Subscribe
News

Looking Back at Local: 2025

Some of the biggest themes, developments and issues we talked about in 2025

Greg Sterling

Looking Back at Local: 2025

AI Invades the SERP

AI features and content continued to encroach on the traditional SERP in 2025. AI Overviews (AIOs), which launched in 2024, were a major subject of debate (and derision) in 2025. Many publishers, agencies, research vendors and tool providers argued that AIOs kill clicks. Google disputed this, claiming the methodologies behind these studies were flawed. (In our behavioral research [in the legal vertical] we didn't see AIOs suppress clicks, but LSAs did.) Beyond AIOs, Google aggressively started exposing people to AI Mode in an effort to get them to use it instead of ChatGPT. AI Mode was announced in March and rolled out in the US in May, and then more countries throughout 2025. Many SEOs believe AI Mode will fully replace the traditional SERP this year. That's unlikely. But it may replace AIOs, which were a kind of answer to ChatGPT and preliminary version of AI Mode. In local specifically, Ask Maps (or "Ask a Question") was introduced with content from GBPs, reviews, websites and third-party sources. Throughout 2025, Google experimented with AI features in the Pack/GBP, such as AI-enabled calling for price information, AI-generated attributes, automated menu/service updates and others. It will continue to push more AI features into local SERPs and Maps in 2026. Google has also started to test inserting a version of the Pack into AIOs. We can expect an acceleration of these efforts and a further blurring of AI and traditional SERPs (a la Web Guide) in 2026.

It's an Ad, Ad, Ad World

Google had its first $100 billion quarter in 2025, which won't be its last. Even with the challenge of AI, this reflects Google's continued market dominance. It also reflects Google's increasingly aggressive monetization of SERPs. Google is putting more ads in more places and making them harder to distinguish from organic content. A large percentage of Local Packs now feature ads (top or bottom). Another case-in-point is the recent change in how ads are presented at the top of the SERP. Google reduced individual ad labeling to a single scrolling mention at the top of the page and added a "hide sponsored results" button that allows people to collapse ads if desired. This appears to benefit users but, as a practical matter, further blurs the ads-organic content distinction. The more clearly ads are labeled, the less likely they are to generate clicks. Google understands this and has been incrementally trying for years to desensitize people to ads in the SERP. In Near Media's research we see people who explicitly resist traditional PPC ads. But that's less likely to happen with LSAs, which often capture a huge percentage of clicks and conversions because they carry many of desirable features of GBPs (e.g., reviews, images). Google is also testing larger local ads/LSAs that further reduce the visibility of organic links. And the company has started rolling out ads/local ads in AI Mode results and Gemini. Yet ads are the challenging, gatekeeping problem that Google must solve before it can fully deploy any AI-SERP substitute.

In August 2024, Judge Mehta of the US DC Circuit Court held that Google had created monopolies in search and search advertising. This year we got that court's largely cowardly and ineffectual ruling on remedies, with mostly wrist slapping for Google. Meanwhile, another federal court in a separate ad-tech antitrust case, in April, found Google liable for monopolizing the ad server and ad exchange markets. The court is now deciding on remedies. There were also two big privacy cases where Google was found liable. In a federal court class action, a jury awarded plaintiffs $425 million in damages. Google separately settled with the Texas attorney general for $1.4 billion in the second privacy case. Back in California, Yelp's local search antitrust case against Google survived a motion to dismiss the case and is moving forward. In Europe, the European Commission fined Google €2.95 billion ($3.4 billion) for abuse of market position for its self-preferential use of advertising technology (similar to the US ad-tech case). Though not yet final, Google also seems poised to lose its appeal of a 2018 fine of €4.3 billion ($5 billion) for anticompetitive practices regarding phone partner rules around access to the Google Play Store and app pre-install requirements. Then, in December, the European Commission opened a new antitrust investigation into Google's use of third party content in its AI answers. Despite all this litigation and regulatory activity, Google's power and position are mostly unscathed.

ChatGPT's Local Progress

More than 60% of ChatGPT users conduct local searches. Satisfaction and trust are mixed. Regardless it's not as popular an AI use case as, say, product research. And as OpenAI has pushed e-commerce and shopping improvements this year it wasn't clear they were going to meaningfully invest in local. That would've been a strategic mistake in their competition with Google and Google's AI tools. OpenAI recently regrouped and refocused in the face of Gemini 3's quality and speed improvements and it appears to have made some progress in local. The company rolled out Knowledge Panels across categories, which we first noticed in local results. In addition, although it has historically used Bing Places and other vertical and editorial sources for local business information, ChatGPT has more recently leaned more heavily on Google Business Profiles, directly and indirectly. Google has sought to disrupt this by suing OpenAI's scraping vendor SerpApi. That litigation will likely be resolved in 2026. Regardless, this will be a pivotal year for ChatGPT, which has ambitions of replacing Google as the "front door to the internet." That's unlikely to happen; more likely is that users will increasingly turn to AI for some types of research in specific contexts and verticals followed by Google searches for "last mile" information.

Suspensions, Verifications, Disappearing Reviews

There were a lot of issues this year around GBP suspensions, reinstatement delays and verifications, as well as confusion and frustration with disappearing reviews. Starting in Q1 and growing throughout the year, suspensions were up for both individual and bulk accounts, especially for service area businesses. It was never entirely clear why. We (Mike in this case) speculated that it might have been related to Google algorithm changes to improve data quality, perhaps tied to AI. Suspensions happened when users made even minor profile changes. The flip side of suspensions is reinstatement. And there were extended backlogs and wait times as the volume of suspensions grew. At the same time Google changed the verification process, limiting the number of unsuccessful attempts available to a business. Google also seemed to be tweaking and tightening its review algorithm to address suspicious reviews, throughout 2025. Whether it was this or bugs, businesses consistently complained of missing reviews (there were patterns). And once again, because of the volume of complaints, Google ended 2025 with a backlog of pending review removal appeals. Review extortion was a growing problem too, which Google initially dismissed then began to address. In addition, Google introduced review nicknames and said it would moderate review responses before publishing them. We haven't seen the last of these issues and problems.

But Is It SEO?

Even though many digital marketing folks and SEOs (unwisely) have adopted the term "GEO" for AI optimization, it hasn't fully stuck – thankfully. However the naming debate has always been a surrogate for the real question of whether SEO and AEO/GEO are the same or different, and if different how different? Is there really a separate discipline or is "good SEO" enough to generate AI visibility? Google continues to argue good SEO is enough and that SEO for LLMs is merely a sub-discipline of SEO. There are philosophical and technical arguments in support of and against that position. But a healthy contingent of SEOs argue that while the disciplines may overlap, they're distinct. Those on team GEO would argue the goals and methodologies are different. SEO is about rankings and clicks, AEO/GEO is about becoming a trusted source and being cited in AI responses. SEO seeks and assumes a human click; AEO/GEO more often assumes zero-click. Success in AEO/GEO is about inclusion and visibility rather than visits. Backlinks play less of a direct role with AI, though may play an indirect one (varies between Google AI and ChatGPT). SEO is focused on pages, AEO/GEO is focused on content and information that can be extracted, understood and presented by LLMs. There are also different KPIs and success metrics emerging for AEO/GEO. And, finally, probabilistic results and personalization make AEO/GEO a different beast than traditional SEO.

Clean-up: Bing, Siri, Meta & Alexa Plus

Microsoft momentarily seemed to have the upper hand when it partnered with OpenAI in 2022 and launched Bing Chat in early 2023. There were some preliminary indications that Bing might be able to steal a point or two of share. Microsoft CFO Amy Hood said, "For every one point of share gain in the search advertising market, it’s a $2 billion revenue opportunity for our advertising business." But that early optimism has faded. Bing market share is flat and Copilot adoption has apparently stalled, relegating Microsoft (despite its $3.6 trillion market cap) to continued also-ran status in search and now AI usage. Apple too has struggled to make AI headway, with Siri being the most obvious example, though that hasn't really hurt iPhone sales. Despite promises, AI-Siri has still not materialized and the ChatGPT integration is unimpressive. Apple has decided to go with Gemini for the guts for Siri 2.0. There has also been a shakeup among Apple executives. After multiple delays Apple is now expected to unveil the new Siri in Q1. Meanwhile, Amazon has been very busy with AI on the B2B side, but just rolled out a web version of Alexa Plus. It's no ChatGPT killer but could become widely adopted for shopping. Finally, Meta has been poaching everyone else's AI employees and overspending on AI. Most recently it bought Manus for more than $2 billion. This could be a prelude to making its own AI tools more useful or an admission of a confused AI strategy that hasn't worked – or both.

A busy and confusing as 2025 was this year promises to be even more consequential. More changes are coming to Google and AI will continue change consumer behavior – and the internet itself. We'll continue to track and analyze it all, through a local lens.

Thanks to all our readers, podcast listeners and those who subscribed to our newsletter in 2025. Slightly belated Happy New Year from David, Mike & Greg.

Drop us a line.