SignalSumo

SignalSumo Data at a Glance: What Powers Every Report You Run

Written by SignalSumo Team
26 Aug 2026
6 min read
Key Takeaways
  • 5.4 billion keywords across 184 geographic databases, with volume, difficulty and SERP history recalculated for each market.
  • 2.65 trillion backlinks from 411 million indexed domains, with 5.88 billion pages crawled per day and 57% of links re-verified within 90 days.
  • 24.9 million AI-search prompts tracked across five AI engines — ChatGPT, Google AI Overview, Google AI Mode, Perplexity and Gemini — with brand mentions, cited sources and outbound links captured every check.
  • The full stack is available through one API and a native MCP server, so the same data feeds dashboards, spreadsheets and LLM workflows.
Core Definition

"SignalSumo is powered by a proprietary index of 5.4B keywords, 2.65T backlinks and 24.9M tracked AI prompts, refreshed by our own crawlers rather than a re-sold third-party feed."

Every report you open in SignalSumo — a keyword idea, a backlink profile, an AI-search visibility check — is answered from an index we build and refresh ourselves. This page is a plain-language tour of what's in that index today, how often it refreshes, and which markets are covered.

Numbers below are current as of this quarter and grow week over week.

The keyword database

Keyword data is the foundation of almost every SEO decision, so it's the part of the platform we've invested in most heavily.

  • 5.4 billion keywords in the live index
  • 184 geographic databases, split by country and language
  • 2.15 billion domains profiled with their ranking keywords, estimated traffic and paid-search footprint
  • Search volume, keyword difficulty, cost-per-click, intent and SERP features recomputed independently per market

Splitting the index by market matters because "best crm" in the United States is not the same query as "best crm" in Germany or Brazil. Different competitors rank, different SERP features appear, and difficulty can swing by 30+ points. SignalSumo scores each market on its own data instead of applying a global fudge factor.

Coverage by market

A few examples of what a single country database looks like:

  • United States — 582M keywords, 141M ranking domains
  • United Kingdom — 749M keywords, 104M ranking domains
  • Germany — 657M keywords, 63M ranking domains
  • Canada, Australia, India, France, Spain, Italy, Brazil and 176 more markets sized on request

Every one of these databases feeds the same tools — Keyword Research, Rank Tracker, Competitor Research, Content Editor and the AI Overview checker — so the market you pick in one tool is the market you get everywhere.

Backlinks are only useful when the index is both wide (finds obscure links) and fresh (drops the dead ones quickly). Ours is designed around both.

  • 2.65 trillion live backlinks
  • 411 million indexed root domains
  • 5.88 billion pages crawled every day
  • 57% of the index re-verified inside 90 days, so lost and redirected links surface fast
  • Per-link toxicity scoring, follow/nofollow, link placement, first-seen and last-seen dates

That refresh cadence is why the Backlink Monitor can page you within a few days of a link going missing, and why the Backlink Gap tool is not showing you competitors' links from three years ago.

The raw index is only step one. On top of it we compute:

  • Domain and page authority signals — trust flow, spam markers, topic clusters
  • Historical link velocity — new-vs-lost curves used by the Backlink Overview health gauge
  • Referring anchor and IP diversity — the inputs behind toxic-link alerts
  • AI-explained context — one-click "why is this link valuable?" summaries in the drawer

SERP and rank-tracking data

Rank Tracker doesn't rely on a general web crawl. Every keyword you add is checked at your chosen frequency through a dedicated SERP-monitoring layer that captures the full result page — not just position numbers.

  • Top 100 organic results per keyword
  • All SERP features present — AI Overview, People Also Ask, featured snippets, video packs, image packs, local packs, shopping, sitelinks
  • The domains that own each feature, saved so you can see who's taking your click share
  • Historical snapshots so ranking gains and losses can be replayed day-by-day

Local Rank Tracking uses the same layer with a geolocated header, so a scan for "coffee shop" in one ZIP code returns the pack you'd actually see standing there.

AI search visibility

This is the newest part of the index and the one changing fastest. AI answer engines don't publish rankings, so we have to prompt them, watch what they say, and record it.

  • 24.9 million+ prompts tracked across five AI engines
  • Engines covered: ChatGPT, Google AI Overview, Google AI Mode, Perplexity and Gemini
  • Every check captures the full answer text, which brands were mentioned, which sources were cited, and which outbound links were served
  • Monthly cadence with historical trend so mention rate, sentiment and citation share can be graphed over time

That's the data behind AI Visibility Checker, GEO Rank Tracker and AI Competitor Research — the three surfaces most customers use to answer the "am I showing up in ChatGPT?" question.

Site-audit and on-page data

The audit crawler is separate from the backlink crawler because the jobs are different: an audit needs to see everything on your site, render JavaScript, and check things like Core Web Vitals, whereas the link crawler is optimised for scale.

  • Renders JavaScript pages the way a modern search engine does
  • Checks 120+ on-page and technical rules per URL
  • Streams Core Web Vitals directly from Chrome UX Report
  • Diffs successive crawls so you see what changed, not just what's broken

How the data gets collected

All of this is collected by infrastructure we own and operate. SignalSumo is not a wrapper around a re-sold feed. That's a deliberate choice for three reasons:

  • Refresh control. When Google changes a SERP feature, we can update the parser the same week. Wrappers wait for their upstream vendor.
  • Cost control. Owning the crawl is how we can offer real AI-search tracking on a small-business plan instead of a $500 enterprise SKU.
  • New signals. The AI-answer index, the local grid scanner and the AI competitor research module all needed data types no vendor was selling — so we crawled them ourselves.

Getting the data into your stack

Every dataset above is available three ways:

  • The SignalSumo web app — dashboards, exports and one-click reports
  • The v1 REST API — five categories (keywords, domains, backlinks, AI search, site audit), with the same pricing you see in the app
  • The native MCP server — so ChatGPT, Claude and other LLM clients can query the same endpoints in natural language, no adapter code required

If you're building a dashboard, wiring an internal Slack bot, or letting your team ask questions in an AI client, all three paths hit the same numbers on this page.

What "our data" is not

Two honest disclaimers, because vendor comparison posts rarely include them:

  • No index is complete. 2.65 trillion backlinks is a large fraction of the visible web, not the whole thing. If a link lives on a page nothing else links to and no sitemap references, no crawler on earth will find it.
  • AI-search coverage moves. The five engines tracked today are the ones with meaningful query share. As new engines pass that bar (or drop below it), the list will change and this page will be updated.

Everything else on this page is a live, running number — not a marketing round-up.

Put the guide into practice

The fastest way to make advice stick is to run the checks, see how competitors handle it, and monitor for regressions.

Frequently Asked Questions

5.4 billion keywords, split across 184 country and language databases. Volume, difficulty, SERP features and click potential are recalculated independently per market — a keyword's difficulty in Canada is not just its US score with a flag on it.

2.65 trillion live backlinks across 411 million indexed domains. Our crawlers process about 5.88 billion pages every day, and 57% of the index is re-verified within 90 days so lost, redirected and toxic links surface fast.

Five, at present: ChatGPT, Google AI Overview, Google AI Mode, Perplexity and Gemini. For every prompt we log the answer, whether your brand was mentioned, which sources were cited and which outbound links were served.

No. SignalSumo runs its own SERP monitors, backlink crawlers and AI-prompt harvesters. That lets us control refresh cadence, add features like the AI-answer index without waiting on a vendor, and keep the price point sustainable.

Yes. A single REST API exposes keyword, domain, backlink, AI-search and site-audit endpoints. There's also a native MCP server, so ChatGPT, Claude and other LLM clients can query the same data conversationally without a custom integration.
About The Author
SignalSumo Team

We are the core analysis and optimization team at SignalSumo, dedicated to making SEO simple and actionable for everyone.