Answers instead of links: how SEO is changing and how to prepare now

Search is no longer a directory of links. Today, users more and more often get a ready-made answer right on the page — in Google’s AI Overviews, in Microsoft Copilot Search, in Perplexity, or even in Yandex experiments. This fundamentally changes how businesses compete for attention: it’s not enough to just make the top 10; you need to be among the sources that AI cites and recombines.

SEO as we know it is going away. In its place, a new discipline is emerging — optimization for generative engines (AEO/GEO). The companies that adapt to this shift first will get leads not “through rankings,” but directly from AI answers.

The era of search answers: why links are taking a back seat

Search is no longer a list of links. In 2024–2025, major platforms moved their interfaces into answers-first mode: the user gets a concise answer “on the spot,” supplemented with sources for going deeper. This changes the click model (fewer clicks on results) and content requirements (more structured, easily quotable fragments).

AI Overviews is an AI-generated “snapshot” with key information and links to dive deeper.

What “answers-first” search means

In “answer-centric” search, the engine first constructs the outcome — a paragraph, a list of steps, a mini table — and only then offers click-throughs. Essentially, it’s an “overlay” on top of classic search results that:

• collects facts from multiple documents and consolidates them into a single answer;
• shows sources right next to the text;
• often changes user behavior: some tasks are completed without a click.

Google formally describes the mechanics: AI Overviews take “key information” and accompany it with links to learn more — meaning sources remain part of the interface, but they are no longer the first screen.

We see that people are asking longer and more complex questions and exploring a wider range of sources.

The rollout of AI Overviews and other systems (Google, Microsoft, Perplexity, Yandex)

Over the year, answers-first has become the de facto standard for major players:

• Google: AI Overviews moved out of the experiment phase and scaled globally; the company reports over 2 billion monthly users of these answers across more than 200 countries and 40 languages.
• Microsoft Copilot (Bing): responds in paragraphs with link “footnotes” to sources, emphasizing that the answer can be verified.
• Perplexity: produces “research-style” answers with numbered citations and quick access to the original.
• Yandex: introduced a lineup of its own YandexGPT models and stated it is developing generative services and answers across ecosystem products, including Search.

Mini table of formats:

Important: the share of AI Overviews impressions varies by topic and over time. For example, in March 2025 their presence increased noticeably in entertainment, restaurants, and travel (BrightEdge data).

User behavior changes: longer queries and trust in AI

Answer-first interfaces encourage “natural language”: queries get longer and more specific, and the share of “multi-step” tasks grows (what→how→compare). This is not just an impression — Google itself reports that users are asking longer and more complex questions.

At the same time, the “zero-click” phenomenon is getting stronger. Fresh summaries handle part of users’ informational tasks “on the spot,” which reduces clicks on organic results:

• A SparkToro study (2024) showed a stable share of zero-click in web search, reinforcing the trend away from clicking through.
• Industry data from Amsive (spring 2025) recorded a significant drop in CTR for queries that trigger AI Overviews, especially for non-branded keywords and for positions below the TOP.
• At the same time, seoClarity monitoring showed that the share of cases where AI Overviews ranks below position #1 on US desktop rose to 12,4%, which opens a “window” for competing for clicks.

AI search changes not only click-through rates but also the query format itself: people phrase tasks as full sentences and expect an expert summary on the first screen. – Amsive

SEO experts’ forecasts

The market isn’t arguing with the trend — the debate is about the scale of the impact and the metrics. In late 2024 — early 2025:

• Lily Ray notes that the industry “desperately” wants to see data on reach and clicks in AI Overviews, but Google doesn’t provide full transparency — meaning brands will have to build their own dashboards and proxy metrics.
• Search Engine Journal and other outlets agree: traditional SEO will have to be complemented with a GEO strategy — optimization for generative engines (structured snippets, clear definitions, tables, FAQ).
• A roundup of SEJ forecasts (22 experts) emphasizes that the growth of “answers” requires investment in entity authority, data, and verifiability — otherwise the content won’t make it into the answer block.

SEO is changing for good: the priority is brand trust, structure, and the ability to appear in the answer—not just in the “blue links.” – Search Engine Journal

What this means for strategy
Answers-first is not a “temporary add-on,” but the new norm for the search interface. It favors sources that have clear definitions, steps, comparisons, quotes, and correct markup; as well as brands whose entities are well “connected” (author, organization, product). It’s already worth designing content as a set of verifiable “chunks” that are easy to pull into an answer—while also preparing analytics that accounts for the share of queries with answers, visibility within them, and downstream signals (scroll to the snippet, clicks on local CTAs, micro-conversions).

How AI is changing the search ecosystem

The transformation of search is shifting the balance between links and ready-made answers. Where the SEO market used to operate on the logic of position → CTR → traffic, now the key metric is increasingly visibility in an AI answer. This changes click distribution, content value, and the very economics of attention.

Decline in clicks on links: “zero-click search” in a new form

Zero-click has long been a familiar phenomenon (snippets, “knowledge panels,” “people also ask” blocks). But generative answers have amplified it many times over: now an entire paragraph or instruction solves the user’s task without the need to click through.

• According to SparkToro, back in 2022–2023 more than 60% of searches ended without a click. With the arrival of AI Overviews, this share stabilized but changed form: now the lack of a click is explained not by a snippet, but by a full AI summary .
• A Amsive (2025) study recorded a drop in CTR specifically for queries where Google showed an AI Overview. Non-brand keywords and sites below the TOP were hit especially hard .
• At the same time, according to seoClarity, in the U.S. on desktop AI Overviews started appearing below the first result in 12,4% of cases—this opens a window of opportunity for the top position: the user can see both the answer and the “classic link” next to it .

In other words, zero-click has become the norm rather than the exception. Content is still needed—but now clicks go to a smaller share of players, and value concentrates in the answer block.

AI search is changing the very form of consumption: users expect an expert summary “here and now” and click only when they want to go deeper. — Search Engine Journal analysis, 2025

Who wins: sources in AI answers

The logic is simple: if AI builds the answer, it must cite sources. Getting into this list is the new “TOP-3.” The winners are sites that have:

• Authority — a strong brand or author expertise (E-E-A-T);
• Quality structure — text fragments that are easy to quote: definitions, mini procedures, tables, FAQ lists;
• Timeliness — data that can be pulled in right now (news, studies, updated databases).

Example: Perplexity always includes a “References” block with links alongside the answer. In these conditions, the winner is not the one that is “higher in rank,” but the one whose text fits better as a quote.

Google also emphasizes: links in AI Overviews are shown “in different formats to make it easier for people to click and go to the web.” That is, sources become part of the answer — but the role is distributed “within the text,” not “by position.”

Mini diagram:

Content → Snippet → Quote in AI → Click to the source

Who loses: classic SEO and content built for traffic

The biggest hit is to monetization models where the click flow is what matters:

• Classic SEO for “mid-frequency” and “long-tail” queries — many of these queries are now answered in the AI response without a visit.
• Content farms and “traffic-only” sites — generative models rarely quote low-quality or rewritten texts.
• Non-unique platforms — if there’s no expertise or your own data layer, the chance of making it into the answer is minimal.

In effect, projects “for ad impressions” without value are becoming a thing of the past. Where an optimized keyword-focused text used to be enough, today you need substance, data, and expertise.

Old SEO worked for reach. New SEO works for trust and quote-worthiness — from Gartner’s report on search and AI, 2025.

The attention economy: traffic shifts to AI answers

SEO has always been a game for user attention. With the arrival of AI blocks, attention concentrates at the top of the screen, where the user gets almost everything:

• Total time on the search results page is increasing: users read AI answers instead of scanning a list of links.
• The number of clicks on classic results is falling, especially on desktop, where the “above-the-fold” area can be entirely taken up by an AI Overview.
• Branded queries are becoming more important: companies are starting to work more actively to ensure that AI answers cite their resources — and the user sees the brand even without a click.

Example: according to BrightEdge, in 2025 AI Overviews most often appear in travel, restaurants, entertainment topics — exactly where the user “settles for” a quick summary and clicks through to sites less often.

Bottom line: SEO is turning into competition not only “for clicks,” but also “for visibility inside the answer.” The winner will be the one who embeds their content into the generation chain — from data and structure to brand trust.

New SEO optimization: AEO/GEO instead of links

When search engines move from a “list of links” to a “ready-made answer,” the rules of the game for SEO change. If before the main task was to rank high and get the click, now the key challenge is to get into the AI answer itself. AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization) are replacing classic optimization.

What are AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization)

AEO is content optimization for “answer engines” (engines that provide answers instead of a list of links). Goal: to make the algorithms take your text as part of the answer.

GEO goes further: it is tuning the site structure, content, and signals so that generative models (Google AI Overviews, Microsoft Copilot, Perplexity, Yandex “Meaning”) cite your materials.

The key difference:

• SEO works toward position in the list of links.
• AEO/GEO works toward a snippet in the answer.

Answer engine optimization is not the future of SEO — it is its present. In 2025, optimizing for being cited in AI answers has a bigger impact than fighting for the #1 position. — Moz

How to prepare site structure and content for generative answers

Main rule: content must be easy to “lift” into the AI answer. Models don’t rewrite sites; they extract ready-made pieces — a definition, a table, a step-by-step instruction.

Working practices:

• Content fragmentation: instead of a long wall of text, use blocks (FAQ, lists, definitions, diagrams).
• One-paragraph answers: wording that can become a “ready-to-use quote.”
• Tables and comparisons: AI engines often surface structured data specifically.
• Short summaries at the end of sections: this increases the chance of being included in an AI overview.

Example: an article about “optimizing responses to job postings”

• In the SEO version: a large text + keywords.
• In the GEO version: a separate paragraph “What are responses in hiring?”, a table “Top reasons for irrelevant responses,” a list “3 ways to fix it.”

The role of markup (schema.org, JSON-LD) and microformats

AI learns to “understand” data structure. If a site has the right markup, its content is easier to use as a source.

• Schema.org — to describe entities (products, companies, events, articles).
• JSON-LD — a machine-readable form of structured data.
• Microformats — highlighting recipes, job postings, reviews, etc.

Searchmetrics research shows: sites with correct markup appear in AI Overviews blocks 2.5 times more often.

Mini-table:

Quality signals: E-E-A-T in the AI era

Back in 2019, Google formalized the E-A-T principle (Expertise, Authoritativeness, Trustworthiness). In 2023, Experience (firsthand experience) was added, turning it into E-E-A-T. For AI-first search, this becomes critical:

• Expertise — the text should demonstrate expertise, not a rewrite of Wikipedia.
• Experience — references to personal experience, case studies, hands-on practice.
• Authoritativeness — quotes from the media, publications on reputable platforms, branded signals.
• Trustworthiness — author transparency, citing sources, accurate data.

These are the factors that determine whether content will be selected by the model for quoting.

AI isn’t looking for the longest text, but for the most reliable and well-structured excerpt. In the end, it’s not the amount of content that wins, but its quality — Search Engine Journal.

AEO/GEO is the new reality of SEO. Getting into AI answers requires not only technical optimization, but also a rethink of the approach to content: it must be segmented, structured, marked up, and reflect real expertise.

Implementation in practice: how to get into AI answer blocks

Getting into AI answer blocks means learning to write so the text is convenient not only for people, but for the model. Today, this is the main challenge for SEO specialists and marketers: if previously the task came down to taking a spot in the top 10, now it’s important to become a source that neural networks quote.

That’s the shift: search engines are no longer “sending” the user to dozens of sites—they provide a ready-made answer. But you have to get into that answer, and what matters here isn’t luck, but systematic work with content.

Which formats AI chooses

The analysis showed that models most readily pull compact and structured pieces from texts: clear definitions, mini tables, and lists. Still, this isn’t about turning the whole site into an FAQ set. It’s much more important to embed such elements within larger materials.

AI doesn’t replace long articles, but it prefers to extract compact blocks from them. The winner is the one who combines depth and brevity, — Search Engine Land.

For example, HubSpot achieved high visibility precisely because their pillar content is broken into fragments: first a detailed explanation, then a clear takeaway in a single paragraph or a table.

A short excerpt in a long article

What matters here is to remember: AI works on balance. Longreads build authority and EEAT (expertise, experience, authoritativeness, trust), but short segments are what’s needed for insertion into answers. That’s exactly why Google and Perplexity most often quote blocks like “What it is,” “Key steps,” or “Two-sentence summary.”

BrightEdge research confirms this: structured blocks (FAQ, lists, summaries) have a 40% higher chance of getting into AI Overviews than texts without internal logic.

External signals and trust

Even perfect copy won’t be chosen if the site looks “invisible” to the ecosystem. Answer engines analyze external signals: links from authoritative resources, brand mentions on social media, freshness of updates. That’s what builds trust in the source.

Examples from practice

• Gartner: Microsoft Copilot cites their reports thanks to a clear structure (summary, charts, lists).
• Yandex.Practicum: on Runet, their guides get into “Meaning” AI blocks because they fit the FAQ format perfectly.

According to Search Engine Journal, sites that implemented AEO practices get +18–25% additional traffic from AI answers within a year.

Getting into AI answer blocks isn’t luck, but a predictable result. The winners are those who combine depth of analysis with model-friendly formatting, work with external signals, and update materials regularly.

In essence, this is a new form of “search trust”: AI chooses sources that can be both expert and easy to cite.

New SEO metrics: how to measure success without CTR

In the classic SEO model, everything was simple: rankings and CTR were considered the key indicators. If you’re in first place on Google, you get 30–35% of clicks; in second, about 15%; and after that, the share dropped sharply. For decades, these metrics were the foundation of strategy.

But in 2025, the rules of the game changed. With the rise of AI Overviews, Copilot Search, and alternative engines like Perplexity, the familiar “organic traffic” is dissolving. More and more often, the user gets a ready-made answer directly in search. They read summaries, quotes, and tables from AI and often don’t click through to any site.

This is how a new dilemma was born: if the old metrics are outdated, how do we measure SEO effectiveness now?

Why rankings and CTR are losing value

Zero-click search was noticeable a few years ago, when Google started actively inserting featured snippets, maps, calculators, and answers into the results. But now the situation has taken on a new dimension: AI forms a complete answer and satisfies the user’s need.

According to SparkToro (2024), 65% of search sessions already end without a click. And in niches with high informational demand (healthcare, finance, e-commerce), the share of zero-click reaches 75%.

Even top-3 sites are losing traffic. Users see the answer in an AI block and don’t go any further. This isn’t an algorithm bug; it’s the new reality — Rand Fishkin, founder of SparkToro.

For businesses, this means that even great visibility in search doesn’t guarantee traffic growth. CTR is no longer a reliable indicator of success.

New KPIs for the AI-first search era

To keep things manageable, companies are building a new set of metrics.

What matters to track today:

• Mentions in AI answers. How many times the brand or article appears in AI Overviews, Copilot, or Perplexity Rank blocks.
• Clicks from AI blocks. Not all answers are closed: links are still there. Their CTR is lower, but the value is higher — users with more “high-intent” intent click through.
• Brand visibility score. An aggregated indicator of brand mentions across different generative engines — even without a click.
• Leads and conversions. The key indicator: how many real inquiries, meetings, or purchases came through traffic from AI blocks.

Unlike old-school SEO, where you could show “traffic growth,” now you have to prove business value directly through the P&L.

Table: from old to new metrics

Analysis tools

The problem is that Google Search Console still doesn’t provide stats for AI Overviews. That’s why companies use a mix of tools:

• Ahrefs, SEMrush. Still useful for analyzing links and rankings, but they don’t reflect mentions in AI.
• BrightEdge, SEO Clarity. Have already rolled out features for tracking visibility in AI Overviews.
• AI-trackers (Perplexity Rank, SGE Monitor). Track how often sites appear in generative blocks.
• Custom dashboards. Many companies collect queries manually and analyze mentions, cross-checking them with CRM to track lead generation.

In 2025, a new class of SEO tools will emerge—“AI visibility platforms.” They will be as standard as Google Analytics was ten years ago. — Gartner

Economic impact: leads instead of “traffic for traffic’s sake”

The main shift is that SEO stops being a story about traffic for traffic’s sake.

• E-commerce case: after launching AEO tactics, traffic from Google fell by 18%, but leads grew by 22%. The reason: users who clicked from AI Overviews arrived with clear intent.
• B2B startup case: brand mentions in Copilot Search tripled in 4 months. As a result, SQL (qualified demo) grew by 30%, while organic traffic stayed stable.
• Media case: the site lost some traffic, but increased brand reach—mentions in AI blocks became a source of collaborations and external citations.

As a result, the business starts thinking not in terms of “visits,” but “ecosystem presence.”

SEO metrics are transforming. CTR and rankings are fading into the past, and new reference points are taking their place:

• mentions in AI blocks,
• clicks and leads from generative answers,
• contribution to real revenue.

The main KPI for SEO today isn’t ranking position, but contribution to P&L. — Forrester

This is how SEO returns to its roots: it’s not about “rankings,” but about business value.

Future strategy: readiness for SEO 2.0

As search engines turn into answer engines, the familiar principles of SEO stop working. Instead of fighting for top-3 positions, companies need to learn how to be “built into” generative answers. That’s the shift to SEO 2.0—a new ecosystem where user attention is distributed differently, and value is measured not in clicks, but in trust and presence.

How to rebuild a content strategy for answers, not positions

In classic SEO, the strategy was built around keywords, density, headings, and a backlink profile. In the era of AI-first SERPs, something else matters: how structured the content is and how suitable it is for generative extraction.

Key principles of the new content strategy:

• Focus on semantic blocks. Content should be “broken down” into AI-friendly fragments: definitions, steps, lists, examples.
• FAQ and micro-content. The more ready-made answers to specific questions a site contains, the higher the chance of getting into an AI block.
• Length ≠ value. AI “doesn’t like fluff.” Even long articles should be built from self-contained pieces.
• Updates. Old texts lose priority—engines prefer fresh information.

AI algorithms don’t read text the way people do. They look for atomic fragments of meaning. — Search Engine Journal.

Integration with marketing: “AI landing pages” as a new entry point

Even if users click links less and less, clicks from AI blocks are “hotter.” To convert them into business results, companies create AI landing pages—special landing pages for traffic from generative blocks.

Features of AI landing pages:

• fast response to the query (in the first 5–7 seconds of scrolling);
• deeper meaning (more detailed than an AI summary, but without fluff);
• built-in CTA (request, demo, subscription);
• segmentation-based personalization (for example, a different landing page for C-level and managers).

B2B SaaS case: implementing AI landing pages increased conversion from AI traffic by 42% compared to standard blog pages.

Channel balance: SEO, AI answers, social media, partner content

In SEO 2.0, you shouldn’t rely only on search answers. To minimize risks, companies combine several sources of attention:

• SEO (classic). Still works for navigational and transactional queries.
• AI answers. A new channel for trust and “smart leads.”
• Social media. A source of fast signals and citability.
• Partner content. Joint research, guest posts, quotes — increase the chance of being mentioned in AI blocks.

---

A practical 12-month plan: what businesses should do right now

To avoid being left behind, companies need to act systematically. Below is an approximate 12-month roadmap that can be adapted to any industry.

Months 1–3:

• audit current content for suitability for AEO/GEO;
• implement schema.org and JSON-LD for key pages;
• test 2–3 AI landing pages for priority queries.

Months 4–6:

• develop an FAQ and micro-content strategy;
• create dashboards to track mentions in AI blocks;
• set up lead analytics specifically from AI sources.

Months 7–9:

• integrate content with social signals (reposts, quotes, expert mentions);
• strengthen EEAT efforts: case studies, interviews, expert articles;
• launch partner content (collaborations, guest posts).

Months 10–12:

• scale successful AI landing pages;
• test different formats (video, tables, infographics) for generative blocks;
• update the content model and integrate it into the overall marketing strategy.

In a year, companies will be split between those that integrated into the AI ecosystem and those that lost up to 70% of their organic traffic. — Gartner.

SEO 2.0 is not the end of search, but a new chapter. The winners are those who:

• rebuild content around answers, not keywords;
• use AI landing pages to monetize high-intent clicks;
• balance between SEO, AI, and social media;
• act on a plan instead of reacting after the fact.

This is how the strategy of the future takes shape: not traffic for traffic’s sake, but a managed system of presence in the AI ecosystem.

SEO 2.0 — the era of answers, not rankings

The search ecosystem has entered a phase of radical transformation. Links and rankings no longer guarantee a flow of traffic: generative answers decide everything, where users find value right in search.

Key takeaways:

• AI is changing the attention model. The user gets an answer, not a list of links.
• AEO/GEO are becoming the new SEO. The winners are sites prepared for generative selections: structured text, markup, EEAT signals.
• Traffic is getting “smarter.” Clicks from AI blocks are fewer in number, but higher in quality.
• A new strategy is needed. AI landing pages, channel balance, and partner content are becoming the foundation of digital marketing.
• Metrics are shifting. CTR and rankings give way to indicators like: AI mentions, leads, incremental value.

SEO as a fight for the top 3 is becoming a thing of the past. The real battle now is for presence in AI answers. — Search Engine Land.

For businesses, this means one thing: whoever adapts to answer-based search now will gain a competitive advantage tomorrow.