Product Display Ads on Amazon: What They're Called Now, and What They Cost

What "product display ads" means today

"Product Display Ads" was Amazon's name for this format before 2018. Amazon renamed it Sponsored Display, then folded it into one "Display ads" umbrella in the ad console alongside Amazon DSP. If you searched the old name, you landed in the right place — the format still exists, it just has two doors in now instead of one.

One thing worth settling before anything else: Amazon DSP is Amazon's Demand-Side Platform, the programmatic buying system for display, video and audio. It has nothing to do with Delivery Service Partner, the courier franchise program. Same three letters, completely different business — worth knowing before you go looking for the wrong sign-up page.

The two ways to buy display ads on Amazon

Everything that used to be called Product Display Ads now runs through one of two doors. The first is self-serve display ads in the Amazon Ads console — what most people still call Sponsored Display. No minimum spend, CPC or vCPM bidding, and you manage it yourself, the same way you'd manage a Sponsored Products campaign.

The second is Amazon DSP: programmatic buying across Amazon-owned inventory and the open web, run as a managed buy rather than a self-service auction you log into daily. Amazon's own guidance puts the typical DSP entry point around $50,000 USD a month in the US, and that figure moves by country — treat it as an order of magnitude, not a quote, and confirm current minimums directly with Amazon before you plan a budget around it.

• Self-serve display ads require Brand Registry for product-targeted campaigns and run mostly on Amazon-owned inventory, with some off-Amazon reach.
• Amazon DSP is open to vendors, sellers, and non-endemic advertisers who don't sell on Amazon at all, and reaches Amazon O&O plus the open web, Fire TV, and Echo Show.

How the pricing actually works, with real numbers

Self-serve display ads run on cost-per-click or vCPM (cost per 1,000 viewable impressions). There's no minimum budget. You set a daily amount, Amazon suggests a starting bid, and you adjust it against performance — the mechanics are close to Sponsored Products.

DSP works differently underneath: it's a CPM auction, but once you're running enough volume a blended CPC number falls out of the math, and that's the number worth watching against total spend. Here's what that looked like on real accounts: across 30 advertisers we managed in July 2026, the book ran 78.4 million impressions at a $4.00 CPM and a blended $1.42 cost-per-click — roughly $313,600 in tracked media spend. That spend produced 57,137 attributed purchases at a blended $5.49 CPA, which multiplies back out to almost the same $313,600 — the kind of arithmetic check that should hold if the reporting is honest. Of those purchases, 20.1% came from shoppers new to the brand, and the book returned 6.04x ROAS.

Those CPC and CPM numbers move with category, competition, and creative format — video inventory generally costs more per impression than static image. Treat any single "average Amazon display CPC" you find elsewhere as an orientation figure, not a benchmark to hit, since your own auction will land somewhere else entirely.

The mistakes that quietly wreck a display campaign

The most common one: judging display purely on last-click ROAS. Display's job is mostly upper-funnel — introducing the product, staying in front of a shopper who's already searched a competitor. Last-click attribution either undercounts that work entirely or, just as often, overcounts it by crediting a sale that would have happened through search anyway.

We've made that exact mistake ourselves. Early on, a 6x ROAS read as proof the campaign was working — until a matched holdout showed a meaningful chunk of that lift would have converted through search regardless. Last-click attribution can't prove incrementality. It never could. Holdouts and matched controls can, and reconciling DSP against sponsored ads inside Amazon Marketing Cloud is the only honest way to see whether the two are quietly double-counting the same shopper's click.

• Running Sponsored Display and DSP at the same audience without checking overlap — you pay twice for one click.
• Setting vCPM too low, getting starved of inventory, and concluding "display doesn't work" instead of raising the bid and testing again.
• Letting auto-generated creative go stale after a price or image change on the product page, then wondering why CTR dropped.

When the numbers look wrong, here's what to check

If CPC or CPA spikes suddenly, check the attribution window first — conversions keep landing until the window on the column you are reading closes, and Sponsored Display reports on a 14-day click-or-view basis, so a mid-flight dip is often a reporting lag, not a performance drop.

If ROAS looks strong but you're not sure it's real, that's an AMC question, not a console question: pull the overlap between DSP and sponsored ads exposure and see how many "display-driven" purchases were also touched by a search ad in the same window. If a fix doesn't move the number after a full attribution cycle, don't keep running it on hope — roll it back and re-test the next lever instead of the same one twice.

Where Dr. DSP fits

Dr. DSP is Amazon DSP run as a managed product by Fable 5, part of Full Circle — a full-service Amazon management company with $500M+ in managed revenue across 100+ brands. Every change we run carries the evidence behind it, a measurement plan, and a rollback trigger before it goes live, and clients choose their own autonomy level — full approval, supervised, or fully autonomous inside agreed guardrails. There's no published price: a demo, the first 30 days free, and pricing set on the call against real budget and scope, with the full Orbit software suite included either way. If you're testing a few hundred dollars a month, you don't need any of this — the self-serve console is the right tool. If you're already spending enough that a 1% pricing difference is a four-figure question, this is the layer that tells you whether the display spend actually did anything, separate from what it whatever vendor runs it.