Most Amazon PPC budgets for next quarter are set by taking this quarter's spend and adding a percentage that feels right. It is quick, and it ignores everything that will be different: the season, the products launching, the events on the calendar, the campaigns that are already running out of budget.

This playbook builds a quarterly forecast from your own data in five steps. The result will not be exact. It does not need to be. A useful forecast is one that sets a range, makes the assumptions visible, and tells you early when the quarter is going off plan.

Step 1: Build the baseline

Pull the last three to six months of monthly data: ad spend, ad sales, ACoS, total sales and TACoS. Break it down by product or product group and by campaign type (brand, generic, competitor, discovery).

Look for the run rate: what a normal month looks like right now. Exclude one-off months, such as a big sale event or a stockout, or adjust them back to normal. That run rate is your baseline. It answers the question "if nothing changed, what would next quarter look like?"

Step 2: Apply seasonality

Nothing stays unchanged, and the season is the biggest reason. If you have last year's data, compare each month of next quarter with the baseline months last year. If January last year ran at a certain ratio of September, apply a similar ratio this year.

If you do not have last year's data, use what you know about the category. Gift products swing hard into Q4 and drop in January. Outdoor products peak in spring and summer. Consumables tend to be steadier. Write down the assumption for each month, even if it is a rough one. The seasonal ad calendar covers the main shifts through the year.

Apply seasonality to sales demand and to CPC separately if you can. In busy months, both tend to rise, and they do not always rise together.

Step 3: Add the planned changes

List what will be different next quarter because you decided it. A product launch. A new marketplace. A deal event. A product being discontinued. A budget shift from discovery to exact match.

Estimate each one separately and add it to the baseline. A launch needs its own budget line, usually with a higher ACoS target for the first months. Launching a new product covers how to size it. An event needs extra budget for a few days. A discontinued product removes its spend.

Keeping these separate from the baseline matters. When the quarter ends, you will want to know whether the baseline was right and whether each planned change performed as expected.

Step 4: Check against constraints

A forecast that ignores limits is a wish list. Check three.

Margin. Forecast ACoS for each product should sit below its break-even ACoS, or above it only for a deliberate reason, such as a launch. What is a good ACoS covers break-even.

Inventory. Forecast sales should not exceed the stock you expect to have. Spending into a stockout is the most common way forecasts waste money.

Demand. Some campaigns cannot spend more no matter the budget, because search volume runs out. If a campaign already spends well below its budget, raising the budget will not raise sales. Check impression share and how often campaigns hit their budgets before assuming more money means more sales. Scaling profitably covers the signs of a demand ceiling.

Step 5: Turn it into a range

A single number invites false confidence. Express the forecast as a range: a low case, an expected case and a high case. The low case assumes CPC rises more than expected or conversion softens. The high case assumes the opposite, or that a launch takes off.

Decide in advance what you would do in each case. In the low case, which campaigns get cut first? In the high case, where would extra budget go? Writing these down now saves rushed decisions mid-quarter.

Tracking the forecast during the quarter

Compare actuals with the forecast monthly, and with prorated targets weekly. The end-of-month review is the natural place for the monthly check.

When actuals drift, split the gap into its parts. Was it spend, CPC, conversion rate, or demand? Spend running ahead with ACoS on target is often fine. Spend running ahead with ACoS rising needs action. Sales running behind with spend on target points to conversion or demand, and more budget will not fix it.

Revise the forecast once mid-quarter if something big changes, like a major stockout or a competitor's move, but do not rebuild it every week. A forecast that is rewritten constantly stops being something you can measure against.

Using the forecast after the quarter

At the end of the quarter, compare each part of the forecast with what happened: baseline, seasonality, each planned change. Note which assumptions were wrong and by how much. Over a few quarters, your seasonality ratios and launch estimates will get much better, because they will be based on your own results rather than guesses. Setting PPC budgets covers how to turn the forecast into daily campaign budgets.

Frequently asked questions

How do you forecast Amazon PPC spend?

Start from a baseline of recent monthly spend, sales and ACoS. Adjust it for seasonality using last year's pattern or category knowledge, add the effect of planned changes such as launches or events, and express the result as a range rather than a single number. Then compare actuals against it monthly.

How accurate can an Amazon PPC forecast be?

Rarely very precise, because CPCs, competition and conversion rates move in ways you cannot control. A forecast that lands within a reasonable range most months is doing its job. The value is in the decisions it forces and in spotting early when the quarter is going off plan.

Should I forecast by campaign or by product?

Forecast at the level you make budget decisions, usually by product or product group and campaign type. Forecasting every campaign individually adds work without much extra accuracy, because individual campaigns are noisy.

Off Hours budget rules and spend alerts keep each campaign inside the forecast you set, and every change is logged for the end-of-quarter comparison. Start a free 14-day trial.