Holiday Amazon ads: ranking, remarketing, video and listing-image tests

Illustrative studio camera for advertising production, not a campaign screenshot or measured result.
A 2025 holiday advertising tutorial by Amazon PPC manager Chris Rawlings argued that higher seasonal sales need not come at the expense of profit. Published on October 27 and circulated to sellers again on November 11, it described preparation before the rush, low-bid campaigns, tailored discounts, remarketing, video testing and stronger listing images. The figures and recommendations below are his dated examples, not guaranteed outcomes or a current statement of Amazon policy.
Build organic rank before clicks become dearer
Rawlings warned that holiday clicks can become much more expensive even while conversion improves. He highlighted the two weeks after Black Friday as especially important. The November email described a broader three-week peak. His advice was to establish organic positions before this period, rather than first trying to buy rank when competitors are spending most heavily.
His ranking setup was a single-keyword, exact-match Sponsored Products campaign, using a top-of-search bid adjustment where affordable. If that placement was too costly, he suggested using the placement that performed best within the budget. Keyword-targeted Sponsored Brands video appearing in search was another option he said could support ranking when Sponsored Products clicks were unaffordable.
Low-bid scavenger campaigns
The scavenger setup uses Amazon-suggested keywords in a Sponsored Products campaign but bids at about one-tenth of the suggested amount or less. The intention is to find cheap auction gaps when competitors run out of stock, exhaust budgets or are absent at particular times. Rawlings stressed that volume is usually limited. His example spent about $16 for roughly $1,500 in sales, around 1% advertising cost of sales, against a 21% account average. He presented it as a low-cost supplement, not the whole growth plan.
Tailored discounts for people already close to buying
Brand Tailored Promotions were described as a different model from paying for every click. In his 2025 explanation, Amazon supplied the promotion's exposure while the seller funded the chosen customer discount. He compared a 10%, 15% or 20% discount with a fixed promotional cost on the resulting sale. That comparison was his explanation of this offer, not a calculation of every fee or the seller's final profit.
The demonstrated route was Advertising, Brand Tailored Promotions, then Create a Tailored Promotion. Options shown were new-customer acquisition, retention, re-engagement and cross-selling. He prioritised the 90-day brand-cart-abandoner audience within acquisition: people who had put a product in their cart but not checked out. His example showed an audience of 60,000, followed by naming the promotion, choosing a discount and setting its budget. He cited several campaigns with reported sales between about $5,000 and $22,000 at a 15% discount, while also supporting ordinary holiday deals and coupons.
Separate viewers from previous buyers in remarketing
Sponsored Display remarketing was his next overlooked tactic. Gift buyers may choose a product they already know from an earlier purchase. He showed views-remarketing and purchase-remarketing examples with ACoS of 14.4%, 13.9%, 6% and 5.8%, presenting them as account examples rather than a universal benchmark. ACoS here means advertising spend relative to attributed ad sales, not total business profit.
The tutorial created a Sponsored Display campaign in Campaign Manager, added the advertised product and used its default image as a starting point, with video or custom imagery as tests. Rawlings suggested roughly 5% to 10% of the product's overall budget for this activity, conversions as the initial optimisation goal and page visits as a possible later test. He warned that reach used viewable-impression pricing rather than PPC and was generally unsuitable for his smaller-brand examples.
He removed automatically added audiences and selected the advertiser's own viewers or buyers. The windows shown extended to 90 days for views and 365 days for purchases. He recommended keeping views and purchase campaigns separate and making the lookback windows explicit. His starting-bid formula was target ACoS multiplied by conversion rate multiplied by product price, with bids adjusted as results arrived.
Test video, including AI-generated versions
Rawlings argued that holiday shoppers are more open to discovering products and that video can explain a product, revive a weak launch or help a listing overcome poor performance. Sellers without footage could try the AI video generator in Amazon's advertising console. He recommended generating several versions and split-testing them, rather than assuming AI or human-made video would always win.
The examples included AI videos with 18% and 21% ACoS in an account averaging above 30%; comparisons of 33% versus 67% and 20% versus 37% for AI and human-produced versions; other AI results of 25% and 21%; and one video at 3% in an account averaging 18%. These were selected campaign outcomes. The wider claim was to test economical creative alternatives, not that every generated clip would produce those figures.
Win more of the available clicks
His final method, which he called clicks maxing, focused on making the main listing image stand out within a finite pool of search impressions. He cited the bright packaging of Angry Orange and claimed a twentyfold sales increase after its acquisition. In his own examples, a stronger image sometimes raised click-through enough to compensate for a lower conversion rate, bringing more sessions, sales and organic ranking.
Examples involved more visible packaging; showing a hand using the product; better-lit tea and an origin claim supported by the seller's keyword data; replacing a photograph with a rendering; showing serving count and ingredients; adding the product beside its packaging; improving contrast; showing the grip; and making key features clearer on a label. Reported gains ranged from about 50% or 66% more clicks to doubling, tripling or, in one case, sixfold click-through. A candle example added a flame, flowers, brightness and contrast. These are his test reports, not proof that the same edits suit another listing.
His checklist also included showing the intended user, primary use, ingredients, product alongside packaging, a useful measurement, colour or style variants, extra angles, the product in action, a string tag and brighter colours. He discussed the tension with Amazon's plain-white-background guidance and argued that informative, non-deceptive edits would not be penalised, citing an Amazon Basics image. That assurance was his opinion in the tutorial, not platform approval of a particular seller's image.
The video closed by promoting a fuller holiday workshop and the presenter's PPC-management service. Those were 2025 commercial invitations. The supplied tutorial gives the methods above but not the separate workshop's complete playbook or every linked follow-up lesson. The studio-camera photograph is illustrative, not a campaign screenshot or evidence of any reported return.