v2.14.1 docs header polish: the docs-center search trigger is now a prominent search field right next to the brand — placeholder text, ⌘K hint, icon-only on narrow screens.
AnvilWiki
Powered by Pagefind
中文
Lesson 4 · Pick & Validate Updated Sep 2, 2026

Reading Google Trends: Five Minutes to Confirm Real Search

Validate each candidate in Google Trends compare mode against a baseline query: read level, slope, and cycle. Two-source rule: two unrelated places minimum.

A true scene first

Two games in the candidate table. The first: high YouTube view counts. The second: modest views, but comments full of “any new codes?”. Which one?

Probably the second — but don’t guess; verify with data. This lesson’s job: stamp every row of the candidate table with a rising / flat / falling label.

Step 1: open compare mode (1 minute)

Go to trends.google.com, hit Compare, enter:

  • your game codes
  • a baseline query whose scale you roughly know (no idea? use roblox codes — a perennial giant)

Step 2: read exactly three things (2 minutes)

WhatHow to judge
LevelYour line at what fraction of the baseline? Half or more = a big query; hugging the floor = niche — drop it or watch it longer
SlopeRising over the last 90 days = the golden window may be open now; flat = an old game with a settled competitive board; falling = fading, don’t catch the knife
CycleSwings with seasons/patches are a normal shape; a long slide with no bounce = demand is leaving

Best practice: screenshot the curve, feed it to AI with the three questions, and let it read the chart against the table above — but you place the label. The judgment is yours, not its.

Step 3: the two-source rule (2 minutes)

Trends says rising — still not enough. Verify once more in a second, unrelated place: a relevant YouTube video passing 50k views in 7 days, or Roblox concurrent players.

Two independent sources both hot → next lesson. Only one source hot → mark the row “watch”, and look again in a week. The most common single-source mirage: huge YouTube views, flat Trends — viewers watch and leave; they never search for guides, and that traffic can never become your visits.

Three classic mistakes (made for you in advance)

  • Slope without level: 0.1% → 0.3% is “rising”, but that scale cannot hold up a site. Level too low means “watch”, however steep the climb.
  • Deciding on one source: trusting Trends alone is this lesson’s most expensive mistake. The two-source rule has no exceptions.
  • Verifying an English market with Chinese-language heat: your site is in English; validate the words English players type (game codes), not the buzz in your own feed.

Three words to know (just these)

  • Baseline query: a query whose scale you know, used as the ruler for your candidate.
  • Two-source rule: popularity evidence from two unrelated places, so one platform’s fake heat can’t fool you.
  • Watch: the middle state in your table — not rejected, just “look again in a week”.

✅ Acceptance (all must hold)

  • Every candidate row carries a rising / flat / falling label with its evidence
  • Every “rising” has two independent sources behind it
  • ☐ You skipped no verification because something “felt hot”

Next lesson

Real searches confirmed — half a question left: can you still get onto page one? Next lesson dissects the Google results page in two minutes: open seats or red ocean. Go to Lesson 5 · Reading the Results Page