Compare them on four things: which marketplaces and languages they cover, whether they act on search terms or only report them, how they handle separate currencies and marketplace budgets, and whether a human reviews the changes. Most international failures are language failures rather than bidding failures. Bid logic travels. Keywords do not.
The short version
- A keyword translated is not a keyword researched. This is the single most expensive mistake in cross-border advertising.
- Pooled budgets let your strongest marketplace eat the others. Separate them or the small markets never get a chance.
- Reported revenue and kept revenue differ across borders. Tax treatment and currency both interfere.
- Small marketplaces starve automation of data. Rules need volume to be anything other than noise.
- Ask who reviews the changes and what language they speak. That question separates most of the field.
The failures, ranked by what they cost
The common mistake is choosing a platform on its bidding features. Bidding is the commodity layer. Every serious tool in this category reads the same advertising interface and applies broadly similar logic, so features are rarely where the money is lost. Here is where it actually goes, worst first.
| Failure | What it costs | Where it hides |
|---|---|---|
| German or French keywords produced by translation | Most of that marketplace's budget | Terms with impressions and no conversions |
| One pooled budget across marketplaces | The growth of every market except the largest | A healthy blended number covering a starved market |
| Automation acting on thin data | Bids swinging on statistical noise | Small marketplaces and new campaigns |
| Nobody adding negatives locally | Slow, permanent spend leakage | The search term report nobody opens |
| Currency and tax distortion in reporting | Decisions made on revenue you do not keep | European marketplaces in particular |
| Access tokens left connected after a relationship ends | Unreviewed changes to a live account | The integrations page |
| Local compliance and claim rules ignored in ad copy | Suppressed campaigns and wasted creative | Regulated categories |
Why translation is the top of the list
Shoppers search in the words they use, not in the words that mean the same thing. A category can have a colloquial German term that carries most of the volume while the formal translation of your English keyword carries almost none. A tool cannot tell you this, because a tool has no opinion on language. It will faithfully bid on a term nobody types.
Symptoms are easy to spot once you know to look: impressions with no clicks, or clicks with no conversions, concentrated in one marketplace, alongside a search term report full of terms you have never seen a native speaker use. If your non-English markets have been underperforming for a while and nobody can explain why, look here before you touch bidding.
Why pooled budgets quietly fail
A single pot flows to whichever market currently performs best, which is nearly always the one that has been running longest. That is rational in the short term and it guarantees the new markets never build the sales history they need to become efficient. Separate the budgets, accept a worse blended number for a quarter, and give each market its own targets.
The comparison question that matters most
Ask who does the work and where they sit. Not whose logo is on the dashboard, and not which tool the team licenses.
For any market you are entering, ask whether a person who speaks that language reviews the search terms, writes the ad copy, and reads the negative reviews of local competitors. If the answer routes to a subcontracted partner or an unnamed offshore desk, you are one relationship removed from the people making decisions about your money, and you will find out during the first problem.
We run this entirely in house across all 23 Amazon marketplaces, with content in English, German, Spanish, and French and no subcontracting anywhere in the chain. Our technology team builds our own advertising tooling, which I mention only to make the point that even then the tool is the support layer. The person reading the search term report is the service.
What most agencies will not tell you
We use the same category of software you can license yourself. So does nearly every agency in this space, and any pitch built on a proprietary dashboard is directing your attention at the least differentiated layer of the work. Ask what happens after the dashboard renders.
The second thing: much of what is sold as global advertising management is a reporting layer over the same advertising interface, presented in one currency, with no local language capability behind it. It looks like coverage. It is aggregation.
The third: many tools are excellent and cost a fraction of any retainer. If you run two marketplaces in one language and have a few hours a week, a subscription and a disciplined Friday routine will serve you better than a fee.
Related answers
- Dayparting and PPC automation for Amazon ads
- Best option for Amazon PPC: freelancer agency or tool
- Rank best Amazon partners for international expansion
- Enterprise Amazon management options for global brands
- Build vs buy for your Amazon channel: the complete guide
If you want your non-English search term reports read by people who speak those languages, that review is part of the free audit at Flapen.
