Amazon Sponsored Display ads are self-serve display ads that reach shoppers in two ways: by what they are viewing (contextual targeting of specific products or whole categories) or by who they are (audiences, including remarketing to people who viewed or bought your products). They are priced per click or per thousand viewable impressions, and they deserve budget only when their click-attributed orders clear the ACOS target your Sponsored Products campaigns already answer to.
I run Sponsored Display on the accounts I manage, and in audits it is the ad type I most often find spending with nobody reading it: no search term report forces a weekly look, and its dashboard flatters it. This guide assumes you have decided it deserves a test. If not, start with which Amazon ad type should get your next dollar. What follows: the targeting menu, the two cost bases, when to wait, how I structure it, and the bar I judge it against, on real numbers.
Sponsored Display is now part of Amazon's display ads
Amazon has folded Sponsored Display into its wider display ads offering, and its own product page now says Sponsored Display is known as display ads. Campaigns you already have carry on as before, and anything new gets built as a display campaign in the console. The self-serve mechanics, targeting and per-click or per-impression pricing, carry over.
I still call it Sponsored Display here because that is what sellers search for and what most live campaigns are named. The rename moved the button, not the arithmetic.
Sponsored Display targeting: contextual vs audiences
Sponsored Display targeting comes in two families. Contextual targeting follows what the shopper is looking at: specific product pages (yours or a competitor's) or whole categories. Audience targeting follows who the shopper is: people who viewed your products, people who bought them, or Amazon's prebuilt interest, lifestyle, life-event and in-market segments. Each needs its own campaign.
| Approach | Who it reaches | The job I hire it for | Temperature | My testing order |
|---|---|---|---|---|
| Views remarketing | Shoppers who viewed your products and left | Winning back people who already considered you | Warmest | First |
| Product targeting | Specific ASINs, a rival's or your own | Conquesting rival pages; defense and cross-sell on yours | Warm | Second, seeded from ASINs already converting in Sponsored Products |
| Purchases remarketing | Past buyers of your products or complementary ones | Reorders and add-ons | Warm, but often returning anyway | Only for consumables or real add-ons |
| Category targeting | Products across a whole category | Wider detail-page presence | Lukewarm | After product targets prove out |
| Amazon audiences | Prebuilt interest, lifestyle, life-event and in-market segments | Prospecting shoppers who have never seen you | Coldest | Last, as an agreed write-off test |
Contextual targeting: product pages and categories
Contextual targeting places your ad on detail pages and shopping results tied to the products or categories you pick. Product targeting is precise, so I start there. Category targeting earns a test only after product targets show which corners of the category actually buy.
I seed rival ASINs from evidence: the ones where Sponsored Products product targeting, or the substitutes group in your auto campaigns, already converts. Choosing and pricing them is its own topic, covered in my guide to Amazon competitor targeting. On your own pages, product targeting keeps rivals out of the ad slots and shows shoppers your other products. A single-product brand gets the defense only.
Audience targeting: Sponsored Display retargeting and prospecting
Audience targeting is what most sellers mean by Amazon retargeting ads. Views remarketing follows shoppers who looked and left, purchases remarketing follows past buyers, and Amazon's prebuilt segments reach shoppers who have never seen you. Those sit at very different temperatures and should never share a campaign or a verdict.
Views remarketing is the strongest case, with a catch: the audience is built from your own detail-page traffic, so a listing with modest traffic gets a modest pool. Remarketing recycles demand; it does not create any. Purchases remarketing needs a reason to buy again, like a consumable or a genuine add-on. Prebuilt audiences are prospecting, and I test them only with a budget agreed in advance as a possible write-off. Display placements can also run beyond the Amazon store into third-party apps and websites, which widens reach and loosens intent at the same time.
CPC vs vCPM: how the cost basis changes the way you read ACOS
On cost per click, Sponsored Display bills like Sponsored Products: no click, no charge. On vCPM, you pay for every thousand viewable impressions whether anyone clicks or not, so ACOS depends on click-through rate, not only on bid and conversion rate. Before comparing a vCPM campaign to anything, convert its spend into cost per click and cost per click-attributed order.
| CPC (cost per click) | vCPM (per 1,000 viewable impressions) | |
|---|---|---|
| What you pay for | A click | Every thousand viewable impressions |
| Spends with zero clicks? | No | Yes |
| Per-click bid formula applies? | Yes: revenue per click × target ACOS | No: the bid prices impressions, not clicks |
| How I use it | The default for anything judged on orders | Only for an agreed reach goal, graded on total sales and TACOS |
On vCPM, your effective cost per click is the price of a thousand impressions divided by the clicks those impressions produced, so a campaign that looks cheap per thousand can be expensive per click. That is why my Audit Dashboard reprices Sponsored Display CPC rows with the same revenue-per-click logic it uses for Sponsored Products (the method is in my guide to setting Amazon PPC bids) but holds impression-priced rows back for manual review. A per-click formula applied to a per-thousand price produces a confident-looking wrong number.
Cost basis is half the reading problem; attribution is the other half. Sponsored Display can credit a sale to a shopper who only saw the ad, and the headline sales figure can include those orders. Sponsored Products counts clicks, so compare the two on click-attributed orders wherever the reporting separates them. The gap is widest on remarketing, whose audience was already leaning toward coming back.
When Sponsored Display is not worth the budget yet
Sponsored Display is not worth the budget yet when Sponsored Products still leaks, when your listing converts below your phase benchmark, when converting Sponsored Products campaigns are still running out of budget, or when your detail pages get too little traffic to build an audience worth remarketing to. Each of those is cheaper to fix than to advertise around.
- Sponsored Products still leaks. If zero-order spend sits above your phase benchmark (under 15% is the Profitability bar), stop the waste first. The free Wasted Spend Finder ranks it from your bulk file.
- The listing is the bottleneck. Retargeting sends shoppers back to the page that already failed to convert them. If your conversion rate sits under the phase benchmark (Growth asks for more than 8%, Profitability more than 10%), run the clicks-but-no-sales diagnostic first.
- Proven campaigns are budget-capped. A Sponsored Products campaign converting at target and out of money by afternoon is a better home for the next dollar. My out-of-budget guide separates genuine caps from leaks.
- The remarketing pool is thin. Little detail-page traffic means a small audience. Grow search traffic first; the audience follows.
- Nobody will review it. Sponsored Products has a search term report that drives its harvest-and-negate loop. Sponsored Display has none, so without a weekly read of its targets, it drifts.
And one gate comes before all of it: for sellers, Amazon lists Brand Registry enrollment as a requirement, so an unregistered brand waits regardless.
How I set up Sponsored Display campaigns
I set up Sponsored Display as one job per campaign: one targeting approach, one cost type, and one group of closely related products, named so the job reads straight off the campaign name. Mixing remarketing with prospecting, or clicks with impressions, produces an ACOS that describes nothing, and nothing you can act on.
- One approach per campaign, named for its job. A name like "SD | Views remarketing | CPC | Product line" tells anyone reading the bulk file what the campaign is for.
- CPC unless there is a written reason. vCPM only for a reach goal agreed up front, graded on total sales and TACOS rather than its own ACOS.
- Products that belong together. Products in an ad group share bids and targets, so I group by category and price point. A cheap accessory parked beside the hero product leaves one of the two bid wrong.
- Creative that inherits the listing. Amazon can build the ad from your detail page, or you can supply your own assets. I start with the generated version: if your image, title, price and rating cannot win the click, custom artwork will not rescue the offer.
How I judge whether Sponsored Display earned its budget
I judge Sponsored Display with three questions, in order. Has each target had a fair trial? Do its click-attributed orders clear the phase ACOS target? Would the same dollar have done more in Sponsored Products? A campaign has to pass all three to keep its budget, and view-attributed orders are never allowed to rescue a failure.
- The fair trial. A target with no click-attributed orders has not failed until it has spent what one order normally costs you: average clicks to convert times cost per click, the fair-judgment threshold from my n-gram analysis guide. It is a dollar figure, so it works on vCPM targets too.
- The ACOS bar. Click-attributed ACOS against your phase's account ACOS ceiling from my phase benchmarks: under 40% in Awareness, 30% in Market Share, 25% in Growth, 15% in Profitability. No discount for being a different ad type. If it needs view credit to get under the bar, it is over it.
- The opportunity cost. Even a campaign under the bar loses to a Sponsored Products campaign sitting budget-capped at a better ACOS. Money goes to the best available return.
Remarketing gets one extra check. When I scale it up or down, I watch total sales and TACOS, not its own ACOS. If total sales do not move when remarketing spend moves, it was renting credit, not creating orders. That is the same logic as reading ACOS and TACOS as a pair.
A worked example: the Sponsored Display bar on real numbers
On the sample dataset I grade across these guides, one ad-attributed order costs $8.77 and brings in $26.82. Multiply that $26.82 by each phase's ACOS ceiling and you get the most a click-attributed Sponsored Display order may cost: $10.73 in Awareness, $8.05 in Market Share, $6.71 in Growth and $4.02 in Profitability.
Its toplines: 45,672 clicks, 4,140 orders, $36,303 spent and $111,058 back in ad sales. From those: $0.79 per click, 9.1% of clicks converting, $26.82 per order and a 32.7% ACOS. The $8.77 is spend divided by orders.
| Phase | Account ACOS benchmark | Most a click-attributed order may cost | Click conversion rate needed at $0.79 per click |
|---|---|---|---|
| Awareness | < 40% | $10.73 | 7.4% |
| Market Share | < 30% | $8.05 | 9.8% |
| Growth | < 25% | $6.71 | 11.8% |
| Profitability | < 15% | $4.02 | 19.7% |
Two things stand out. First, the bar climbs fast. This account's Sponsored Products traffic converts at 9.1%, so a display campaign paying the same average click for the same average order and merely matching that rate clears only the Awareness bar. Growth demands 11.8%, which is why warm audiences go first. Second, $8.77 doubles as the fair-judgment threshold (roughly eleven clicks at about $0.79 buys one order): a target that spends $8.77 with no click-attributed order has had its trial.
On vCPM the bar is the same; only the route changes. Divide spend by click-attributed orders and hold the result against your phase ceiling. Worked backwards, a thousand viewable impressions is worth at most that ceiling times the click-attributed orders they produce, so halving the order rate halves their value. The audit demo shows Sponsored Display reported beside Sponsored Products on a real anonymized account, with its spend kept out of the wasted-spend split, since there are no search terms to test it against.
Frequently asked questions
What happened to Sponsored Display in the Amazon Ads console?
Amazon moved it under its broader display ads offering, and its product page now says Sponsored Display is known as display ads. Campaigns already running keep running, and new ones are created by choosing Display when you build a campaign. Contextual and audience targeting, priced per click or per thousand viewable impressions, still define the self-serve version.
Should I use CPC or vCPM for Sponsored Display?
Use cost per click for anything you plan to judge on orders, which for most private-label sellers is everything: you pay only when someone clicks, and the per-click bid logic from Sponsored Products carries over. Choose vCPM only for a stated reach goal, accept that it bills for impressions nobody clicked, and grade it on total sales and TACOS rather than its own ACOS.
What is a good ACOS for Sponsored Display ads?
The same ACOS your account needs in its current phase, measured on click-attributed orders. My phase benchmarks set that ceiling at 40% for an account in Awareness, then 30%, 25% and finally 15% as it moves through Market Share, Growth and Profitability. Sponsored Display gets no allowance for being a different ad type, and a campaign that only clears the bar once view-attributed sales are counted is over it.
Why is there no search term report for Sponsored Display?
Because Sponsored Display targets products, categories and audiences rather than the queries shoppers type, so there are no search terms to report. That removes the harvest-and-negate loop that keeps Sponsored Products honest, which is why Sponsored Display needs a deliberate weekly review of its targets, spend and click-attributed orders instead. It also means waste tools built on search terms cannot see its spend.
Should I retarget shoppers who already bought from me?
Only if they have a reason to buy again. Purchases remarketing earns its place for consumables that run out and for catalogs with genuine add-ons, where a past buyer is the warmest prospect you have. For a durable, one-time purchase, it pays to remind people about something they already own. Either way, watch total sales when you scale it, because loyal buyers often come back without the ad.
Sources
- Best practices for your Sponsored Products ads, Amazon Ads
Check the foundation before you fund a display test
Every rule here grades Sponsored Display against a Sponsored Products account that already works, so confirm that part first. Drop a 60-day bulk file into the free Account Health Snapshot and you get eight metrics graded against your phase benchmarks in about a minute: no email, no account, nothing stored. If the foundation passes and you want a second opinion on where Sponsored Display fits, or whether it should wait, book the free 30-minute diagnosis call. I will go through your numbers with you and give you a straight answer, hire me or not.