The Problem
Google SERPs in 2026 contain AI Overviews, featured snippets, People Also Ask, local packs, knowledge panels, video carousels, and shopping results. A change in any feature can impact your traffic. Most SEO tools track organic positions but miss feature-level changes. You might maintain your #3 organic position while losing featured snippet ownership, or an AI Overview might start answering your target query directly, both situations that reduce clicks to your site without changing your rank. Without feature-level monitoring, you only see the traffic drop, not the cause.
The Scavio Solution
Scavio returns SERP features as structured fields in the search response: ai_overview, featured_snippet, people_also_ask, local_pack, and more. Build a monitoring pipeline that snapshots every SERP feature for your target keywords and diffs against the previous snapshot. The pipeline detects when AI Overviews appear or change content, when featured snippets change owners, when new PAA questions surface, and when local pack composition shifts. Every change that could impact traffic is caught within 24 hours.
Before
Before feature-level monitoring, teams saw traffic drops without knowing the cause. A lost featured snippet or new AI Overview was invisible until someone manually checked the SERP days or weeks later.
After
After building feature detection, every SERP surface change triggers an alert within 24 hours. The team knows exactly why traffic changed and can respond to feature-level shifts the same day they occur.
Who It Is For
SEO teams who need to understand why traffic changed, not just that it changed. Anyone whose traffic dropped without a ranking change and suspects SERP feature shifts are the cause.
Key Benefits
- All SERP features returned as structured JSON fields
- Daily diff detects changes in AI Overviews, snippets, PAA, local pack
- Feature-level alerts explain traffic drops that rank tracking misses
- Catches AI Overview appearances that preempt organic clicks
- Historical feature data shows SERP evolution per keyword
Python Example
import requests
import json
from pathlib import Path
from datetime import datetime
API_KEY = "your_scavio_api_key"
def snapshot_features(keyword: str) -> dict:
res = requests.post(
"https://api.scavio.dev/api/v1/search",
headers={"x-api-key": API_KEY},
json={"platform": "google", "query": keyword, "ai_overview": True},
timeout=15,
)
res.raise_for_status()
data = res.json()
return {
"keyword": keyword,
"ai_overview": bool(data.get("ai_overview")),
"ai_overview_text": (data.get("ai_overview", {}) or {}).get("text", "")[:200],
"featured_snippet": bool(data.get("featured_snippet")),
"featured_snippet_source": (data.get("featured_snippet", {}) or {}).get("link", ""),
"people_also_ask": [q.get("question", "") for q in data.get("people_also_ask", [])],
"local_pack": bool(data.get("local_pack")),
"organic_top_3": [r.get("link", "").split("/")[2] if r.get("link") else "" for r in data.get("organic", [])[:3]],
}
def detect_changes(current: dict, previous: dict) -> list[str]:
changes = []
if current["ai_overview"] != previous.get("ai_overview"):
changes.append(f"AI Overview {'appeared' if current['ai_overview'] else 'disappeared'}")
if current["featured_snippet"] != previous.get("featured_snippet"):
changes.append(f"Featured snippet {'appeared' if current['featured_snippet'] else 'disappeared'}")
if current["featured_snippet_source"] != previous.get("featured_snippet_source", ""):
changes.append(f"Featured snippet source changed to {current['featured_snippet_source']}")
if current["local_pack"] != previous.get("local_pack"):
changes.append(f"Local pack {'appeared' if current['local_pack'] else 'disappeared'}")
return changes
def monitor_surfaces(keywords: list[str]):
history_path = Path("serp_features_history.json")
history = json.loads(history_path.read_text()) if history_path.exists() else {}
all_changes = []
current = {}
for kw in keywords:
snapshot = snapshot_features(kw)
current[kw] = snapshot
if kw in history:
changes = detect_changes(snapshot, history[kw])
if changes:
all_changes.append({"keyword": kw, "changes": changes})
history_path.write_text(json.dumps(current, indent=2))
if all_changes:
print(f"SERP CHANGES DETECTED: {len(all_changes)} keywords")
for item in all_changes:
print(f" {item['keyword']}:")
for change in item["changes"]:
print(f" - {change}")
else:
print(f"No feature changes across {len(keywords)} keywords")
monitor_surfaces(["best search API", "SERP API pricing", "web scraping tools"])JavaScript Example
const API_KEY = "your_scavio_api_key";
async function snapshotFeatures(keyword) {
const res = await fetch("https://api.scavio.dev/api/v1/search", {
method: "POST",
headers: { "x-api-key": API_KEY, "content-type": "application/json" },
body: JSON.stringify({ platform: "google", query: keyword, ai_overview: true }),
});
if (!res.ok) throw new Error(`scavio ${res.status}`);
const data = await res.json();
return {
keyword,
aiOverview: !!data.ai_overview,
featuredSnippet: !!data.featured_snippet,
featuredSnippetSource: data.featured_snippet?.link ?? "",
localPack: !!data.local_pack,
paa: (data.people_also_ask ?? []).map((q) => q.question ?? ""),
};
}
const keywords = ["best search API", "SERP API pricing"];
for (const kw of keywords) {
const snap = await snapshotFeatures(kw);
console.log(`${kw}: AIO=${snap.aiOverview}, snippet=${snap.featuredSnippet}, local=${snap.localPack}, PAA=${snap.paa.length}`);
}Platforms Used
Web search with knowledge graph, PAA, and AI overviews