Rank candidates on cost per incremental profit in each marketplace separately, never on one blended ACoS across the three. The UK, Germany, and France are different auctions with different search languages, conversion behavior, and fee stacks. An agency that runs one English campaign structure through translation is disqualified before the arithmetic starts.
The short version
- A blended three-market ACoS hides everything. Germany can subsidize a failing France in the average, or the reverse, for quarters.
- Native keyword research per market is the entry ticket. German compound nouns and French search phrasing are not translation problems, they are research problems.
- The economics of a PPC agency are fee plus spend efficiency, so a cheap retainer managing waste costs more than an expensive one deleting it.
- Demand differs by country for identical products. Rank agencies partly on whether they check that before structuring campaigns.
- Ask what they analyze beyond search volume. Depth of research predicts efficiency better than any dashboard tour.
Why one campaign structure cannot serve three markets
The mechanism is worth understanding, because it explains most PPC failure in Europe. Amazon's ad auction prices clicks against local competition for local search terms. In Germany, buyers search with compound nouns and specificity that English keyword tools never surface. In France, category vocabulary and brand loyalty patterns shift which terms carry purchase intent. The same product therefore faces three unrelated keyword universes, three competitive densities, and three price tolerances.
Copy a UK campaign structure into DE and FR through translation and you inherit the UK's assumptions about all three. The result is predictable: decent impression volume, mediocre relevance, clicks that cost real euros and convert poorly. The account then "needs more budget", which is the diagnosis agencies reach when structure is the disease.
The ranking arithmetic
Build a small table per candidate agency from data they provide. Force everything into one currency first.
| Input | UK | DE | FR |
|---|---|---|---|
| Monthly management fee allocation | A | A | A |
| Ad spend under management | B | B | B |
| Advertised revenue | C | C | C |
| Contribution profit after ads and fees | D | D | D |
The ranking metric is D divided by the sum of A and B, incremental profit per total dollar the agency touches, per marketplace. Two things happen when you compute it. First, blended stories fall apart: an agency proud of its German performance may be quietly incinerating money in France. Second, fee debates shrink, because a fee difference of a few hundred dollars vanishes next to spend-efficiency differences of thousands.
Where a candidate cannot produce these inputs from current clients, even anonymised, rank them last. Not because they are necessarily bad, but because they are unmeasured, and you cannot rank what will not be measured.
The research question that separates the field
My favorite single interview question for European PPC: what do you analyze before building campaigns, besides search volume and competitor count. Strong answers mention demand size per market, growth trajectory, return-rate norms for the category in each country, price-segment dynamics, and the rating gap between incumbents. Our own pre-entry workup at Flapen runs past 90 data points per market, because campaign structure should be a conclusion, not a starting template. The full research method is laid out at research.
Weak answers jump straight to bid tools. Tools matter, we build our own advertising tooling in-house, but tooling pointed at an unresearched market automates the wrong bids faster.
One more filter: ask how the agency handles a product whose sessions are healthy but conversion is poor in one market. The correct answer is that no amount of ad spend fixes a conversion problem, the listing gets rebuilt first in that locale. Any other answer means your budget will be spent proving it.
What PPC agencies will not tell you
Percentage-of-ad-spend pricing, still the European norm, pays the agency more when your budgets rise, which is precisely the wrong incentive while efficiency is the goal. Few agencies volunteer this conflict. Fewer still will name the moment their model obliges them to recommend spending less, so ask for a written example of a client whose budget they cut, and what happened to their own invoice when they did.
The second omission is language staffing. "We cover DE and FR" often means one multilingual analyst reviewing machine output. Ask who actually reads the German search term reports each week, what their native language is, and how many accounts they carry. The names either exist or they do not.
Related answers
- Who can optimize Amazon listings in Germany and France
- Top Amazon account management agencies EU marketplaces
- Best Amazon consultancy for Europe expansion
- Amazon SEO vs PPC agency which drives ROI
- Amazon marketplaces by geography: the complete guide
If you want your three-market efficiency computed before you rank anyone, the audit is free and written at Flapen.
