kalu Noise Index
October 2026 · Updated · Leer en español
Of 3,006 stories kalu published between June 16 and October 1, 2026, not one went out without the URL of its original source. To get there, the engine assessed 1,040 news domains that were not on its list of recognized outlets and rejected 229 of them, 22%.
What we measured
Period: to . Source: kalu's production database.
22%
of unknown news domains were rejected as a source
229 of 1,040 domains assessed between August 7 and October 1, 2026.
- Sample
- 1,040 domains
- Method
- Every domain that is not on the allowlist of established press and official sources is classified before any of its stories can be published. The test is whether there is journalism or an institution behind it. Social networks, tabloids, satire and disinformation sites are blocked earlier, by a denylist, and are not counted here.
0
stories published without the URL of their original source
0 of 3,006 stories published between June 16 and October 1, 2026.
- Sample
- 3,006 stories
- Method
- Count of stories published to feeds whose source field is not an http(s) URL. The database has required that field since the first version of the schema, so the zero is not an accident.
791
distinct domains behind the published stories
791 domains across 3,006 stories: on average, fewer than four stories per outlet.
- Sample
- 3,006 stories
- Method
- Count of unique domains in each published story's source URL, without “www.”.
| Front | Stories | Share of total |
|---|---|---|
| Your customers' industry | 1,482 | 49.3% |
| Geographic risk | 413 | 13.7% |
| Your industry | 401 | 13.3% |
| International | 315 | 10.5% |
| Regulatory | 205 | 6.8% |
| Prospects and opportunities | 190 | 6.3% |
| Total | 3,006 | 100 % |
What this index does not say
- This is a pre-launch period with few active radars. The breakdown by front describes those radars, not the market.
- The rejected-domains figure starts on August 7, 2026, when each domain's verdict began to be stored.
- Nothing on this page identifies an organization: these are aggregate counts across all of them.
What others measured
Third-party figures that frame the problem. Each one links to the original publication.
58.5%
of Google searches in the United States end without a single click
In the European Union, 59.7%. Searchers get the answer on the results page itself, without visiting the source.
SparkToro, “2024 Zero-Click Search Study”, Rand Fishkin, . Sample: Datos clickstream panel (tens of millions of users, desktop and mobile).
Read the original source34.5%
fewer clicks go to the top result when Google shows an AI Overview
Compared with similar informational keywords without an AI Overview. It is a correlation in average CTR, not a controlled experiment.
Ahrefs, “AI Overviews Reduce Clicks by 34.5%”, Ryan Law, . Sample: 300,000 keywords.
Read the original source40%
of their time is what the average seller spends selling
The other 60% goes to work that is not selling. Among Gen Z sellers, time spent selling drops to 35%.
Salesforce, “State of Sales Report”, State of Sales, 2026, . Sample: 4,050 sales professionals in 22 countries, Mexico included; fieldwork August to September 2025.
Read the original source40%
of people say they sometimes or often avoid the news
In 2017 it was 29%. Among the reasons the report gives are information overload and a sense of powerlessness.
Reuters Institute, Universidad de Oxford, “Digital News Report 2025”, Digital News Report 2025, . Sample: Online YouGov survey in 48 markets, mid-January to late February 2025.
Read the original source
How to cite this data
Use it freely with attribution and a link to this page, or to the specific figure with its anchor.
kalu, “kalu Noise Index”, October 2026. https://kalu.lat/statistics
For the full method or a figure that is not here, write to us from the contact link in the footer. Who makes kalu and why: about us.