If you learned eBay Promoted Listings from a guide written before 2026, most of what you know about the fee model is now wrong, and it is wrong in the direction that costs you money. eBay changed how General campaigns attribute sales, Australian sellers were in the first wave of that change, and the practical effect is that a much larger share of your sales now carry an ad fee, including sales you would very likely have made anyway. This guide covers how the system actually works now, when it is worth paying for, and how to structure campaigns around it. The numbers and structures in it come from a client account we manage: an Australian automotive parts catalogue that has put through $325,620 in ad-attributed sales across eight months under the current rules. The full story of that account is in case study 02. Adgenix runs eBay Promoted Listings alongside Google Shopping as part of ecommerce PPC management.
What changed, and why your old playbook mispredicts your fees
eBay’s General campaign strategy has always been pitched as the safe option: cost-per-sale, no sale means no fee. What changed is the definition of an attributed sale. Under the current model, a fee is charged when any buyer clicks your promoted listing and then any buyer, including a completely different person, purchases that item within 30 days. Australian, UK, and German sellers received this model first, and the US followed in January 2026.
The consequence is mechanical. Under the old direct-attribution logic, roughly half of a promoted item’s sales might be attributed to ads. Under the new logic, sellers report attribution rates of 80 to 90 percent and higher. Your listing only needs one promoted click in a 30-day window for subsequent sales of that item to carry the fee, whether or not the eventual buyer ever saw an ad. For a listing that sells steadily on its own, promotion at a careless ad rate stops being a growth lever and becomes a tax on sales you already had.
This does not make General campaigns bad. It makes indiscriminate promotion bad. The distinction runs through the rest of this guide.
The second change matters just as much for structure: Priority campaigns now hold exclusive access to the top ad slot in eBay search results. General campaigns cannot appear there at any ad rate. The best placement on the platform is only purchasable through the CPC product, which reshapes what each campaign type is actually for.
General vs Priority in 2026: what each one is now for
General (cost-per-sale). You set an ad rate as a percentage of the sale price, eBay controls placement below the top slot, and the fee accrues on attributed sales as defined above. Its strength is unchanged: it is the only sane way to advertise a large catalogue, because it prices thousands of listings into visibility with no per-click risk and no bid management overhead. Its weakness is new: the attribution model means the effective cost of promoting your steady sellers is higher than the ad rate makes it look.
Priority (cost-per-click). You pay per click, with targeting controls and, since the change, exclusive rights to the premium placement. It demands enough sales data to bid sensibly and enough attention to manage, which is why most sellers run it badly or not at all. Run properly on segmented product groups, it is where the efficiency now lives.
In the client account, that split is not theoretical. The launch structure was General coverage across the catalogue and it built the channel: roughly three quarters of all ad sales at an 8.5x return on ad fees. The rebuilt structure runs Priority campaigns with Smart Bidding on segmented product families, and those campaigns return 14.9x, nearly double the efficiency, at an average CPC around fifty cents. General built the base. Priority is where the margin is.
The selective-promotion maths every guide skips
The question is never whether Promoted Listings work. It is whether the fee on a given listing buys incremental sales or taxes organic ones. The test is simple enough to run in a spreadsheet.
For each candidate listing, take its trailing sales rate without promotion. Estimate the attributed share under the current model honestly, meaning assume most of its sales will carry the fee once it is promoted, not half. Then ask what sales uplift the promotion would need to generate for the incremental margin to exceed the fee on the whole attributed base. A listing already selling five a week at a healthy organic rank needs a large uplift to justify a 10 percent ad rate on what is now nearly all of its sales. A listing with strong conversion but weak visibility needs almost none, because the fee is buying sales that genuinely were not happening.
That is the practical rule: promote items whose problem is visibility, not items whose sales you simply want to count in the ads dashboard. The attribution change made this arithmetic mandatory. Before it, sloppy promotion was mildly wasteful. Now it is systematically expensive, and the reported return in Seller Hub will still look respectable while it happens, because the dashboard counts the attributed organic sales as wins.
One number worth internalising from the client account: eBay ad fees were 3.4% of eBay sales. That is the figure to manage, fees against the whole channel, not just the flattering per-campaign return, because under current attribution the two tell different stories.
Breadth beats heroes: the position most guides get backwards
Almost every Promoted Listings guide says the same thing: promote your best sellers, cap promotion at your winners, keep the rest organic. For hero-product catalogues that advice holds. For long-tail catalogues it is exactly backwards, and most Australian ecommerce sellers on eBay, in parts, components, accessories, and variant-heavy categories, run long-tail catalogues.
The client account advertises 48,756 listings. The single best performer contributes barely twelve hundred dollars of the $325K total. There is no hero to promote. The channel’s engine is thousands of vehicle-specific items each selling a handful of units, and the only structure that monetises that shape is broad coverage priced by product group, with the ad rate tiered to margin. Promoting only the winners would have captured a rounding error of the demand.
The synthesis of this section and the last one is the actual strategy: broad General coverage across the long tail, where each individual listing’s organic sales are too thin for the attribution trap to bite, and selective, data-driven Priority campaigns on the product families with proven conversion, where the premium placement earns its clicks. Breadth where attribution risk is low. Precision where placement is now exclusive.
Test in tranches, then segment: the sequencing that worked
The account’s structural history is the useful part, because the sequence is repeatable.
Launch was broad General coverage in November 2025, tiered ad rates by product group. That validated the channel fast, with the first full month clearing $31K in attributed sales, and then plateaued in the mid-thirties to mid-forties per month, which is what broad coverage does: it finds the demand ceiling of its structure.
The second move was a deliberately short structured test. Two weeks, two competing hypotheses run as separate campaigns: a hero-style top-100 priority set against long-tail coverage tiers. Both were ended on schedule with their answer in hand. The top-100 test returned 32x and the long-tail tiers 25x, which proved the efficiency ceiling of the launch structure was roughly double, and, just as usefully, that neither test structure was the final answer, because hero promotion cannot scale across a catalogue with no heroes.
The rebuild segmented the catalogue by vehicle make and region, Australian and American, Asian, German, and European, on Priority campaigns with Smart Bidding. Segmentation by make matches how parts buyers actually search, and it gives the bidding algorithm coherent product sets instead of a 48,000-listing blur. The result was the 14.9x figure above, and the month this guide draws on, the account’s twentieth day of July had already outsold every prior full month.
The generalisable sequence: launch broad to find the demand, test structures in short deliberate tranches with scheduled end dates, then rebuild around the segmentation your buyers already use. What killed accounts we audit is the opposite pattern, launching everything at the suggested ad rate and never running a controlled test at all.
Where eBay sits next to Google Shopping for Australian brands
The two channels are not competitors for the same dollar. Google Shopping captures buyers searching the open web with a store-shaped intent. eBay Promoted Listings capture buyers who start and finish inside the marketplace and were never going to reach your website at any bid. For categories with heavy marketplace-native demand, automotive parts being the archetype, the eBay budget reaches people the Google budget structurally cannot.
The proof standard should be the same one we applied: website revenue held at its established level through the entire marketplace expansion, verified in GA4, which means the marketplace channel was additive rather than a reshuffle of existing customers. If you cannot demonstrate that increment, you have not expanded, you have redistributed. The feed and account fundamentals that make Google Shopping work, covered in the feed optimisation framework, have direct eBay equivalents: item specifics, titles, and category placement do for eBay search what the product feed does for Google.
For the broader question of running paid across a whole ecommerce operation, website channels and marketplaces as one system, that is the scope of our ecommerce PPC management engagement.
Frequently asked questions
Do eBay Promoted Listings work in Australia? Yes, and Australia runs the same current attribution model as the US and UK, having received it earlier. The account referenced throughout this guide is Australian: $325,620 in ad-attributed sales across eight months, and eBay ad fees were 3.4% of eBay sales. The qualifier is structure. Indiscriminate promotion at suggested ad rates underperforms badly under the current attribution rules.
What is a good ROAS for eBay Promoted Listings? Blended across a whole account, high single digits is strong for General-led structures. Well-run segmented Priority campaigns can run materially higher, the referenced account holds 14.9x on its Smart Bidding set. Treat dashboard ROAS with scepticism under current attribution though, since it counts attributed organic sales. Fees as a percentage of total channel sales is the honest health metric.
Should I use General or Priority campaigns? Both, for different jobs. General for broad catalogue coverage, priced by margin tier. Priority for proven, segmented product families, because it now holds exclusive access to the top search placement. Small catalogues with a few strong sellers can run Priority-led. Large long-tail catalogues cannot skip General.
What ad rate should I set? Work backwards from margin, not from eBay’s suggestion. The suggested rate optimises for eBay’s fee revenue, not your profit. Tier rates by product margin group, and under current attribution, assume the fee will apply to most of a promoted item’s sales, not a fraction.
Does promoting listings hurt organic rank if I stop? eBay does not penalise stopping promotion, but a listing whose sales velocity was carried by ads will lose the velocity contribution when they stop, and velocity feeds eBay’s search ranking. That is another reason to promote for visibility problems rather than to inflate items that already rank.
Where Adgenix sits
Adgenix runs paid media for Australian ecommerce brands as one system: Google Ads, Shopping, and Performance Max on the website side, with marketplace advertising on eBay and Amazon where the category’s demand lives there. Senior operator on every account, flat fees, and the same measurement standard everywhere, channel growth only counts when the rest of the business holds.
If your catalogue is long-tail and your category has marketplace-native demand, the Free Revenue Leak Report covers the website account you have and the marketplace opportunity you might be leaving unworked. Written inside 48 hours, yours to keep either way.
