%20(1).jpg)
Google Discover Is Testing “Dive Deeper” AI — What Publishers Need to Know
That may sound like a small Discover interface experiment, but it could matter to publishers who depend on Discover for traffic.
The new experience creates another step between a reader seeing a publisher's content in Discover and actually visiting that publisher's website.
Key Takeaway: Google is testing “Dive deeper” in Discover, initially with video content. The feature opens an AI-generated topic overview containing links to related stories, community reactions and original reporting. Google has not yet published enough data to determine how the experiment will affect publisher traffic.
What Is Google Discover “Dive Deeper”?
Google is experimenting with a new “Dive deeper” button inside Discover.
According to Google's announcement, tapping the button opens a short topic overview designed to help users explore related information. The overview can include links to related stories, community reactions and original reporting.
Google says the experiment will begin with videos and that it will test multiple designs in the coming weeks.
That last part is important.
This is currently an experiment, not a permanent Discover feature that every user should expect to see.
How the New Discover Experience Works
The basic journey is different from the traditional Discover experience.
| Step | What Happens |
|---|---|
| 1. Discover feed | A user sees a recommended piece of content in Google Discover. |
| 2. Dive deeper | The user can tap the new “Dive deeper” option. |
| 3. AI overview | Google presents a short AI-generated explanation of the topic. |
| 4. Related sources | The experience can show links to related stories, community reactions and original reporting. |
| 5. Publisher visit | The user can then choose one of the linked sources and visit the publisher. |
So the important change is not simply that Google is adding AI.
It is that AI becomes another layer inside the Discover journey.
Why Publishers Are Paying Attention
Discover has traditionally worked differently from normal Google Search.
A user scrolls through a personalized feed, sees a headline or video that interests them, and then decides whether to open it.
The website is often only one tap away.
With “Dive deeper,” Google is testing an additional experience where users can first consume an AI-generated overview of the topic.
That creates an important question for publishers:
Will users still click through to the original article after reading Google's AI-generated overview?
There is currently no public data from Google showing that the experiment is increasing or decreasing publisher traffic.
Industry observers have nevertheless raised concerns that an AI summary could answer enough of a user's question that some users may decide not to continue to a publisher's website.
That is a concern worth watching, but it should not be treated as an established traffic loss from this experiment.
Google Says the Goal Is Easier Topic Discovery
Google's explanation focuses on helping people explore a topic and discover related web content.
The company says the “Dive deeper” experience can surface related stories, community reactions and original reporting.
That means the feature is not designed as a simple replacement for publisher websites.
Instead, Google is positioning it as a discovery layer that organizes additional information around the topic.
Whether users actually continue from that layer to publisher websites is something the experiment will help Google evaluate.
The AI Summary Is Not the Same as a Normal Search Result
It is easy to confuse the new Discover experiment with Google's AI Overviews in Search.
They are related to Google's broader use of generative AI, but the user experience is different.
AI Overviews appear within Google Search results, while this experiment is being tested directly inside the personalized Discover feed.
That distinction matters for publishers because Discover traffic behaves differently from traditional search traffic.
| Google Search | Google Discover | |
|---|---|---|
| User intent | User is actively searching | User is browsing recommended content |
| Main discovery mechanism | Search query | Personalized feed |
| AI experience | AI Overviews and AI Mode | “Dive deeper” experiment |
| Publisher opportunity | Search visibility and clicks | Recommendations and feed engagement |
Could “Dive Deeper” Reduce Discover Traffic?
It is too early to answer that question.
There is currently no public Google study showing how the experiment affects clicks to publisher websites.
However, the design creates a possibility publishers should monitor.
If a user sees an AI-generated overview that answers the basic question, the user may or may not feel a need to open the linked articles.
On the other hand, users who want more detail can still follow the prominent source links shown in the experience.
That means the actual result could depend on the type of content.
| Content Type | Possible User Behavior |
|---|---|
| Breaking news | User may want the original reporting or latest updates. |
| How-to content | User may still need the publisher's detailed instructions. |
| Simple factual stories | An AI overview could potentially answer the basic question quickly. |
| Opinion or analysis | Users may want the original article to understand the full argument. |
| Original research | Links to the original source can remain important for deeper information. |
What This Means for Google Discover Publishers
For publishers, the safest response right now is not to completely change your Discover strategy because of an experiment.
Instead, use this as another reason to make your content useful after the initial headline click.
A Discover article should give readers something that a short summary cannot easily replace.
- Original information
- First-hand experience
- Practical instructions
- Useful screenshots or examples
- Clear explanations
- Unique analysis
- Updated information when the story changes
- Details that require visiting the original page
This does not mean writing longer articles simply to make them longer.
The goal is to provide more useful information than a short AI-generated overview can reasonably contain.
What Panstag Publishers Should Monitor
If your website receives Discover traffic, this is a good time to establish a baseline.
Before the experiment becomes widespread, record your normal Discover performance so you have something to compare later.
Pay particular attention to:
- Discover impressions
- Discover clicks
- Discover CTR
- Top-performing articles
- Traffic changes after publication
- Traffic from individual article types
- How quickly new articles begin receiving Discover impressions
Do not assume that every change is caused by “Dive deeper.” Discover traffic naturally fluctuates, and Google can change recommendation systems independently of this experiment.
Don't Change Your Headlines Just Because of This Test
One tempting reaction would be to make every Discover headline more aggressive.
That would be the wrong takeaway.
Strong headlines matter, but the headline must accurately represent the article.
Google Discover is designed around personalized recommendations, so misleading curiosity gaps can create a poor user experience when the reader opens the article and doesn't find what the headline promised.
For Panstag, the better strategy is a combination of:
- Clear news value
- Strong curiosity
- Accurate claims
- Fresh information
- Useful original details
What About AI-Generated Content?
The arrival of more AI-generated summaries makes original content even more important.
If Google can summarize a topic itself, publishers need to give readers a reason to continue to the original source.
That does not necessarily mean avoiding AI tools.
AI can still help with research, outlining, editing and productivity.
But the finished article should provide genuine value rather than simply repeating information that readers can already obtain from an AI summary.
For a technology publisher, that could mean showing exactly what changed, explaining how a feature works, testing something yourself, providing screenshots, comparing options or explaining what the change means for a specific type of user.
A Bigger Change May Be Coming to Discover
“Dive deeper” is interesting because it represents another step toward making Discover more than a simple list of links.
Google is experimenting with an experience where the platform can help users understand a topic before they decide which source to visit.
That could make Discover more useful for readers.
At the same time, publishers will naturally want to know whether that additional Google-generated layer sends more qualified visitors to original sources or keeps more of the discovery experience inside Google.
That question cannot be answered from the current test alone.
What Publishers Should Do Right Now
There is no reason to panic or make major website changes solely because Google announced this experiment.
Instead, publishers can prepare by strengthening the parts of their content that provide value beyond a short summary.
| Action | What to Do |
|---|---|
| Track Discover | Monitor impressions, clicks and CTR in Search Console. |
| Improve originality | Add original reporting, testing, examples and useful analysis. |
| Keep content fresh | Update articles when facts, products or features change. |
| Improve headlines | Use curiosity without making claims the article cannot support. |
| Study winners | Identify which topics and formats consistently receive Discover exposure. |
| Don't overreact | Wait for measurable traffic changes before making major strategy decisions. |
If you're also working on improving how your publication appears across Google's ecosystem, see our guide on adding a Google Preferred Source button to a Blogger site.
You can also compare Discover changes against broader traffic movements using our guide on why website traffic can suddenly drop.
Frequently Asked Questions
What is Google Discover “Dive deeper”?
It is an experimental Discover feature that opens an AI-generated topic overview when users tap “Dive deeper.” Google says the overview can include links to related stories, community reactions and original reporting.
Is Google Discover “Dive deeper” available to everyone?
No. Google has described it as an experiment. The company says it is starting with videos and will test multiple designs in the coming weeks.
Will “Dive deeper” reduce website traffic?
There is not enough public data to say. The feature creates another AI-generated step before a user reaches linked sources, so publishers are watching its potential traffic impact, but Google has not published traffic results from the experiment.
Does “Dive deeper” replace the original publisher?
Not according to Google's description. The feature is designed to show an AI-generated overview while also providing links to related stories, community reactions and original reporting.
Should I change my Discover SEO strategy now?
There is no strong reason to make major changes solely because of this test. Continue focusing on accurate headlines, original information, useful content, freshness and a good reader experience while monitoring your Discover data.
Does Google Discover still send traffic to websites?
Yes. Discover remains a traffic source for publishers, but individual performance can fluctuate significantly. Publishers should monitor their own Search Console data rather than assuming that a single Google experiment explains every traffic change.
Key Takeaway
Google is testing a new AI layer inside Discover that could change how users explore stories before visiting publishers. The “Dive deeper” feature currently starts with videos and provides an AI-generated topic overview with links to additional sources. There is not yet enough evidence to determine its effect on publisher traffic.
Bottom Line
Google Discover is becoming more AI-driven, and the new “Dive deeper” experiment is another sign of that direction.
For publishers, the important part is not to assume that this automatically means less traffic.
The experiment is still new. Google says it will test different designs, and there is no public traffic data showing what happens after users interact with the new experience.
What publishers can control is the quality of the destination they offer.
If a reader can get a basic summary from an AI layer, the original article needs to offer something more valuable: original reporting, practical information, detailed explanation, evidence, examples or analysis.
For now, watch the data rather than reacting to speculation.
Google is testing how people discover information. Publishers should be testing how effectively their content gives those people a reason to continue beyond the summary.
Source note: This article is based on Google's September 18, 2026 announcement as reported by current search-industry coverage. Google has not publicly disclosed the test's full geographic rollout, audience size or traffic impact.
