34 brands · 69 brand-markets in 13 marketplaces · December 29, 2025 to November 16, 2026 · Off Amazon seen in UK, DE, FR, IT, ES, NL, PL, CA and US
01Sponsored Products now run off Amazon too, and our European accounts got it from September
Amazon shows Sponsored Products ads on other websites and apps as well as on Amazon. The placement report lists those clicks as Off Amazon, next to top of search, product pages and the rest of Amazon. You can raise bids for the Amazon placements. There is no bid adjustment for Off Amazon.
What you get instead is a campaign setting for ads served off Amazon, with two choices: “maximize reach” and “minimize spend”. Trade write-ups describe it, for example on intentwise.com and feedvisor.com. In the Amazon Ads API the two values read MAXIMIZE_REACH and MINIMIZE_SPEND, and that is what our app stores. Section 06 shows what the second one did.
Our US campaigns had Off Amazon traffic before our data starts. In our European accounts the first impressions came between September 19 and October 19. By November 16, 23 of our 42 European brand-markets, at 10 brands, had some.
- United Statesbefore Dec 29, 2025
- CanadaApr 1, 2026
- United KingdomSep 19, 2026
- ItalySep 19, 2026
- FranceSep 20, 2026
- GermanySep 21, 2026
- SpainOct 1, 2026
- PolandOct 18, 2026
- NetherlandsOct 19, 2026
02It takes a small share of the budget, until a burst hits
Across the 48 brand-markets with any Off Amazon traffic, a median 0.25% of Sponsored Products spend went there (the middle half from 0.05% to 0.99%, the largest 8.8%). In Europe, from each marketplace's first Off Amazon day, the median was 0.22%. The US over the same weeks: 0.80%.
Eurozone: Germany, France, Italy, Spain, the Netherlands and Belgium. The eurozone peak (6.1%) is one German account: 99% of that week's eurozone Off Amazon spend. Section 08 is about these bursts.
03In Europe, Off Amazon clicks were cheap and rarely turned into an order
We compared Off Amazon with all on-Amazon placements of the same brand in the same marketplace. In Europe we counted from each marketplace's first Off Amazon day to November 16. In the US we used August 1 to November 16. We left out October 5 to 9, around Prime Big Deal Days, and kept every brand-market with at least 100 Off Amazon clicks: 10 in Europe from 4 brands, and 18 in the US.
In Europe an Off Amazon click cost 0.27 times an on-Amazon click, 73% less. It turned into an order 0.13 times as often: 2.0% against 16.8%. Off Amazon returned a median 0.63 times the on-Amazon return on ad spend, lower in all 10 European brand-markets. Brand C, whose click-through numbers look broken (section 05), is 3 of the 10. Without it the median is 0.74 times, still lower in 7 of 7.
October 5 to 9 left out. European brands are anonymous letters; the same letter is the same brand in every marketplace.
Show the numbers
| Europe | ROAS vs on Amazon | Conversion off | Conversion on | CPC vs on Amazon | Off Amazon clicks |
|---|---|---|---|---|---|
| Brand A · DE | 0.75× | 5.6% | 18.8% | 0.31× | 444 |
| Brand A · IT | 0.85× | 0.7% | 7.8% | 0.19× | 1,267 |
| Brand A · UK | 0.79× | 1.4% | 22.2% | 0.12× | 845 |
| Brand B · DE | 0.74× | 2.6% | 9.5% | 0.31× | 727 |
| Brand B · FR | 0.05× | 0.2% | 15.1% | 0.23× | 405 |
| Brand B · UK | 0.30× | 0.8% | 13.9% | 0.21× | 791 |
| Brand C · DE | 0.53× | 11.3% | 22.8% | 0.91× | 204 |
| Brand C · IT | 0.52× | 2.8% | 17.0% | 0.32× | 499 |
| Brand C · UK | 0.07× | 0.3% | 17.9% | 0.19× | 379 |
| Brand D · DE | 0.72× | 5.0% | 16.6% | 0.56× | 101 |
04Europe looks like the US, which has had it all year
The US result is close: Off Amazon returned a median 0.60 times the on-Amazon return, lower in 13 of 18 brand-markets. 5 US brand-markets did better off Amazon than on it, so it can pay for some brands. European clicks were relatively cheaper (0.27 times on-Amazon, against 0.46 in the US) and converted relatively worse (0.13 times, against 0.28). The return lands in about the same place.
The arrival of Off Amazon didn't change the on-Amazon return in any way we can detect. We compared the 28 days before each European marketplace's first Off Amazon day with the 28 days after, against the US over the same dates. On-Amazon return on ad spend moved -4% at the median and did better than the US in 13 of 33 brand-markets. In 17 of them Off Amazon spent nothing in those 28 days, so this test only rules out large changes.
05Don't trust the Off Amazon click-through rate in Europe yet
The European reports show far more clicks per impression than is plausible: 56% in the first week after launch. Most of that is one brand. Brand C had 86% of Europe's Off Amazon clicks in the first week and 93% in the third, and in the UK its clicks equalled its impressions on 15 of 16 days (99.7% overall). It stayed between 16% and 40% in the weeks after. Without brand C, the first four weeks ran at 16.3%, 6.9%, 8.5%, 6.5%.
Brand C's bursts look the same: 13 of the 14 European burst days in section 08 are brand C's. Across those 14 days, Off Amazon clicks were 78.5% of impressions, against 5.4% on other European days. Those days were weeks old when we pulled the data, so a late report doesn't explain it. From the fifth week on Europe as a whole sat at 4.5%, against 0.8% in the US from October 17 to November 15, and its clicks convert worse too. The impressions look incomplete, and we don't know why.
Week 9 is partial.
06The “minimize spend” setting switched it off
We have 40,722 Sponsored Products campaigns in our data. 32,297 of them (79%) have no value in the Off Amazon setting, 8,148 are on “minimize spend” and 277 on “maximize reach”. None of our European campaigns is on “maximize reach”.
From November 1 to 16, 0 of 791 campaigns on “minimize spend” with any spend had a single Off Amazon impression. Of the campaigns with no setting, 1,274 of 4,736 did (27%), and 50 of 209 on “maximize reach” (24%). Counting only brand-markets that have Off Amazon at all, it is 0 of 639 against 31%. Inside the same brand-market, campaigns on “minimize spend” spent a smaller share off Amazon than campaigns without a setting in 10 of 10 brand-markets.
The same campaigns tell the same story over time. Of the campaigns that had Off Amazon traffic between July 1 and August 15 and still spent in November, 0 of 204 now on “minimize spend” still get it, against 143 of 190 without a setting (75%). Our app keeps no history of this setting, so we can't see the day each campaign was switched. The chart shows when their Off Amazon traffic stopped.
The line for campaigns without a setting rises in late October as European campaigns start getting Off Amazon. Campaigns on “maximize reach” are left out.
07Exact-match campaigns got more than their share
We sorted every campaign by its targeting, from its active keywords and targets, and looked at October 1 to November 16 in the 39 brand-markets with Off Amazon spend. Campaigns with only exact-match keywords took 24% of Sponsored Products spend on average, but 42% of the Off Amazon spend. Product-target campaigns took 19% of the spend and 7% of Off Amazon.
Inside the same brand-market, among campaigns not on “minimize spend”, exact-match campaigns spent a larger share off Amazon than product-target campaigns in 25 of 27 brand-markets (lower in 1; a median 0.65% of spend against 0.081%). Against broad match it was 14 of 19 (lower in 4), against phrase match 15 of 22 (lower in 6) and against auto targeting 22 of 31 (lower in 8); the rest are ties. Auto campaigns touch Off Amazon most often, 53% of them against 24% of exact-match campaigns, but spend little there. We don't know why exact match gets so much. Amazon doesn't say how it picks the campaigns it shows off Amazon.
Show the numbers
| Campaign type | Share of SP spend | Share of Off Amazon spend | Not on minimize: had Off Amazon |
|---|---|---|---|
| Exact match only | 24% | 42% | 24% of 1,808 |
| Phrase match | 10% | 13% | 26% of 610 |
| Broad match | 12% | 7% | 19% of 461 |
| Product (ASIN) targets | 19% | 7% | 32% of 773 |
| Category targets | 2% | 1% | 38% of 78 |
| Auto targeting | 31% | 26% | 53% of 1,403 |
| Mixed, or no active targets | 2% | 5% | 24% of 139 |
08Bursts: a day of Off Amazon traffic that can use up the budget
Sometimes a campaign that normally gets almost no Off Amazon clicks gets hundreds in one day. We counted a burst when a campaign had at least 100 Off Amazon clicks and at least 10 times its median of the 28 days before. We found 63 burst days in 32 campaigns at 6 brands. On a burst day Off Amazon took a median 55% of the campaign's spend, about 50 times its usual Off Amazon clicks. All lasted one day (39) or two (12), and where a campaign had more than one, a median 18 days apart.
41 of the 63 were in one US account, which had set “maximize reach” on 241 campaigns. All 14 burst campaigns on “maximize reach” are in that account, so we can't tell the setting apart from the account. There the burst clicks turned into an order 0.44% of the time (83 orders on 19,020 clicks), against 2.7% for the same campaigns' Off Amazon clicks on other days and 13% on Amazon on the burst days. Burst days held 37% of that account's Off Amazon spend.
Bursts eat budget. 17 of 37 burst days with a known budget reached at least 95% of it (46%), against 6% of the same campaigns' other days. On 43 US burst days without a budget change, most of them in that one account, the campaigns' on-Amazon spend was 0.85 times its usual level (0.90 after allowing for the rest of the account that day). So part of a burst replaced on-Amazon spend: about 25% of it measured against the rest of the account, about 38% without that adjustment, with wide uncertainty either way. The rest came on top.
09What we can't tell from this data
- Credit for the sale. Amazon credits an order to the last ad click. A shopper who sees an Off Amazon ad, comes back later and buys through a search gives Off Amazon no credit. Our return on ad spend can undercount what Off Amazon adds.
- Where the ads ran. The report doesn't say which websites or apps showed them.
- When the setting changed. We see each campaign's current setting. Our app keeps no history of it.
- How many European accounts. 10 European brand-markets from 4 brands had enough clicks to compare, over two months that include Prime Big Deal Days (left out).
- Click-through rate. European impressions look incomplete, at least in the first weeks.
- Cause. The numbers describe what happened in the accounts we manage. They don't prove what Off Amazon does for every brand.
10What we'd do in Europe
- 01
Put your efficiency campaigns on “minimize spend”. It took Off Amazon to zero in our accounts. Most campaigns have no value in this setting, and those do get Off Amazon traffic. Start with exact match, which gets the biggest share.
- 02
If you want the reach, test it on one campaign. Judge it on the brand's total sales. Amazon credits the last click, so the campaign's return on ad spend misses an Off Amazon ad that leads to a later purchase.
- 03
Check the placement report for Off Amazon every month. Per campaign. There is no bid adjustment for it, so the setting is your only lever.
- 04
Watch for bursts. A campaign that normally gets almost no Off Amazon clicks and suddenly gets 100 or more in a day can run into its budget. Put it on “minimize spend”.
- 05
Ignore the Off Amazon click-through rate in Europe for now. At one brand it was too high to be real, and Europe as a whole is still far above the US.
11How we measured it
- Data. Amazon's daily Sponsored Products placement report, as our app stores it: one row per campaign, placement and day, for 34 brands in 69 brand-markets, December 29, 2025 to November 16, 2026. Orders and sales use Amazon's 7-day attribution. The campaign settings, keywords and targets are as of November 18, 2026.
- Brand-markets. One brand in one marketplace, with at least 30 days of data. Shares are worked out inside a brand-market, in one currency, and then summarised as a median or an average across brand-markets.
- Performance. Off Amazon against all on-Amazon placements of the same brand-market, from each European marketplace's first Off Amazon impression (US: August 1) to November 16, October 5 to 9 left out, at least 100 Off Amazon clicks.
- The setting. Each campaign's current Off Amazon setting. A campaign counts as having Off Amazon traffic with at least one Off Amazon impression in the period.
- Campaign types. From each campaign's active keywords and targets: exact match only, phrase match, broad match, product targets, category targets, auto targeting, or mixed (which includes campaigns with no active keywords or targets). Shares are averages of each brand-market's shares.
- Bursts. At least 100 Off Amazon clicks and 10 times the campaign's median of the previous 28 days. Budgets come from our app's change history, which starts on August 20, 2026; days with a budget change, or with spend above 135% of the budget (where an Amazon budget rule probably raised it), are left out of the budget figures. On-Amazon spend is compared with the campaign's median of the 14 days before, and with the same account's other campaigns that day.
- Re-check. Separate checks recomputed the numbers from the raw data without seeing our code, first for the findings and then for this page. They found our budget dates were in the wrong time zone, which had hidden part of the swap in section 08, rounding slips in the tables, that the odd click-through rates come mostly from one brand, and several claims worded more strongly than the data. Those are fixed.
- Limits. 34 brands we manage, 4 of them with enough European Off Amazon clicks to compare. The numbers describe what happened in these accounts.