SERP feature types

format: "advanced" returns every SERP feature Serplify detects. Each item is discriminated by its type and carries the shared ranking fields (rank_absolute, rank_group, page).

Fully-typed features

These types have rich, stable, documented fields:

type	Description	Notable fields
organic	Classic organic result	domain, title, description, url, breadcrumb, highlighted, links, rating, price, images, faq
paid	Paid ad	domain, title, description, url, breadcrumb, description_rows, price, rating
featured_snippet	Answer box pulled from a page	domain, title, description, url, featured_title, table
ai_overview	Google AI Overview	items[] (text/markdown/links/images/references), references[], asynchronous_ai_overview
people_also_ask	”People also ask” accordion	items[] with title, seed_question, expanded_element[]
related_searches	Related search chips	items[] (strings)
refinement_chips	Refinement/filter chips	items[] with title, options[]
knowledge_graph	Knowledge panel	title, subtitle, description, url, image_url, cid, items[]
local_pack	Google Business (map 3-pack)	title, description, domain, phone, url, rating, cid

GMB rank tracking

Each business in the local pack is its own local_pack item and exposes cid (Google Maps customer id) — the most stable identifier for tracking a business’s map rank. Combine with rank tracking.

Additional detected features

These are also returned when present, with a common, flexible field set (title, description, url, items[], images[], rating, price, and type-specific extras):

images, video, short_videos, top_stories, twitter, map, shopping, popular_products, answer_box, currency_box, math_solver, people_also_search, jobs, events, recipes, carousel, multi_carousel, mention_carousel, app, commercial_units, compare_sites, courses, discussions_and_forums, explore_brands, find_results_on, found_on_web, google_flights, google_hotels, google_posts, google_reviews, hotels_pack, knowledge_graph_expanded_item, knowledge_graph_images_item, local_services, perspectives, product_considerations, questions_and_answers, scholarly_articles, third_party_reviews, top_sights.

Working with mixed types

Because items is a mixed list, switch on type:

for (const item of data.items) {
  switch (item.type) {
    case "organic":
      console.log(item.rank_absolute, item.domain, item.url);
      break;
    case "local_pack":
      console.log("GMB:", item.title, item.cid);
      break;
    case "people_also_ask":
      console.log("PAA questions:", item.items?.map((q) => q.title));
      break;
  }
}

Use feature_types on the result to see which types are present before iterating.