Amazon Ads on October 15, 2026 published a three-minute-43-second episode of its Ad Lab video series on YouTube comparing fixed bidding with dynamic bidding for Sponsored Products campaigns. The video opens by promising to "test Amazon ad strategies and break down the experiments," but the comparison it presents is a hypothetical pair of campaigns, and no performance figures from any experiment appear on screen or in the description.

In Short

Amazon made a short video explaining two ways to bid on its search ads: one where your bid never moves, and one where Amazon can push it up or down depending on how likely a shopper is to buy. It matters for anyone paying for clicks on Amazon, because the second option lets Amazon spend more per click on your behalf. The video says neither choice is always better, gives four numbers to watch, and does not show any test results or say how high Amazon can raise a bid.

What the video covers

The episode, titled "Should you use dynamic bidding or fixed bidding for Sponsored Products campaigns?", sits on the official Amazon Ads YouTube channel, which the page capture shows with 29,400 subscribers. When the capture was taken, the video had 55 views, one like and no comments. The page lists the publication date as 15/10/2026. No presenter is named in the description or in the transcript.

The video begins with a question and a framing device. "Should you use dynamic bidding or fixed bidding for your Amazon ads? Welcome to the Ad Lab, where we test Amazon ad strategies and break down the experiments," the presenter says. The setup then shifts to the option most advertisers have already seen. "If you've ever launched a Sponsored Products campaign, you've probably seen the option for dynamic bidding. It lets Amazon adjust your bids based on the likelihood of a conversion."

The description is more precise about the menu than the video itself. According to Amazon, advertisers running Sponsored Products campaigns "have three bidding strategies to choose from: fixed bidding, dynamic bidding (up and down), and dynamic bidding (down only)." The episode compares only the first two. "Amazon also offers dynamic bidding down only. But for this walkthrough, we're focusing on up and down," the presenter says at 0:54.

That choice matters. Down-only bidding cannot increase what an advertiser pays; up-and-down bidding can. By setting the down-only option aside, the video concentrates on the one strategy where Amazon is allowed to bid above the number the advertiser typed in.

Two identical campaigns, on paper

The core of the episode is a thought experiment rather than a test. "To understand the difference, imagine two identical Sponsored Products campaigns. Campaign A uses fixed bidding. Campaign B uses dynamic bidding up and down," the presenter says at 0:41. According to the video, holding everything else constant "makes it easier to isolate the impact of the bidding strategy."

The description repeats the claim, listing among the things viewers will learn "How to set up two identical Sponsored Products campaigns to isolate the performance impact of each bidding strategy." Neither the video nor the description reports what happened when such campaigns were run. There are no CPC figures, no conversion rates, no sample sizes, no product categories and no dates for any experiment. The word "experiments" in the opening line is not matched by any published result.

How each strategy enters the auction

On fixed bidding, the explanation is direct. "With fixed bidding, your bid stays exactly where you set it. If you've bid a dollar, Amazon enters the auction at a dollar every time," the presenter says at 1:06.

Dynamic bidding is described in terms of conversion likelihood. "Dynamic bidding works differently. Amazon can raise your bid when a click looks more likely to convert, and lower it when it doesn't," according to the video. The next line puts it more bluntly: "So instead of setting one fixed number, you're giving Amazon permission to react in real time. Pay more when a click looks valuable. Pay less when it doesn't."

Earlier, at 0:23, the presenter summarised the arrangement in plainer terms still: "You're basically telling Amazon, raise or lower my bid when you think it makes sense. Some advertisers love that." The following line acknowledges the other camp: "Others want to keep both hands on the wheel."

What the video does not say

The episode never states how far Amazon can raise a bid under the up-and-down setting. That limit is the single most consequential number for an advertiser deciding whether to grant the permission the video describes.

PPC Land's Sponsored Products explainer, published on September 20, 2026, set out the figures from other sources. Amazon's current guide sets the up-and-down increase at up to 100% for all placements, while earlier versions allowed 100% at top of search and 50% elsewhere. Placement adjustments, configured separately at campaign level, can lift the base bid by up to 900% for top of search, rest of search and product pages. Combined, a $1.00 bid with a 900% placement adjustment and up-and-down bidding can resolve to a theoretical $20.00 per click. The Ad Lab video's own example uses exactly one dollar and stops at the fixed case.

The video also does not mention placement adjustments at all. Nor does it describe how the bid relates to the price an advertiser is actually charged. In a second-price auction, the winner traditionally pays the runner-up's bid plus a small increment, so a higher bid does not automatically mean a higher cost per click. Whether that still holds on Amazon is now contested in federal court.

The auction pricing backdrop

The video arrives six weeks after the Federal Trade Commission and 22 states sued Amazon over about $20 billion in alleged hidden ad surcharges on August 31, 2026, in the U.S. District Court for the Western District of Washington. The 181-page complaint alleges that Amazon applied an undisclosed soft reserve after its generalized second-price auction, capped only by the winner's own bid.

The figures in that case bear directly on the dynamic bidding question. According to the complaint, the share of clicks charged the advertiser's full bid rose from about 4% in late 2020 to 79.1% in 2024. The mechanism allegedly reached Sponsored Products second-price auctions in mid-2019. The complaint also cites an internal observation that dynamic bid adjustments, which can raise a bid by up to 900% for certain placements, mean advertisers may not know what bid they actually entered. Amazon, the complaint says, was concerned that advertisers would respond with bid shading, the practice of bidding below true value that is common in first-price auctions.

Amazon denies overcharging. It has said that inflation-adjusted average cost per click was flat from 2019 to 2024, that about 92% of selected Sponsored Products ads in 2024 were not the highest bid, and that its Support Center now says the price charged can exceed the runner-up bid but never the advertiser's maximum. The difference regulators describe, between the auction price and the amount charged, is what PPC Land has termed the Amazon ad surcharge.

If the price paid tracks the bid in most auctions, then a strategy that lets Amazon raise the bid is close to a strategy that lets Amazon raise the price. The Ad Lab episode does not address that link. It treats a higher bid as a lever for winning more valuable clicks, without discussing what the advertiser is then charged.

The four metrics

The middle section of the video turns to measurement, and here it offers its most concrete guidance. First comes a caution about time. "The question is, will those adjustments actually improve performance? If you're comparing these bidding strategies, don't judge performance after a single date," the presenter says at 1:35. The transcript reads "single date," which is most likely an automatic transcription error for "single day." The next line supports that reading: "Let them run long enough to reveal a pattern, ideally a few weeks."

The four metrics follow in order. "Here are four metrics I pay close attention to. First cost per click. How did CPC change? Second conversion rate," according to the video. The presenter then links the first two: "If dynamic bidding is working as intended, conversion rate may improve even if CPC increases."

The third and fourth metrics are efficiency ratios. "Third, advertising cost of sale or ACOS," the presenter says at 2:06, followed by: "Are we becoming more efficient or just spending more money and finally return on ad spend or Roas?" The justification for watching ROAS comes at 2:17: "Because higher bids aren't automatically bad if the return grows faster than the cost."

ACOS and ROAS are two sides of the same calculation. ACOS divides ad spend by attributed sales; ROAS divides attributed sales by ad spend. A 25% ACOS corresponds to a ROAS of 4. Listing both as separate metrics gives the comparison four headings, though they measure one relationship.

Attribution caveats left out

None of the four metrics is independent of how Amazon counts a conversion. Conversion rate, ACOS and ROAS all depend on attribution windows, and Amazon changed those this year. In January, PPC Land reported that Amazon moved view-based conversion metrics to a shopping-signal model across 28 markets for Sponsored Brands, Sponsored Display viewable CPM and DSP Store campaigns, effective January 1, 2026, with windows shorter than the previous 14-day standard. That change did not cover Sponsored Products click attribution directly, but it shows that the denominators behind these ratios are set by the platform. The video does not raise attribution.

When the video says each strategy fits

According to the episode, dynamic bidding depends on data volume. "Dynamic bidding can be helpful when your campaign has enough insights for Amazon to recognize which clicks are more likely to convert," the presenter says at 2:22. What counts as enough is not defined - no minimum number of clicks, conversions or days is given.

The expected pattern when it works is described as a trade. "When it works well, you may see higher CPCs, but stronger conversion rates and better return. That's the trade off," according to the video. The framing then moves from cost to judgement: "You're not just asking, did I spend more? You're asking, did I spend smart?"

Fixed bidding is presented through control. "Fixed bidding has one big advantage, control," the presenter says at 2:40, continuing: "You decide exactly what you're willing to pay, and Amazon sticks to it. That can be useful if you're testing a new product, protecting a tight budget, or you simply want a predictable spend with fixed bidding. You may give up some automation, but you know exactly what you're bidding every time."

The wording is careful. Fixed bidding guarantees what an advertiser is "bidding," not what the advertiser pays. Under the pricing rules described in the FTC complaint, the two have converged for most clicks, but they are not the same by definition.

The conclusion declines to pick a side. "So should you use dynamic bidding or fixed bidding? It depends on your objective. If you want Amazon to optimize bids based on conversion likelihood, dynamic bidding may be the better fit," the presenter says at 3:08. "If predictable spend and complete control are your priority. Fixed bidding may be the better choice." The closing lines: "Neither strategy is universally better. The right one depends on how much automation and how much control you want in your campaign."

Where the transcript and the page differ

The transcript provided and the YouTube page capture disagree in a few places, none of which changes the substance.

The description lists seven chapters: 0:00, 0:41, 1:05, 1:35, 2:22, 2:45 and 3:08. The transcript groups the content into six, folding "Setting up two identical campaigns to compare bidding strategies" and "How fixed bidding and dynamic bidding work in Sponsored Products auctions" into a single chapter starting at 0:41. In the description, those appear as separate chapters at 0:41 and 1:05.

The chapter "Why fixed bidding gives you control over Sponsored Products spend" begins at 2:45 according to the description. In the transcript, the line "Fixed bidding has one big advantage, control" is timestamped 2:40 and sits under the previous chapter, while the fixed-bidding chapter heading appears before the 2:49 segment. The player in the capture also shows "3:26 Choosing the right bid..." as the current chapter label at one point, while the description places that chapter at 3:08; the capture appears to have been taken mid-playback, so the label likely reflects playback position.

The description says advertisers have three strategies; the spoken script mentions the third, down only, once and then sets it aside. The total running time shown in the player is 3:43, while the last transcript segment starts at 3:30.

Why it matters for advertisers

Amazon has published Sponsored Products guidance before. In March 2025, PPC Land covered Amazon's Sponsored Products best practices documentation, which described dynamic and fixed bidding in similar terms, recommended a $10 starting daily budget and noted placement adjustments of up to 900%. The Ad Lab episode covers less ground than that document. It does not mention placement adjustments, the 100% cap on dynamic increases or budget settings.

What the video adds is a measurement frame: run both strategies for weeks, then compare CPC, conversion rate, ACOS and ROAS. That frame is reasonable as far as it goes. Its limit is that every input to it - the bid adjustment, the auction price, the conversion attribution - is calculated by Amazon, and the episode presents no data of its own showing what an advertiser might expect to see.

Amazon has been shifting optimization decisions toward its own systems across ad products. In April 2025, PPC Land reported that Amazon applied conversion-focused optimization automatically to existing Sponsored Display Reach and Page Visit campaigns, with ROAS and ACOS as priority metrics. In October 2025, it offered reserve share of voice for branded search, a fixed-price route to Sponsored Brands top-of-search placements that sidesteps the auction for those keywords. Dynamic bidding sits at the other end of that spectrum: the advertiser sets a base bid, and Amazon decides, auction by auction, how far above or below it to go.

The video's own summary of the choice - "how much automation and how much control" - is accurate. Left unanswered is how much more an advertiser can be charged once control is handed over, a question now before a federal judge in Seattle.

Timeline

Summary

Who: Amazon Ads, through an unnamed presenter in its Ad Lab YouTube series, addressing advertisers who run Sponsored Products campaigns.

What: A 3:43 video comparing fixed bidding, where the bid never changes, with dynamic bidding (up and down), where Amazon raises or lowers the bid based on conversion likelihood. It proposes running two identical campaigns for a few weeks and comparing CPC, conversion rate, ACOS and ROAS, and concludes that neither strategy is universally better. It presents no experiment results and does not state how far Amazon can raise a bid.

When: Published on YouTube on October 15, 2026.

Where: The Amazon Ads YouTube channel, at https://www.youtube.com/watch?v=CfoPTLaOVXo, covering Sponsored Products auctions on Amazon.

Why: Dynamic up-and-down bidding gives Amazon permission to bid above the advertiser's set amount. With the FTC and 22 states alleging that Amazon charged the full bid on 79.1% of clicks in 2024, how a bid translates into a price paid is a live question that the video does not address.