An Amazon Q4 PPC strategy is a three-phase plan run around each peak event (the October Prime member sale, then Black Friday through Cyber Monday): fix leaks and pre-set budgets in the weeks before, keep proven campaigns funded through the whole event day, then walk every temporary change back. Grade the result on TACOS and profit, because deal prices push ACOS up by arithmetic alone.
I have run Q4 on Amazon Ads for six-plus years, including as lead operator of a 10-account agency portfolio, and the costly peak-season mistakes I find are rarely timid bids. They are a proven campaign dark by mid-afternoon on the biggest day of the quarter, a Top of Search premium multiplied by up-and-down bidding, and event settings still running in January. This guide, the seasonal chapter of my Amazon PPC strategy guide, covers the calendar, an eight-step prep sequence, event-day rules, the stacking math, grading, and the two weeks after.
The Amazon Q4 calendar: two peaks, a tail, and a January bill
Q4 on Amazon has two peaks and a tail. Amazon's October Prime member event comes first, Black Friday through Cyber Monday follows at the end of November, and gift buying runs on until holiday shipping cutoffs. Each peak needs its own before, during, and after, and the October after-period runs almost directly into November's before.
For 2026, Amazon has set Prime Big Deal Days for October 6–7: 48 hours of Prime member deals, opening at 12:01 a.m. Pacific. Reading this in late September? You are already in the final week, so use the short version under the prep steps. The same phases work for the summer Prime Day.
| Window | What changes | The PPC job | What I leave alone |
|---|---|---|---|
| 3–4 weeks before each event | Data still reflects normal traffic | Baseline, restore point, leak cleanup, stock check | Event bid raises |
| The final week | Auctions get busier as competitors ramp up | Lock the roster, pre-set budgets, pick one bid lever per campaign | Structural rebuilds |
| Event days | Deal prices go live and traffic peaks | Keep roster campaigns funded all day; watch spend pace and orders | Intraday ACOS, discovery budgets |
| The two weeks after | Attributed sales settle; organic sales show whether rank moved | Walk changes back, quarantine event data, grade on TACOS and profit | Promoting deal-day search terms |
| December into January | Gift demand runs to shipping cutoffs; returns come back | Fund gift converters to the cutoffs; close the Q4 books in January | Calling Q4 profit early |
How to prepare your Amazon PPC for Prime Day and Black Friday
Preparation is structural work done three to four weeks before the event, while your data is still normal: record a baseline, save the current settings, close the leaks that peak traffic would multiply, confirm stock, choose which campaigns get event budgets, pre-set those budgets, pick one bid lever per campaign, and date every change's reversal.
- Record the baseline. Write down trailing 30-day TACOS, average daily total sales, and profit per unit at your planned deal price. Without them, "did the event work" has no answer. For Black Friday, take the baseline from clean weeks, not the days around the October sale.
- Save a restore point. Download a full bulk file before changing anything: a dated record of every budget, bid, placement modifier, and bidding strategy. Afterward, put the old values back on only the rows you changed, working in a fresh download. Never re-upload the stale file whole.
- Close the leaks while there is time. Peak traffic multiplies whatever an account already wastes. Negate the zero-order terms the free Wasted Spend Finder ranks, give each bleeding search term one exact match owner (my search term bleed guide covers how), and pull placement premiums the data no longer supports.
- Confirm stock can carry the event. Amazon's Sponsored Products best-practices guide is plain that an out-of-stock product cannot be the featured offer, so it cannot be advertised. A product that cannot cover event demand keeps its normal budget.
- Pick the event roster. Event budgets go to proven exact match performance campaigns at or under target, branded defense, and the campaigns behind your deal ASINs. Auto campaigns and broad discovery stay at normal budgets: search terms bought at deal prices make poor discovery data.
- Pre-set the budgets. Raise roster budgets before the event starts, not after they run dry: the evening before, or in advance with a schedule-based budget rule, which the console offers for special events and custom date ranges. Also set a total event spend ceiling, so top-ups on the day have a hard limit.
- Choose one bid lever per campaign. Raise the Top of Search modifier where placement data earned it, or run dynamic bids up and down on proven terms. Not both by default; the stacking math is below.
- Calendar the walk-back. Next to every change, write its reversal date, normally the morning after the event ends. Event settings nobody scheduled a reversal for are one of the most common Q4 hangovers I find in January audits.
If the event is less than a week away
With under a week to go, skip the structural cleanup: fixes made now cannot show results before the traffic arrives. Do three things instead. Save a restore point, pre-set budgets on proven campaigns only, and leave bidding strategies and placement modifiers where they are. An untested setting change on event day is how accounts overspend.
On event day, the costliest failure is a budget that runs out
The most expensive event-day failure is a proven campaign running out of budget in the afternoon. Every dark hour after that is peak traffic, with your deal live and your listing converting, handed to competitors. A normal Tuesday that runs dry at 4pm costs a few ordinary hours; the biggest day of the quarter has no make-up day.
So roster budgets go up the evening before, not in reaction: an event that opens overnight drains a normal budget before anyone is awake to raise it, and by the time the out-of-budget label shows, those auctions are already lost.
On the day I watch two things per roster campaign: spend pace, and whether orders are arriving at all. One that has burned most of its budget by late morning will not see the evening, so it gets a top-up inside the ceiling, provided orders are flowing. Intraday ACOS I ignore, because sales are still being attributed to earlier clicks.
What I will not do mid-event: raise bids across the board, switch bidding strategies, or hand spare budget to discovery. A roster campaign that dies early at every event needs a diagnosis in the quiet weeks, and my guide to campaigns that run out of budget separates the constrained ones from the leaking ones.
Should you stack dynamic bidding and placement modifiers for an event?
Not by default. A Top of Search modifier multiplies the base bid for the top slot, and dynamic bids up and down can raise that result by as much as 100% on any placement. Together, a +100% modifier and up-and-down make the top-slot ceiling four times the normal bid. Raise one lever per campaign.
Both settings feel like one decision, getting aggressive for the event, so sellers make it twice. Amazon's dynamic bidding guide already tells up-and-down advertisers to be ready to spend double the bid they enter, and a raised modifier multiplies it before that. Take one keyword priced for a normal day: the teaching dataset's $111,058 in ad sales across 45,672 clicks works out to $2.43 per click, so at a 30% target ACOS its fair bid is $0.73.
| Event-day setting | Highest top-slot bid | ACOS at the account's $2.43 per click | Revenue per click needed for 30% ACOS |
|---|---|---|---|
| Base bid, dynamic down only | $0.73 | 30% | $2.43 |
| +100% Top of Search modifier only | $1.46 | 60% | $4.86 |
| Dynamic bids up and down only | $1.46 | 60% | $4.86 |
| Both together | $2.92 | 120% | $9.73 |
The modifier row is fine on its own if the top slot really earns double the revenue per click; my Top of Search guide shows how to size that from placement rows, and the free Placement Bid Optimizer does it on your own file. The bottom row is the problem: a $2.92 ceiling needs $9.73 from every top-slot click to hold 30%, four times the account average, on the day a deal price is already shrinking revenue per click. My event rule is one lever per campaign, only where the data earned it, with base bids left where my bid-setting formula put them.
How to judge Black Friday and Prime Day: TACOS and profit, not ACOS
Event-day ACOS is distorted three ways at once: a deal price cuts the revenue each click brings in, attributed sales keep arriving for days after the click, and organic sales move with the event without ever appearing in the ad console. TACOS measured against your pre-event baseline, plus profit per unit at the deal price, is the honest scoreboard.
The deal effect is plain arithmetic. The teaching dataset runs a 32.7% ACOS at full price. Put it on a 20% deal with identical clicks, CPCs, and conversion rate, and revenue per click drops 20% while each click costs the same, so ACOS reads about 40.9%. Holding 32.7% would take a conversion rate near 11.3%, up from 9.1%. Nothing broke. The price changed.
So I grade event clicks against a deal-adjusted ceiling: normal target ACOS divided by one minus the discount. A 30% target on a 20% deal becomes 37.5%. At or under that line, the clicks worked as hard as they do on an ordinary day.
Normal clicks do not mean the event paid. Each ad-attributed order in that dataset carried about $8.77 of ad spend ($36,303 across 4,140 orders), and at the deal price that $8.77 has to fit inside a margin the discount already cut. Check it against your break-even ACOS at the deal price, which the free ACOS calculator works out from your margin.
TACOS catches what ACOS cannot: the organic orders an event throws off. Compare event-window TACOS with your baseline and confirm total sales rose, not only ad sales; TACOS needs total sales from your business reports, which the ad console never sees. It is the scoreboard my results page uses all year, leading every case with profit added, like the $310K/mo kitchen brand at +$28,400 a month with TACOS down from 11.4% to 7.2%. The full lens is in my ACOS vs TACOS guide.
After the event: the two weeks that decide whether it paid
The period after an event matters as much as the event. Over the next two weeks, walk every temporary setting back on its date, keep event-day data out of your bid calculations, wait for attributed sales to settle before grading, and watch whether organic sales stay above the pre-event baseline once ads return to normal.
Walk it back on schedule. The morning after, return budgets, modifiers, and bidding strategies to their restore-point values. A forgotten event setting costs every day until someone notices, on traffic that no longer carries event demand.
Quarantine the event data. Event days corrupt the revenue-per-click input every bid pass reads: deal prices shrink it, deal-driven conversion inflates it. For the next few passes I price from windows that exclude the event, then grade the settled account against its normal column in my phase benchmarks. Repricing December bids off Black Friday conversion rates is how accounts overpay into January.
Look for the rank dividend. The strongest sign an event paid is organic sales holding above baseline after ad spend returns to normal. If total sales drop straight back, the event sold discounted units and bought nothing durable. My organic ranking guide covers how paid sales feed rank.
Harvest with a full-price trial. Some new search terms converted only because of the deal, so each gets a fair trial at full price before promotion to exact match. Terms with real clicks and zero orders even at a discount are among the easiest negatives of the year, and the free Negative Keyword Finder turns them into an upload-ready list.
Frequently asked questions
When should I start preparing my Amazon PPC for Q4?
Start the structural work three to four weeks before each peak event, while your data still reflects normal traffic: record a TACOS and profit baseline, save a bulk file as a restore point, and close the leaks peak traffic would multiply. The final week is for pre-setting budgets, not rebuilding. For the October Prime event, that means starting in September.
Should I raise my bids for Black Friday and Cyber Monday?
Not across the board. Raise budgets on proven campaigns first, because running out of budget costs more on a peak day than a slightly low bid does. For more top-slot share on deal products, raise one lever per campaign, the Top of Search modifier or dynamic bids up and down, not both, and keep base bids priced from normal-day data.
Why is my ACOS so high on Prime Day?
Mostly arithmetic. A deal price cuts the revenue each click brings in while the click costs the same, so ACOS rises even at your usual conversion rate: a 20% discount turns a 32.7% ACOS into about 40.9%. Attributed sales also keep arriving for days, so the event-night figure is incomplete. Grade event clicks against your target ACOS divided by one minus the discount.
How much extra ad budget does Prime Day need?
There is no honest universal percentage. Size it per campaign from your own history: last year's event days if you have them, otherwise how early the campaign runs short on a busy normal day. The goal is keeping each proven campaign live through the whole event. Raise only those campaigns, keep discovery flat, and cap total event spend so top-ups have a limit.
Should I cut my Amazon ad spend after the holidays?
Cut it back to your pre-event settings, not below them. Walk budgets, modifiers, and bidding strategies back to your restore point on schedule, then price bids from clean, non-event data. Slashing spend in January to recoup Q4 costs can give back the organic rank the season paid for. Judge the season on TACOS and profit once holiday returns have landed.
Is Prime Big Deal Days the same as Prime Day?
No. Prime Day is Amazon's summer sale for Prime members; Prime Big Deal Days is its October Prime member event, set for October 6–7 in 2026. The PPC playbook is identical: prepare in the weeks before, fund proven campaigns through the event, and walk changes back afterward. The October event just sits close to Black Friday, so its after-period runs into holiday prep.
Sources
- Best practices for your Sponsored Products ads, Amazon Ads
- Guide to dynamic bidding with Sponsored Products, Amazon Ads
Find the leaks before the peak traffic finds them
Step three of the prep sequence is the one most accounts skip, and you can measure it today. Give the free Account Health Snapshot a 60-day bulk file and it scores eight metrics against the phase benchmarks, wasted spend and Top-of-Search ACOS among them, with no email, no account, nothing stored. For a second read on your event roster, budgets, and bid levers before the traffic arrives, I will go through that same file with you on a free 30-minute diagnosis call, and the read is yours to keep either way.