Seasonality

Seasonality detects repeating demand patterns tied to time of year — holiday spikes, summer slowdowns, back-to-school surges, and similar cycles. It analyzes both sales rank and price history together to find these patterns.

What It Measures

This metric looks for repeating cycles in the product's sales rank. A product with strong seasonality will have demand that rises and falls at roughly the same times each year. The analysis identifies the cycle length (e.g., 52 weeks for annual patterns), the timing of peak and trough demand, and the strength of the pattern.

Price patterns are also checked to see if they align with demand cycles, which increases confidence in the detection.

Why It Matters for Resellers

Seasonal products can be very profitable if you time your inventory correctly, but risky if you do not:

• Peak periods are when demand is highest. Having inventory ready to sell during peaks maximizes your revenue.
• Trough periods are when demand drops. Getting stuck with inventory during a trough ties up capital and may force liquidation.
• Lead time matters. You need to order inventory 6-8 weeks before a peak to ensure it arrives in time.

How We Calculate It

1. We collect at least 26 weeks (6 months) of sales rank data. More data means more reliable detection.
2. We remove long-term trends so that steady growth or decline does not mask seasonal patterns.
3. We test every possible cycle length from 4 to 52 weeks, measuring how strongly the pattern repeats at each length.
4. We pick the cycle length where the pattern repeats most strongly.
5. We check whether price patterns align with the demand cycle, which raises our confidence.
6. If the pattern is strong enough (above a threshold that adjusts based on data length), we classify it by confidence level.
7. For detected patterns, we identify peak and trough timing and your current position in the cycle.

How to Read the Results

| Classification | What It Means | |---------------|---------------| | Not detected | No repeating seasonal pattern found. Demand appears consistent year-round based on available history. | | Low confidence | A weak seasonal signal exists but is not very pronounced. Demand shifts slightly by season but not enough to require major inventory adjustments. | | Moderate confidence | A clear seasonal pattern is present. Plan to increase inventory before peak periods and reduce orders during slow periods. | | High confidence | Strong seasonal swings detected, confirmed by both sales rank and price patterns. Demand changes dramatically by time of year — timing your inventory purchases is critical. |

Confirmation Level

When a seasonal pattern is detected, it is classified as either confirmed or candidate:

• Confirmed — At least 2 full cycles of data are available, meaning the pattern has repeated and is more likely to continue.
• Candidate — Fewer than 2 full cycles of data are available. The pattern is present but has not yet been observed to repeat. Treat with more caution.

Phase Information

When a seasonal pattern is detected, you will also see phase information:

• Peak — You are near the highest demand period. Sell through your inventory now.
• Ascending — Demand is rising toward the peak. Good time to have inventory in place.
• Descending — Demand is falling from the peak. Watch your stock levels and slow down reorders.
• Trough — You are near the lowest demand period. Avoid heavy restocking until demand starts rising again.

Limitations & Caveats

• Requires at least 26 weeks of data. Ideally, you want a full year or more to detect annual patterns reliably.
• Annual patterns need ~78 weeks (1.5 years) of data for reliable detection. With only 6 months of data, the algorithm can detect shorter cycles but may miss annual seasonality.
• Not all repeating patterns are seasonal. Some products have cycles driven by promotional events, restocking patterns, or other non-seasonal factors. The algorithm cannot distinguish the cause.
• Phase predictions assume the pattern continues. If market conditions change (new competitors, Amazon changes, etc.), the seasonal pattern may shift or disappear.
• Weak patterns may be noise. The "low confidence" classification means the signal exists but may not be strong enough to act on with high conviction.

Related Metrics

• Demand Stability — Seasonal products often show as "irregular" or "volatile" in demand stability because their demand genuinely varies by season.
• Amazon Out-of-Stock — Amazon's stock-outs sometimes follow seasonal patterns, aligning with demand peaks.
• Price Trend — Seasonal price movements can look like trends over short time windows.