One domain had 2.7Γ more backlinks and still ranked lower.
We compared 5 domain pairs. Backlink volume alone didn't explain who ranked higher.
So what did? π
We left the last two blank. They depend on your team, setup, and how often interfaces, sessions, or answer formats change. ($44β49 is at list prices.)
Also check each platform's rules on automated access.
π¬ How do you collect AI answers today: API, in-house, or both?
Track which sites Google's AI Overview cites for your keywords, on autopilot.
Free @n8n_io workflow:
β runs on a schedule (weekly by default)
β pulls the AI Overview via DataForSEO SERP API
β extracts each cited source: title, URL, domain
β saves it to Google Sheets
π
Prefer a BI dashboard? Swap Google Sheets for it. You'll need an n8n instance and a DataForSEO account.
Template: l.dataforseo.com/4hVBf4W
Which keywords would you track first?
A competitor outranks you. Don't jump to backlinks.
Check in this order:
1.Technical: can Google index your page?
2.Intent: does their page match the search?
3.SERP: what sits above organic?
4.Keywords: what do they rank for that you don't?
5.Links: only now
More below β
Links go last because they're the most expensive to fix.
Check the cheap things first.
For step 5, our free template finds sites that link to your competitors but not to you: l.dataforseo.com/4jNkFqz
Which of the 5 goes wrong most often for you?
What's your brand's AI visibility?
We measured one brand across 146 AI answers:
β named in the answer: 58%
β site cited as a source: 36%
β share of all cited sources: 5.7%
Same brand. Same answers. Three different numbersπ
Ranking in Google's Top 10 doesn't mean you're in the AI answer.
We compared Google's Top 10 with the AI answer for 20 real searches:
β on average, ~1 in 3 Top 10 domains showed up
β Perplexity: ~1 in 2
β ChatGPT: ~1 in 6
Our website developer built a 19-module SEO app in his spare time.
β Python scripts
β Chrome extension, 47 checks
β macOS app: 19 modules, 109 on-page checks
12 modules run on DataForSEO APIs, 7 are fully local.
Why? He wanted to pay only for the checks he runs π
3 lessons from the build:
- Architecture first. AI fills in the rest, you verify it
- Cache everything. Never pay for the same query twice
- Keep expensive data off by default
Full story, APIs, and costs: l.dataforseo.com/4yDemdS
Ask ChatGPT and Perplexity the same question.
How many of their source sites overlap?
Half? A third? Less?
We tested 60 questions. Drop your guess, answer below π
New free course: GEO Analytics & Automation π
Track your brand in AI Overviews, AI Mode, ChatGPT and Gemini. Then make the report update itself.
7 modules Β· ~2h Β· no code
Start here, templates for Make, n8n and MCP included: l.dataforseo.com/46X3vPO
Stop hunting for the right endpoint.
Tell Codex what you need. The DataForSEO plugin picks the endpoint, pulls live data and returns a structured answer β‘
Same account. Pay-as-you-go. No API keys in chat.
Install guide, no MCP config needed: l.dataforseo.com/4dbwXVJ
Around Google's August spam update, #1 mostly held. #10 turned over.
352 keywords, before vs after:
β 97.7% had a Top 10 domain change
β 77.6% kept the same #1
β #10 went to a domain new to the Top 10 in 50.6%
So where did the new #1s come from? π

















