Keyword Golden Ratio (KGR): How to Calculate It

What is the Keyword Golden Ratio

The Keyword Golden Ratio (KGR), coined by Doug Cunnington, is a simple math test that scores how competitive a long-tail keyword is. It works because Google needs a real reason to rank a new page, and phrases that almost nobody has targeted in their title tag give you that reason for free.

The formula is:

KGR = allintitle results ÷ monthly search volume

• allintitle results — how many pages Google has indexed with your exact keyword phrase inside their <title> tag.
• Monthly search volume — how many times the phrase is searched in an average month.

How to interpret the score

KGR	What it means
Below 0.25	Under-served. A new site can realistically hit page 1 within weeks.
0.25 – 1.00	Some intent match, but the SERP is more crowded. Winnable with authority.
Above 1.00	Well-covered topic. You'll need strong backlinks and depth to break in.

Volume cap matters. KGR only works on keywords with fewer than 250 monthly searches. Above that, Google's ranking factors shift toward domain authority and click signals, and the math stops predicting anything useful.

Step-by-step: calculating KGR by hand

1. Pick a long-tail keyword you're thinking of writing about.
2. In Google, run: allintitle:"your exact keyword phrase" (with quotes).
3. Note the result count Google shows.
4. Look up the phrase's monthly search volume in a keyword tool.
5. Divide: results ÷ volume.

Example: the phrase "best wireless earbuds for small ears" returns 42 allintitle results and gets 210 monthly searches. 42 ÷ 210 = 0.20 → strong KGR candidate.

Don't want to do the division yourself? Use the free Keyword Golden Ratio checker — paste the allintitle count and volume, and it returns the KGR plus its band (Golden / Okay / Hard) instantly, no signup.

Once a keyword passes KGR, model the upside with the CTR curve calculator to see the clicks each ranking position is worth.

How LemRank finds KGR keywords for you

Running allintitle searches by hand doesn't scale past a few queries. Google throttles the operator hard — usually 300 queries per day per IP — so KGR discovery at scale needs caching and budget-gating.

In LemRank, every keyword in your discovery workspace gets a KGR score computed automatically:

1. Volume is pulled from our keyword cache (12 sources, deduped).
2. allintitle is fetched on-demand via a rate-capped SERP probe (300/day, 30-day TTL).
3. Only keywords under the 250-volume ceiling get a score — everything else shows as N/A so you don't waste attention.
4. You can filter the keyword list by KGR < 0.25 to see only easy wins.

Common mistakes

• Using KGR on high-volume keywords. The formula was never designed for anything above 250/mo. Above that, the score is noise.
• Confusing intitle with allintitle. intitle requires only one word in the title; allintitle requires all of them. Use the strict one.
• Ignoring intent. A keyword can pass KGR and still convert nobody. Read the SERP before you commit — if the top pages don't match what you'd write, the keyword is wrong for you, not too hard.
• Chasing dozens of KGR wins at once. Publish 3–5 per week, then internal-link them so each supports the others.

What KGR does not do

• It doesn't guarantee page 1 — it stacks the odds.
• It doesn't measure commercial value — pair it with intent + CPC before committing.
• It doesn't replace on-page SEO — you still need to actually answer the query well.
• It doesn't apply to established sites with strong authority — those can target harder keywords directly.

Related

• How to do a keyword gap analysis
• How to find keyword cannibalization