Marketplaces and comparison shopping: a channel that runs on different rules
On a comparison site you are not buying attention, you are buying a position in a price ranking. That changes the whole logic of optimisation — and it is why a single number for the channel hides the categories that lose money.
- On a comparison site price is the main signal; bidding higher without a competitive price mostly buys clicks that do not convert.
- Results must be evaluated by category — a single figure hides the parts of the range that lose money.
- A marketplace is both a sales channel and a competitor to your own shop. That is a commercial decision, not a marketing one.
Comparison sites and marketplaces behave differently from an advertising auction, and the difference is fundamental. In search you buy the attention of someone who is still choosing. On a comparison site you enter a list sorted by price, where the decision is already being made between specific offers.
Which comparison engines matter depends heavily on the market — Idealo and Geizhals in German-speaking countries, Kelkoo and LeGuide in France, Ceneo in Poland, PriceRunner and Prisjakt in the Nordics, Google Shopping everywhere. The mechanics and the mistakes are the same.
Why it is managed differently
- Price is the main signal. If you are not near the top of the ranking, bidding more for position often just makes clicks dearer that would not convert anyway.
- Conversion rate tends to be higher than in general search, because the visitor is closer to a decision.
- Margin is under pressure. Precisely because your competitors’ numbers sit next to yours.
- Data quality decides. Matching your offer to the right product, availability, delivery, shop rating.
The most common mistake: one number for the whole channel
The channel reports a 12 % cost of sale and it looks fine. Underneath, three categories run at 5 % and five categories at 30 % — and the latter lose money at your margin. Because bids are set per category, this is also a mistake that is easy to fix.
The method:
- Export spend and orders by category over a long enough period.
- Attach each category’s gross margin, not the shop average.
- Calculate break-even cost of sale for each: it equals that category’s margin.
- Categories above break-even either get their bids cut or get removed from the channel.
- For categories well below break-even, consider whether they have room to grow.
Costs include not only bids but also sales commission where it applies, and the cost of processing an order. For categories with a low margin and expensive delivery, an order can be loss-making even at a very good cost of sale.
Data quality decides
Exactly as with the product feed for Google, data quality does more than bids:
- Correct matching to the product page. An unmatched offer loses a substantial part of its visibility.
- Availability and delivery time. Often decides more than the last few units of price.
- Complete parameters. Filters on a comparison site run on them; a missing parameter means nobody finds you in that filter.
- Shop rating. It affects conversion rate across the whole channel, not just one offer.
- Delivery cost. Shipping settings feed into the total price displayed, and therefore into your position.
Marketplaces: a sales channel and a competitor
Selling through a marketplace is a different calculation from a comparison site. You gain reach and traffic you could not build yourself, but you pay for it in three ways:
- Commission, usually higher than the cost of your own advertising at the same revenue.
- The customer relationship. The customer belongs to the marketplace, not to you. No email, no audience to build, no repeat purchase of your own.
- Price pressure. The same goods from five other sellers sit next to yours.
The approach we see working: a marketplace as a channel for goods with sufficient margin and low service cost, while your own shop and advertising concentrate on the range where building a relationship pays. It is a commercial decision rather than a marketing one, and part of it is asking whether you are cannibalising your own sales.
How it fits with the rest of PPC
- One conversion definition. Otherwise channels cannot be compared. See why the numbers never match.
- Evaluate the whole, not channels against each other. A customer may look at a comparison site and buy through search; both will claim them.
- Do not compete on price with yourself. A different price on a comparison site than on your own site leads to disapproval at Google and confuses the customer.
- Share what you know about margin. The margin bands you use for splitting campaigns apply here too.
What to take away
For some shops comparison sites and marketplaces are irreplaceable and for others they are a quiet loss — and you cannot tell which from the headline figure. The one piece of work that genuinely pays here is breaking spend and orders down by category and comparing each against its own margin. Usually a handful of categories turn up whose removal raises profit with no effect on revenue at all.
Common questions
Are comparison shopping sites still worth it?
For some shops yes and for others they are a quiet loss — but you cannot tell from the headline figure for the channel. The result has to be broken down by category with each category’s own gross margin attached, because break-even cost of sale equals exactly that margin. Typically you find several categories well below break-even that are carrying the average, and several loss-making ones whose removal raises profit with no effect on revenue.
How should I set bids on a comparison site?
By category and derived from its margin, not with one figure for the whole shop. Include not only bids but also sales commission where it applies and the cost of processing an order; for categories with a low margin and expensive delivery an order can lose money even at an apparently good cost of sale. Remember too that if you are not near the top of the price ranking, a higher bid often just buys dearer clicks that would not convert anyway.
Should an online shop sell on a marketplace?
It is a commercial decision, not a marketing one. You gain reach and traffic you could not build yourself, but you pay commission that is usually higher than the cost of your own advertising at the same revenue, and above all you lose the customer relationship — the customer belongs to the marketplace, so you cannot build an audience or repeat purchase from them. The compromise that most often works is putting goods with sufficient margin and low service cost on the marketplace, and keeping the range where a relationship pays on your own shop.