst brought 100,000 + 95,000 + 105,000 = 300,000. Peak-to-trough ratio = 850,000 ÷ 300,000 = 2.83×. Taking only the single best and worst month gives 320,000 ÷ 95,000 = 3.37×: the same year looks more seasonal.

Not to be confused with

• Seasonal index — a figure for each month against an average month. The ratio compresses the whole year into one number.
• Peak season — the busy months themselves; the ratio measures how far they stand above the weakest.

Right and wrong readings

• Wrong: “A flatter year is healthier, so we should aim for a ratio near 1.” Right: in GetProfit data on 96 stores, those under 2× grew a median 1.24× over July 2025 – June 2026, those from 2× to 4× grew 1.03×, and those at 4× or more 1.46×. The data does not link an even year to faster growth.

Benchmarks

Benchmark according to GetProfit data (96 online stores, July 2025 – June 2026, ad revenue, 3 best ÷ 3 worst months, incomplete months excluded, growth trend removed): median store 2.62×, middle half of stores 2.01× to 4.01×, top tenth 5.03× or more. Of the 96 stores, 72 were at 2× or more. Your niche may differ.

Sources

• 2.3 Time series patterns — Hyndman and Athanasopoulos, Forecasting: Principles and Practice (3rd ed.): trend and seasonal patterns. Checked 2 October 2026.
• GetProfit data: 96 online stores, July 2025 – June 2026 — best-to-worst months of ad revenue; growth by seasonal swing.