After managing hundreds of Amazon accounts through peak, we can tell you how most Amazon PPC Q4 budgets get set. Somebody opens the console in October, looks at what the account spent in September, adds a percentage that feels bold, and moves on. The number is written down as dollars. It stays as dollars. And in the third week of November it quietly becomes the tightest cap the account has run all year, because a dollar buys fewer clicks at peak than it did in September, and nobody re-ran the arithmetic.
This is not a post about spending more. It’s about the unit you set the budget in. A budget is a decision about how many clicks you’re willing to buy, and if you set it in dollars during a window where the price of a click moves 40-80%, you didn’t set a budget. You set a cut and called it a plan.
The arithmetic nobody runs in October
Take a brand doing $200K/month with $30K/month in ad spend, a blended CPC of $1.10 in September. That budget buys roughly 27,000 clicks a month.
Practitioner data on peak CPC inflation is consistent in direction and wide in range: 40-80% above baseline across Black Friday week, with some categories worse. Call it 50% for the exercise. November CPC is now $1.65.
If the brand raises its November budget to $40K, which feels aggressive, it buys about 24,000 clicks. A 33% budget increase bought 11% fewer clicks than September. If the brand carries $30K forward unchanged, it buys 18,000 clicks, a third fewer than the quietest month of Q3, in the ten days where conversion intent is highest all year.
That is the whole problem in one paragraph. The board approved an increase. The account got cut. And in January the ACOS on those campaigns looks fine because you bought fewer clicks at higher intent, so nobody notices the volume that never happened.
Four budget mechanics that behave differently at peak
Before sizing anything, you need to know what the console does with the number you type, because four different mechanisms are all called “budget” and they don’t behave the same way in November. Confirm the current specifics in your own console; Amazon revises these, and the shape matters more than the exact figure.
Daily budget can overspend by up to 25% on a high-traffic day. Amazon may spend above your daily budget on days it sees more opportunity, averaged so the month stays within the cap. In a quiet month this is invisible. On Black Friday it means a $500 daily budget can run $625 in a single day and then throttle later in the month to compensate. If you set daily budgets assuming they’re hard ceilings, your event days are bigger and your recovery days are smaller than you planned.
Portfolio budget caps don’t pace. A portfolio cap is a ceiling on total spend across the campaigns in it, with a date range. When the cap is hit, every campaign in the portfolio pauses until someone raises it. It does not spread spend across the period. One campaign can exhaust the whole portfolio on a deal day and pause your branded defense campaign alongside it. A cap set in March for “Q4 total” is one of the most common ways we find an entire portfolio dark on the Saturday after Black Friday, with nobody at the desk.
Budget rules are a multiplier on whatever you already set. Schedule-based rules increase a campaign’s daily budget by a percentage across a date range, with Amazon offering suggested event windows. Performance-based rules increase it when a metric clears a threshold. Both are useful and both inherit the problem above: a 50% budget rule on a campaign whose base budget was already a dollar-denominated cut just produces a slightly less severe cut.
Bid rules and budget rules stack. A campaign carrying a 30% event bid increase and a 50% event budget increase on the same day is running both, and the bid increase inflates the CPC that the budget then has to cover. Brands set both in September, in different screens, and meet the combined effect in November.
Size the budget in clicks, then convert to dollars
Here is the sequence we run on every account in the first two weeks of September.
1. Establish the click baseline. For each campaign tier (branded defense, core converting, deal-supported, discovery, conquest), pull August clicks and August CPC. Write down clicks. Clicks are what you’re deciding about.
2. Decide the click target per tier. Branded defense should hold or grow, because competitor conquesting on your name rises in Q4 and that traffic converts at your best rate. Core converting should grow in proportion to expected demand. Deal-supported ASINs should grow sharply inside the deal window and return to baseline after. Discovery should shrink or hold; paying peak prices to learn which keywords work is the worst use of a November dollar. Conquest is a judgment call and usually the first tier to cut when the budget is finite.
3. Apply a CPC inflation assumption per tier. Not one number for the account. Branded terms inflate less than category terms. Deal-window CPCs on hero terms inflate the most. If you have last year’s data, use your own November-over-August CPC ratio by tier. If you don’t, use a range and write the assumption down with a date on it.
4. Convert to dollars. Click target times inflated CPC equals the tier budget. Sum the tiers. That’s the number.
5. Check it against contribution. Total ad spend divided by expected contribution margin, by tier where possible. If deal-supported spend is fine at a thin margin because the deal is buying rank on a newer SKU, say so in writing. If discovery spend is running at an ACOS that exceeds contribution, that tier has no business inflating in November. The point of the check isn’t to cap the number. It’s to make sure every dollar has a reason attached that someone could read in January.
Most brands who run this for the first time find the honest number is meaningfully higher than what they’d planned for the ten days around Black Friday, and lower than what they’d planned for the first three weeks of December, when CPCs stay elevated and gift-buying conversion softens as shipping cutoffs approach.
The shape of the quarter matters more than the total
A single Q4 number spread evenly across thirteen weeks is wrong in both directions. Demand and CPCs move together and not linearly. The practical shape:
- October 1 to the week before Prime Big Deal Days: near baseline. Modest increase on deal-supported SKUs three to four days before the event to build velocity.
- PBDD window: the first test of your peak assumptions. Fund it, read it at seven days minimum, and use the actual CPC inflation you observed to correct the Black Friday assumptions. This is the single most valuable read of the quarter and most brands treat it as a rehearsal instead of a measurement.
- Late October through mid-November: rising. Shoppers are carting and waiting. Cart-add rates rise and purchase rates lag; don’t read that as a conversion problem.
- Black Friday through Cyber Monday: the spike. This is where daily-budget overspend, portfolio caps and stacked rules all fire at once. Somebody with authority needs to be watching spend pacing on Friday morning and Saturday morning. Not ACOS, which isn’t settled for a week. Spend and delivery.
- December 1 to shipping cutoffs: elevated but softening. CPCs stay high; conversion on non-gift SKUs drifts down. This is where a December budget carried at the November level overspends into worse efficiency.
- After cutoffs: cut hard. The last week of December is a different market and most event rules should already be off.
Write the shape down as a weekly table before October. The mistake isn’t picking the wrong total. It’s spending the right total in the wrong weeks.
The three failure patterns we see every January
The dollar carry-forward. Covered above. The tell in the post-mortem is impression share falling on core terms while ACOS looks acceptable. You bought fewer clicks at higher intent and called it efficiency.
The portfolio cap nobody remembered. Campaigns paused on a Saturday, discovered on a Monday, with two of the highest-intent days of the year spent dark. The tell is a flat line in spend across a weekend that should have been a peak.
The rule that never got turned off. A 40% event budget increase still running on January 14 against post-holiday conversion. The tell is January ACOS drift blamed on seasonality, which it partly is, and partly isn’t.
All three are calendar problems, not strategy problems. Each has a date attached and a name that should be next to it.
What to do in the next two weeks
Prime Big Deal Days deal submissions have closed and BFCM deal submissions close October 20, so the SKUs that will be deal-supported are largely known now.
FAQ
How much should I increase my Amazon PPC budget for Black Friday? There isn’t a universal percentage, and anyone quoting one hasn’t seen your CPC data. Size it in clicks per tier, apply your own observed CPC inflation, and convert. For most accounts the honest number on core and deal-supported campaigns is a larger increase than feels comfortable for the event window and a smaller one for early December.
Does Amazon really spend more than my daily budget? Amazon may exceed a daily budget by up to 25% on high-traffic days, keeping the monthly total within the cap. Confirm the current behavior in your console. Practically, it means event days run bigger than the number you typed and quiet days smaller.
Should I use budget rules or just raise the budget manually? Rules are fine on campaigns with real history and a defensible base budget. They’re a multiplier, so fix the base first. And put a date on turning them off; an event rule running into mid-January is the most common January efficiency problem we see.
What’s a good ACOS target for Q4? A blended Q4 ACOS target is an average that describes none of your campaign tiers. Set targets per tier from contribution margin, and grade peak against last peak, not against October. ACOS in November mostly tells you what the market charged for a click, not how well the account was run.
We’re launching a new SKU in October. How should we budget it? Conservatively, and with a human managing bids rather than a rule. A campaign with no baseline can’t have a sensible multiplier applied to it, and paying peak CPCs to establish a baseline is the most expensive way to learn what a click is worth on that product.
If you’re looking for a team that manages every lever — creative, advertising, and operations — Velocity Sellers works with brands doing $100K+/month on Amazon. Contact us for a free account audit.
Related reading: ACoS Amazon Explained: Analysis, Benchmarks, and Strategies · Sponsored Products Moves Into Creator Content on August 10 — And Your Campaigns Are Already Enrolled
