Your product ranks for the wrong search term: finding and fixing irrelevant results on otto.de

Why irrelevant rankings hurt

At first glance every additional ranking sounds good. In reality an irrelevant result costs you: shoppers click, see that the product does not fit, and go back. What otto.de does with that internally, OTTO does not say; on other marketplaces, listings with many clicks and no purchase tend to be rated worse. What is certain: the clicks bring no revenue, and ads on such terms burn budget.

There is a measurement problem too. If your keyword set contains irrelevant terms, you see position gains that are worth nothing and miss drops on the terms that count.

How to discover wrong rankings

By hand this is hard, because you only check the terms you think of yourself. Use your rank tracker's keyword discovery instead: it shows which search terms a product is actually found for. In MarktRank this is called Find keywords. Go through the list and mark every term your product does not satisfy:

• Wrong variant: “kettle 0.5 litre” for a 1.7-litre model.
• Wrong product type: “tea maker” or “milk frother” for a kettle without those functions.
• Wrong audience: “travel kettle” for a countertop appliance.
• Someone else's brand: terms containing a brand name that is not yours.

A second hint is your ad statistics: terms with many clicks and no purchase are usually irrelevant results.

Finding the cause in the listing

Irrelevant rankings come from words in the listing. Check in this order:

1. Title: does it name properties the product does not have? Typical are lists of use cases (“for travel, office, camping”) or comparisons (“like a tea maker”).
2. Bullet points and description: sentences such as “also suitable as a milk frother” or “replaces your old 1-litre kettle” bring exactly those words into play.
3. Attributes: a wrongly set attribute (wrong category, wrong volume, “portable: yes”) has more effect than any text, because filters and search are built on it.
4. Variants: when several variants share one product page, the page ranks for the properties of all variants. That is correct as long as the variant actually exists.

Correcting without losing good rankings

Remove only the words that trigger the wrong term and leave everything that belongs to your good rankings. Change one thing at a time and note the date. Afterwards watch two things over several checks: the irrelevant term should disappear or drop clearly, the relevant terms should stay stable. If an important term drops along with it, the removed word was needed after all – put it back more precisely (“1.7 litre” instead of “litre”).

How to measure such changes cleanly is described in the measurement protocol for listing changes.

Irrelevant or just unexpected?

Not every surprising result is wrong. If your glass kettle ranks for “kettle without plastic”, you may have discovered a term that really moves shoppers and that you have not served so far. The question is always: would a shopper with this search buy my product? If yes, add the term to tracking and to the listing. If no, remove the trigger.

Frequently asked questions