When is the last time a Google search actually sent you to a website instead of answering you right there? That was the opening question on Episode 133 of SEOTalk Spaces, and it sets up the entire conversation. Zero-click searches now account for 60 percent of all Google queries, and roughly 80 percent of AI Overview interactions end without a click. The deal that defined SEO for two decades, publish good content and Google sends you traffic, is breaking down in real time.
Against that backdrop, Google has quietly rolled out a set of features: preferred sources, search profiles, and subscription linking. None of them are built around traffic. All of them are built around loyalty. Host Malhar Barai opened the session with the question worth sitting with: is Google handing publishers a lifeline, or locking them deeper into its own ecosystem? Parth Suba and few members from the SEOTalk community worked through it from there.
Small Publishers Are Losing the Most
The numbers framing the conversation were stark. Small publishers have lost roughly 60 percent of their search traffic over two years. Large players have lost closer to 20 percent. That gap is not closing. If anything, the new loyalty-based features risk widening it further.
Parth’s read was direct: these features reward an audience a brand has already built, rather than helping a brand build one. Discover traffic might increase for publishers who already have engaged readers. Overall organic traffic decline, though, is expected to continue regardless.
What “Preferred Sources” Actually Does
Parth walked through the mechanics. Preferred sources is a feature publishers can add as a button or embedded CTA on a website, in an email newsletter, anywhere a brand has built a community. When a logged-in user opts a site in as a preferred source, that site becomes more likely to surface for that user’s future queries in the relevant category, independent of traditional ranking signals.
That distinction matters. This is not a ranking factor in the conventional sense. It functions more as a user preference signal layered on top of the existing system. Parth was clear that this shows up reliably on the Discover feed; whether and how it influences traditional search rankings is still being tested. He described running live experiments with two clients and called it too early to draw firm conclusions.
What it does appear to do well is build what Parth called an “algorithm-proof” audience. If a user has explicitly chosen a brand as a preferred source, that brand keeps surfacing in their feed even through core algorithm updates. That is a meaningfully different kind of durability than ranking position has ever offered.
Three Hard Questions About Fairness
Malhar pushed on three concerns at once: is there an eligibility bar to participate, does this structurally favor larger publishers who already have bigger audiences, and does it risk turning search into more of an echo chamber, where users only ever see brands that match their existing preferences?
Parth’s answer on the size question was more optimistic than the framing implied. He argued the feature is not inherently tilted toward bigger brands. Two brands with similar organic traffic, one with a large email list and one without, are still on equal footing if the smaller one is more deliberate about cross-channel promotion of the preferred-source CTA. The advantage goes to whoever activates the feature thoughtfully across email, social, and on-site placements, not necessarily whoever already has the bigger following.
Isha offered a sharper, more skeptical framing of the whole premise. Her comparison: this is Google breaking the system and then selling the fix as a cure. The underlying point is that Google created the conditions for zero-click decline through featured snippets and AI Overviews, and is now positioning preferred sources as the remedy to a problem it caused. Whether that remedy meaningfully redistributes traffic back to publishers, or simply gives Google a justification to keep doing what it was already doing, is an open question neither panelist claimed to have resolved.
What Is Google’s Actual End Game?
Parth’s view: Google does not need more text content.
The web is already trained, the models already exist. What Google needs now is for publishers and creators to stay inside its ecosystem rather than leaving for other platforms entirely. Preferred sources functions as what he called a “sidekick” thrown to publishers, an acknowledgment that they are bleeding traffic, paired with a tool whose actual effectiveness remains the publisher’s problem to figure out. Google retains full control over the algorithm regardless of how many preferred-source subscribers a brand accumulates.
Isha widened the frame further. From a pure product standpoint, she argued, Google has to keep building these features because the search landscape itself has fragmented. User behavior has scattered across YouTube, Reddit, social platforms, and community forums in ways that simply did not exist when Google displaced Yahoo. Google remains the largest search engine, but its monopoly on discovery is gone. Features like preferred sources are partly a defensive move to keep users from researching solutions entirely outside Google’s properties.
She also flagged a deeper measurement problem underneath all of this: outside of Google Analytics, there is no reliable, independent tool for tracking what is actually happening to traffic and attention in this fragmented landscape. Citing a widely circulated SparkToro and Semrush data point, she noted that a large share of people who search are not clicking anything at all afterward, yet brand recall is clearly happening. There is currently no good way to measure or prove that recall to a skeptical CMO.
From Keyword Intent to Customer Understanding
One of the more substantive threads in the episode was a challenge to how SEO professionals have traditionally approached intent. Malhar asked directly: for people who built careers optimizing around keywords and phrase variations, how do you shift toward understanding genuine intent? Is that even a different skill, or does it eventually collapse back into traffic as the only KPI that matters?
Isha pushed back on the framing of intent as a keyword-level exercise at all. Her argument: intent understanding is not something you derive by staring at a key phrase and guessing what someone might mean by it. It comes from genuinely understanding the business, the customer, and the different reasons someone might arrive at a brand: a learnability angle, a comparison angle, an awareness-stage angle.
Her practical example came from e-commerce. If a brand sells phones, the work is not just ranking for “buy phone X.” It is making sure the brand shows up when someone searches “best phones in this category” at the awareness stage, and again when someone is comparing two specific competitor products. That requires SEO professionals to sit with brand experts and understand customer touchpoints holistically, rather than reverse-engineering intent from a keyword list.
On whether traffic is fully out of the picture as a KPI, Isha’s answer was nuanced: yes for Google Analytics specifically, but not entirely, because a meaningful amount of activity and impact is currently untracked rather than nonexistent. Impressions, she argued, are becoming a more honest signal of brand exposure than clicks, even though the tooling to measure that well still does not fully exist. She drew a comparison to traditional out-of-home advertising: a billboard’s value has always been estimated, not precisely measured, based on foot traffic and visibility in a given area. Online discovery in the AI era may require a similar shift toward estimation rather than exact attribution.
Is This New, or Just Old Trust-Building With New Packaging?
A useful reframing came late in the conversation. Isha argued that none of this, brand trust, loyalty, reputation, is fundamentally new. What has changed is that previously isolated disciplines, SEO, performance marketing, brand, are now structurally overlapping in ways that make clean KPI ownership difficult. A marketing team used to be able to say “these are SEO’s metrics, these are brand’s metrics, these are performance’s metrics.” Now those lines blur, and that ambiguity itself is part of why this period feels harder to navigate than past shifts.
Malhar offered his own framing in response: is Google effectively saying it no longer fully trusts its own algorithm, and is asking users to supply an explicit human signal, “this is a preferred source,” as a layer on top of what the algorithm alone can determine? Isha’s response, half-serious, was that only Google can really answer that. But the underlying observation holds: SEO began as a heavily human-curated signal system, backlinks and citations as votes of trust, before Google’s algorithm absorbed and automated that judgment. Preferred sources looks like a partial return to explicit human signal, layered on top of an algorithm that may be reaching the limits of what it can infer on its own.
A guest panelist offered the more cynical version of this same idea: Google may simply be capturing a missing signal it never had access to before, with the secondary benefit of feeding that same preference data into how ads get served. From that view, preferred sources is less a publisher lifeline and more a data acquisition mechanism dressed up as a loyalty feature.
Will Clicks Stop Mattering Entirely?
Malhar asked the question directly: are we heading toward a future where the entire SEO effort shifts toward AI Overview discoverability and preferred-source accumulation, with the click as a metric essentially retired?
Parth’s answer reinforced a position he has held across recent episodes: SEO is no longer a traffic channel, it is a brand influence channel. The real marker of success in this environment is whether target audience members remember the brand well enough to search for it by name later. Branded search demand, direct visits, and returning visitors are harder to fake than traditional rankings, which makes them a more durable measure of whether brand-building efforts are actually working.
Isha added an important caveat from her own client base: for enterprise-level brands, this is genuinely a challenging period. But for MSMEs and growing startups, the picture looks different. Several of her clients are seeing branded clicks increase, not decline, because they are in a growth phase where overall brand awareness work, influencer collaborations, offline promotion, even a Shark Tank appearance, is lifting branded search volume across the board. Her broader point: brand, performance marketing, and SEO need to function as one coordinated effort now, because their effects are no longer separable.
The Gamification Question Nobody Has Fully Answered
Malhar raised a concern that ran through earlier episodes too: how easily can preferred sources be gamed? The comparison he drew was to early-2010s affiliate marketing tactics that exploited loose signals before platforms tightened enforcement.
Isha’s response acknowledged the risk without dismissing the feature. Yes, loyalty signals can be manipulated, the same way early SEO was manipulated through keyword stuffing before Google’s algorithm matured enough to penalize it. Her expectation is that Google will iterate the same way: early adopters who invest now will get a real head start, and enforcement against manipulation will tighten over time, the same pattern SEO has followed since its earliest days.
Parth’s closing framing tied the value question to a deeper shift: is search moving into a phase where being chosen matters more than being technically correct or most relevant? His answer leaned toward yes, with the caveat that being chosen and earning genuine trust are, ideally, the same outcome. Brands that get this right early will have a real competitive edge over those still optimizing purely for blue-link position, even as the lack of clean measurement tools makes it hard to prove that edge to a skeptical stakeholder.
The Honest Summary
Episode 133 did not resolve into a tidy verdict, and the panel was upfront about that. What emerged instead was a working hypothesis: Google is not in the business of cheap clicks anymore. The clicks that remain are increasingly concentrated around digital PR, brand recognition, and direct intent, rather than the broad organic discovery that defined SEO for two decades.
Whether preferred sources, search profiles, and subscription linking genuinely return value to publishers, or simply give Google a more defensible story while it consolidates control over discovery, is not yet settled. The panel’s practical advice was to treat these features as worth testing now, while staying clear-eyed that the algorithm, the eligibility, and the long-term value of the badge remain entirely in Google’s hands.
SEOTalk Spaces is a weekly community conversation hosted by Malhar Barai and Parth Suba. Episode 133 featured contributions from Isha and community members from the SEOTalk audience.
Catch the replay: