Inventory how-to

How to adjust a forecast after a viral spike

Do not let one viral spike skew your reorder quantities.

Short answer: Do not let one viral spike skew your reorder quantities. Separate the spike days from your normal sales history, decide if the new demand level is real or temporary, and adjust your forecast inputs by hand until you have enough post-spike data to trust again.

Why spikes break forecasts

A forecast built on a 28-day sales window assumes those days are roughly normal. A viral spike pushes daily sales far above the real baseline. If the spike stays inside the window, it inflates the average and tells you to order far more stock than you actually need once demand settles back down.

Separate the spike from the baseline

Look at daily sales for the SKU and mark the days where volume jumped well above the usual range. Calculate your baseline daily sales using only the days before the spike, or the days after it once sales have settled. Use that baseline, not the blended average, to decide reorder quantities.

Decide if the new level is temporary or permanent

Watch what happens in the days right after the spike. If sales drop back to roughly the old baseline, treat the spike as a one-off event and order based on the old baseline plus a bit extra for the cleared-out stock. If sales settle at a new, higher plateau, treat that plateau as the new baseline and increase your order quantities and reorder points to match it. Give it at least a week or two of stable data before you commit to a new baseline.

Rebuild history gradually

Once you have three to four weeks of post-spike data, your rolling window will naturally exclude the spike and recalculate on its own. Until then, treat any forecast that includes the spike days with suspicion and cross-check it against what you are seeing manually.

Step by step

  1. Pull daily sales for the SKU and identify exactly which days were the spike.
  2. Calculate baseline daily sales using only pre-spike or settled post-spike days.
  3. Watch the week after the spike to see if demand returns to baseline or holds higher.
  4. If it returns to baseline, order against the old daily rate plus enough to refill what sold out.
  5. If it holds higher, raise your daily sales assumption and reorder point manually.
  6. Recheck the forecast once the spike days roll out of the sales window, usually after 28 days.
  7. Keep an eye on the SKU weekly until you are confident the numbers have settled.

Worked example

A store sells a candle at about 5 units a day for weeks. A video goes viral and it sells 220 units in 3 days, then drops back to 6 units a day the following week. If you averaged all 28 days including the spike, you would get roughly 12 units a day and over-order heavily. Instead, using the 6 units a day post-spike baseline with a 10 day lead time and 3 day safety stock, the reorder point is 6 x (10 + 3) = 78 units, which is much closer to reality.

How SteadyShelf helps

SteadyShelf's forecast uses a rolling 28-day window and will blend spike days into the average until they roll out of that window, so right after a viral spike you should expect the suggested quantity to run high. There is no manual override to exclude specific days, so for a week or two after a spike, treat the suggested order quantity as a starting point and adjust it down yourself based on what you know about the real baseline.

$29/month flat, unlimited products and suppliers, 30-day free trial. Run the numbers now with the free reorder point calculator.

SteadyShelf is launching on the Shopify App Store. Join early access to get the install link first.

Related guides

All inventory how-to guides ยท Inventory glossary

FAQ

Will the forecast fix itself automatically after a spike?

Yes, eventually. Once the spike days age out of the 28-day window, the average will reflect normal sales again. That takes up to 28 days from the spike, so you need to manage reorders manually in the meantime.

Should I pause reordering entirely during a spike?

Not necessarily. If the spike is still active and stock is selling fast, you likely need to reorder sooner than usual, just at a quantity based on the real spike rate and how long you expect it to last, not the blended average.

Let SteadyShelf do this math for every SKU

Order-by dates per supplier, seasonal forecasts and draft POs for your Shopify store. $29/month after a 30-day free trial.

Get early access