
Trust per topic, not per site
A website can seem trustworthy on travel but not on health. Google judges this per topic, which is why often only part of the traffic drops.
We examine why your website is losing reach in the Google Discover feed, find the affected topics and content and turn it into a concrete plan within four weeks.

I look at every topic separately.
In plain wordsGoogle Discover is the news feed in the Google app and on the Chrome start page on mobile. It shows articles without anyone searching for them. Whether your content appears there depends less on keywords than on how much Google trusts you on a topic.

A website can seem trustworthy on travel but not on health. Google judges this per topic, which is why often only part of the traffic drops.

For health, money and safety, Google looks closer: who wrote the text, with what experience, which sources?

Rewriting texts blindly wastes weeks. We first compare which pages still perform and which do not, and why.
That is why classic SEO measures often help little after a Discover drop.
| Aspect | Google Search | Google Discover |
|---|---|---|
| Trigger | someone actively searches | Google suggests, without a search |
| What matters | search terms, technical setup, links | trust per topic, freshness, images, authors |
| Typical drop | gradual, single terms | sudden, whole topic areas |
| Measurement | Search Console, queries | Search Console, Discover report |
From data to plan, with clear interim results.

Discover data from the last 12 months from Search Console, crawl of all articles, preparation.
Result: complete data set

All articles are grouped into topic clusters, typically 30 to 50 per audit, sensitive topics flagged.
Result: topic map

Winners and losers compared, authors and sources per article, wording in sensitive topics.
Result: causes per topic

Prioritized actions per article group, quick wins first, final call.
Result: recovery plan
The analysis leads to a decision per page, each with a reason.
Articles that still bring Discover traffic. They show us what works.
Author, sources or freshness are missing. A targeted rework is often enough.
Thin or duplicate articles that can lower trust in a whole topic.

Every topic with its own value.
In the first call we look at your Discover data and tell you whether an analysis is worth it. Free, no obligation, 30 minutes.
Your contact: Aredi Dragusha, [email protected]