How to Target SERP Features
Last updated View as Markdown
Targeting a SERP feature means meeting that feature’s own eligibility rules, and they differ more than the shared label suggests: structured data for rich results, a Merchant Center feed for free product listings, a Google Business Profile for the local pack, and clear, well-structured answers for featured snippets and AI Overviews. Measuring the result is harder than earning it, because Search Console does not report featured snippets separately, and its standard Performance report counts AI Overview and AI Mode appearances within Web search, with only a separate, impressions-only report isolating them.
This guide covers how to optimise for each major SERP feature. For background on what features exist, how they affect click-through rates, and how to factor them into keyword research, see SERP Features.
AI Overviews
What they are
AI Overviews are synthesised summaries generated by Google’s AI (based on Gemini) that appear at the top of the results page for a wide range of queries. Unlike featured snippets, which extract from a single source, AI Overviews draw from multiple pages and present a synthesised answer in Google’s own words, with source citations.
How to target AI Overviews
There is no direct mechanism to submit content for AI Overview inclusion. Google selects sources based on the same signals it uses for quality assessment generally. The overlap with featured snippets is measurable, though: SE Ranking found in November 2024 that when a featured snippet and an AI Overview appeared on the same US desktop results page, the snippet’s source was among the AI Overview’s citations 66% of the time.1 The same content characteristics that earn featured snippets tend to earn AI Overview citations.
Targeting principles:
- Answer the query directly in the opening paragraph or an early H2
- Use clear, factual language rather than hedged or promotional language
- Support answers with specific data, dates, or figures
- Structure content so individual sections are self-contained answers: Google often pulls from mid-page sections, not only introductions
- Build E-E-A-T signals (authorship, cited sources, original research)
Monitoring: there is no AI Overviews filter in the standard Performance report. Google states that sites appearing in AI features “are included in the overall search traffic” and are reported within the Web search type, mixed in with ordinary blue-link impressions.2 The Search Generative AI Performance report (below) is the only surface-level breakdown Google offers, and it carries no click or query data. Anything more granular needs a third-party AI visibility tool.
AI Mode
What it is
A dedicated AI-powered search interface within Google Search, using Gemini to handle complex, multi-step queries through a conversational session. Unlike AI Overviews, which are inline features in standard results, AI Mode is a separate experience where follow-up questions carry context across the conversation. As of May 2026, AI Mode has over one billion monthly users.3
How to target AI Mode
AI Mode draws on the same quality signals as AI Overviews but with greater emphasis on complex, multi-part queries. Because users ask follow-up questions within the same session, content that covers a topic comprehensively, including surrounding questions a researcher might ask, performs better than a single tightly scoped page.
Targeting principles:
- Cover the topic’s related questions, not just the primary keyword. AI Mode uses query fan-out: it generates related sub-queries to build its answer, pulling from pages that address those sub-queries directly
- Structure each section as a self-contained answer: AI Mode pulls passages, not whole pages
- Make your entity clear and consistent: brand, product, or service descriptions should match across your site, structured data, and third-party sources
- The same E-E-A-T signals apply: authorship, cited sources, original research, and external validation
Monitoring: the Search Generative AI Performance report in Search Console shows impressions from AI Overviews and AI Mode as a separate view, with no click or query data; Discover’s AI features have a generative AI report of their own. It launched to UK website owners in June 20264 and reached all websites worldwide on 31 August 2026.5
Featured snippets
What they are
A single extracted answer, shown above position 1 in the organic results (position 0). Google pulls the text, list, or table directly from a page. Three formats:
- Paragraph: a 2–5 sentence answer to a direct question
- List: numbered or bulleted steps or items
- Table: comparative data (prices, specs, comparisons)
How to target featured snippets
Know how much is left to win. Snippets are far rarer than they were: across 1,000,000 US desktop SERPs, Ahrefs found them on 5.53% in June 2025, down from 15.41% in January, while AI Overviews rose from 3.93% to 27.43% on the same sample.6 The work below still pays because the same signals feed AI Overview citations, but a snippet-only target list is a shrinking one.
Identify snippet opportunities. Target queries formatted as direct questions: “what is X”, “how does X work”, “how to do X”, “X vs Y”. These trigger snippet formats most reliably. Search Console will not tell you which snippets you hold: featured snippets have never been a Search Appearance type, and a snippet simply reports as position 1. To find which of your queries trigger a snippet and who owns it, use a rank tracker that records SERP features, or check the SERPs by hand for your priority queries.
Format content to match snippet types:
For paragraph snippets:
- Open your answer with a 40–60 word direct response immediately after an H2 or H3
- Don’t bury the answer in supporting context
For list snippets:
- Use proper
<ol>or<ul>HTML lists - Each item should be a distinct, complete step or item
- 5–10 items is the typical range Google pulls
For table snippets:
- Use proper HTML
<table>markup - Include a header row
- Keep tables comparative (two or more items being compared)
Write for the question, not around it. The opening sentence of a section should answer the question posed by the heading, not introduce it. “Title tags are the text that appears in a browser tab and in search results” earns more snippets than “In this section, we examine what title tags are and why they matter.”
Monitor and iterate. Ahrefs and Semrush both show which of your pages hold featured snippets and which competitor pages hold snippets you could displace. This is the tracking gap Search Console leaves, so the third-party tool is doing real work here rather than duplicating GSC.
People Also Ask
How to target PAA
Mine PAA for content gaps. PAA questions show what users ask after (and alongside) your target query. If your page doesn’t address PAA questions, a competitor’s page that does has an advantage.
Structure content as Q&A. Use question-formatted H2 and H3 headings followed by a concise answer paragraph. PAA often pulls from the same content format as featured snippets.
FAQPage structured data no longer produces a rich result, since Google removed FAQ rich results from Search in May 2026,7 and no platform documents reading it for AI answers.
Knowledge panels
How to influence knowledge panels
You cannot submit directly for a knowledge panel. Google says panels are created automatically “when there is enough information available on the open web”, so the constraint is independent sources describing you consistently, not your own markup. The order matters, and the knowledge panel guide works through it in full:
- Build third-party corroboration first: author bios on publications you do not own, speaker pages, interviews, and coverage in authoritative publications
- Add structured reference entries second, Wikidata before Wikipedia: Wikidata’s bar is “serious and publicly available references”, and a Wikipedia article you do not clearly qualify for gets deleted
- Make your own signals consistent last:
OrganizationorPersonschema on one canonical page,sameAspointing only at profiles that are genuinely the same entity, and the same name and NAP everywhere
Two verification-based shortcuts skip the corroboration work: a verified Google Business Profile for a business with premises or a service area, and a claimed Google Search profile for creators above the follower threshold. If you already have a panel, claiming it lets you suggest corrections.
Local pack (map pack)
How to rank in the local pack
Local pack ranking is determined by three factors:
- Relevance: how well your Business Profile matches the query
- Distance: proximity to the searcher
- Prominence: authority based on reviews, links, mentions, and Business Profile completeness
Practical actions: complete your Google Business Profile fully (categories, attributes, photos, hours), build local citations, maintain review velocity, and earn local backlinks.
Local pack ranking is separate from organic ranking: a site can rank in the local pack without ranking organically, and vice versa.
Shopping carousels
How to appear in shopping carousels
The paid shopping carousel is a Shopping ads placement and needs a Merchant Center account, an approved product feed and an ads budget; no organic signal earns it. Free product surfaces are different. Google’s product structured data documentation says you can “add Product structured data to your web pages, upload data feeds with Google Merchant Center and opt into free listings within the Merchant Center console, or both”, and that doing both “maximizes your eligibility”.8 For e-commerce sites, this means:
- A verified Merchant Center account with a complete, accurate feed (title, description, price, availability, GTIN), opted into free listings
- Merchant listing structured data on every page where the product can be bought, which is the route that works without a feed
- No policy violations in either
See product schema for the difference between merchant listings and product snippets, which have different requirements.
Video carousels
How to appear in video carousels
Video SEO covers the rules in full; the short version is that most pages with a video are not eligible.
- Put the video on a watch page, one whose “main purpose is to show users a single video”. Only a video embedded on a watch page is eligible for video features.9
- Make sure the watch page is indexed. Google adds that it “must be performing well in Search” before the video is considered.9
- Give each watch page a title and description “unique to that video”.9
- Add
VideoObjectstructured data to the watch page. - A YouTube embed can be indexed twice: Google “may index the video both on your web page and on the third-party platform’s equivalent page”.9
Rich snippets
Schema types that enable rich snippets
Rich snippets enhance an existing organic result with additional information from structured data: star ratings, prices, event dates, recipe details. They don’t change ranking position but can improve CTR by making your result more visually distinct.
| Schema type | Rich result elements |
|---|---|
Review / AggregateRating | Star rating, review count |
Product | Price, availability, rating |
Event | Event date, location, tickets |
Recipe | Cook time, rating, calories |
JobPosting | Job title, company, location |
Article | Better title text, images and dates in results; Google states it is not required for Top Stories eligibility |
Implement schema using JSON-LD in the <head> of the page. Validate with Google’s Rich Results Test before deploying.
Note: HowTo rich results were deprecated in September 2023 and FAQPage rich results in May 2026. Neither produces a visual rich result in Google Search any longer, and no platform documents reading either for AI answers, so do not implement them expecting anything.
Sitelinks
Sitelinks appear automatically; they are not submitted or applied for. They were long treated as a branded-query feature, and that is no longer accurate: Ahrefs recorded sitelinks on 84.95% of US desktop results in June 2025, up from 8.44% in January, a jump its data scientist attributes to scroll-to sitelinks, which unlike internal sitelinks appear on non-branded queries too.6 Assume they are in play on most results rather than only your brand terms. To influence which pages appear:
- Ensure your most important pages are prominently linked from the navigation
- Use descriptive anchor text in internal links
UseGoogle’s Sitelinks Search Box has been deprecated and no longer appears as a visual SERP featureSearchActionsitelink searchbox schema if you want a search box shown in sitelinks
Generative UI
What it is
From summer 2026, Google Search can generate custom interactive interfaces for specific queries: comparison tables, calculators, simulations, and data visualisations built directly in the results page.3 Most common for comparative, decision-oriented, and research queries.
How to appear in Generative UI results
Generative UI pulls data from pages that make key information explicit and machine-readable. Pages that state figures clearly, use comparison tables, and label specifications are more likely to be sourced.
Practical steps:
- State facts explicitly: prices, specs, eligibility criteria, and feature comparisons should be clear in prose, not embedded in images or PDFs
- Use HTML tables for comparative data: these are easy for Google to extract and replicate in a generated interface
- Keep structured data accurate and complete (
Product,HowTo,Event, and similar schema types) - For e-commerce, maintain an accurate Merchant Center feed: product comparisons in Generative UI may draw from feed data as well as page content
Tracking SERP feature performance
Google Search Console: the Search Appearance filter in the Performance report breaks out impressions and clicks for the features Google derives from structured data and similar signals: review snippets, product results, videos, events, job postings, merchant listings, translated results and the like. Know its limits before building reporting on it. Featured snippets are not among the types, and AI Overview and AI Mode impressions are folded into the Web search type rather than exposed as a filter.2 For those two, GSC alone will not tell you what you hold.
Third-party tools: Ahrefs, Semrush, and Moz all track SERP feature ownership by keyword, which is what fills the featured snippet gap above. Use these to see which competitors hold features you don’t.
Manual spot-checks: search target queries in an incognito window. Check what features appear, whose content is cited, and how much of the page organic results occupy before the fold.
Footnotes
-
Google AI Overviews Research: 2024 Recap & 2025 Outlook — SE Ranking ↩
-
New opportunities, control and insights for website owners — Google, 3 June 2026; reported by Search Engine Roundtable ↩
-
Introducing Search Generative AI performance reports in Search Console — Google Search Central. Carries Google’s note: “As of August 31, 2026, we’ve rolled out these insights to all websites worldwide.” ↩
-
Goodbye, Featured Snippets: How SERP Features Have Evolved in the AI Era — Ahrefs ↩ ↩2
-
Deprecating the FAQ rich result feature — Google Search Central changelog ↩