A-TIER technique
Jev keyword triage — score SERP beatability before writing content
100k ft — What is this?
Jev keyword triage — score SERP beatability before writing content
50k ft — Why care?
Pull the top 10 results for 72 keywords and have TypeSafe's Jev (a classifier model that returns probabilities) answer literal questions per result: intent match, depth, freshness, site strength, beatable or not. Three or more weak results means target the keyword, then Claude writes. 720 results cost about 4 cents, search volume still needs a separate tool.
10k ft — Key takeaways
• Ask literal yes/no questions for clean classifier answers
• Manually review results the model was unsure about
• Pick keywords by how beatable page one is, not by volume
• Three or more weak top-10 results marks a winnable keyword
5k ft — Tools & techniques
Claude, Jev
LLM-as-classifier, SERP scoring
1k ft — Deep-dive / try it
Comment “JEV” for the full guide.
Everyone’s writing blog posts.
Almost nobody checks who they’re up against before they write.
So I pulled the top 10 search results for 72 keywords in my niche.
720 real results.
Then set Jev up to check which ones are actually beatable.
Here’s the setup.
Pick your keywords. Things your buyers actually search.
Pull the top 10 results for each one.
Ask Jev the same questions about every result.
Does it match what the searcher wants.
How deep is the content.
How fresh is it.
How strong is the site.
Could you beat it.
Thin pages, old posts, forum threads and off-topic results light up orange.
Three or more beatable results in a top 10 means that keyword is yours to take.
Jev is a new AI model from TypeSafe. It doesn’t write. It decides, and every answer…
Transcript excerpt
(no speech detected)