AI SEO In Detail › Module C · Demand and Keyword Research › Episode 22
Rank Keywords by CPC and Commercial Value
Use CPC as a value signal: combine volume, expected clicks and CPC into one keyword score with DataForSEO's Google Ads data, and learn where it misleads.
Playing the video loads it from YouTube (privacy-enhanced mode).
Keyword Value Score Sheet
A spreadsheet that scores keywords by volume x your own CTR curve x value per click, with each input in its own column, optional fields for your conversion rate and order value, high, medium and low tiers, and the shift from volume rank to value rank.
- Fill one column per input — Volume and CPC from the same Google Ads task, about nine cents for up to a thousand keywords. Expected position from Search Console, or a deliberately modest guess. CTR from your own curve per position.
- Read the score as a rough ad cost — Volume x expected CTR x value per click is roughly what the traffic would cost if you had to buy it as ads. If you know your conversion rate and order value, use those and keep CPC as a sanity check.
- Use tiers, not exact ranks — The score multiplies three estimates, so 412 versus 390 is not a real difference. Group keywords into high, medium and low, then combine the tier with intent and difficulty.
Free · Excel / Google Sheets / Numbers · no sign-up
979 words · the narration as spoken in the episode.
Intro
Two keywords from a fictional coffee equipment shop, with made-up numbers for illustration. One has almost ten thousand searches a month. The other has under six hundred. Which one should you go after first?
If you sort by volume, the answer is obvious. But advertisers pay almost ten times more per click for the small one. They know something the volume column doesn't tell you: what a visitor from that search is worth.
In this episode: how to use cost per click as a value signal, a simple formula that combines volume, expected clicks and value into one score, and where that score lies to you.
What CPC and competition actually measure
When you pull search volumes from the Google Ads endpoint, you get a few more fields for free, in the same task.
CPC is the average cost per click that advertisers have paid for ads on this keyword. Competition is how many advertisers bid on it, relative to other keywords: low, medium or high, plus an index from zero to a hundred. And you get a range of top-of-page bids.
Notice what these numbers are about. They're about the ad market. Competition here is not how hard it is to rank organically. It says nothing about who's in the organic results. People mix these up all the time.
But the ad market is useful precisely because it's a market. Advertisers keep paying a high CPC only if those clicks turn into money. So across many keywords, CPC is a decent proxy for commercial value.
The value score
Here's the formula I use as a starting point. Value score equals volume, times expected click-through rate, times value per click.
Volume is the monthly search volume, from one source, as we discussed in episode eighteen.
Expected click-through rate is the share of searchers you'd get at the position you realistically expect. Don't use an industry average. Use your own site's CTR curve per position, the one we built in episode eleven. If you expect position five, use your own CTR at position five.
Value per click is CPC as a proxy. If you know your own conversion rate and order value, even better: use those, and keep CPC as a sanity check.
The result is roughly what the traffic from this keyword would cost if you had to buy it as ads. That's an easy number to explain to a client or a boss.
Filling the columns in practice
Where does each number come from in practice?
Volume and CPC come from the same Google Ads task, so they're consistent with each other. One task, up to a thousand keywords, about nine cents.
Expected position is the hard one. For keywords where you already rank, use your current average position from Search Console, or one step better if you plan to improve the page. For keywords where you don't rank at all, make a deliberately modest guess, like the bottom half of page one. Later in the series we'll replace that guess with a difficulty estimate from the results page.
The CTR curve comes from your own Search Console data. If your site is new and you don't have enough data yet, use one cautious curve for everything, and mark the scores as provisional.
Keep each input in its own column, not hidden inside the formula. When someone asks why a keyword ranks high, you want to show which input drove it.
Worked example
Let's run it on five made-up keywords for our example shop.
Sorted by volume, the how-to guide is first and the dual boiler machine is near the bottom. Now apply the score. The guide has huge volume, but a low CPC, and the results page is crowded, so we expect a low position and a small click-through rate. Its score drops. The dual boiler machine has modest volume, but a high CPC and a winnable position. It moves to the top.
The order changes completely. And that's the point. Volume tells you where the people are. The value score tells you where the money is. And notice that nothing here needed new data. Volume and CPC came from one task, and the CTR curve from your own Search Console.
Where the score lies
Now the caveats, because this formula is easy to trust too much.
CPC measures advertisers, not your business. If you earn from display ads, an affiliate commission or a lead form, your value per click can be very different from what advertisers pay.
Zero CPC doesn't mean zero value. It often just means nobody bids. Informational keywords with no CPC can still build the audience, the links and the internal paths that make your money pages rank.
CPC can be inflated. Brand terms, competitor names and a few aggressive bidders can push it up without telling you much about organic value.
CPC depends on location and currency. Compare keywords within one market, never across.
And the score multiplies three estimates. Each one is rough. Multiplied, the error grows. A score of four hundred and twelve versus three hundred and ninety is not a real difference.
How to use it without fake precision
So I don't rank by the exact score. I group keywords into three tiers: high, medium and low value. Within a tier, they're equal. Then I combine the tier with other signals.
Intent from last episode: a transactional keyword with a high CPC is a strong signal. A high CPC on an informational keyword deserves a second look. And difficulty from the SERP module, because a valuable keyword you can't win is still worth nothing to you.
We'll put all of these together into one gap score in episode forty-three. For now, the takeaway: don't sort your keyword list by volume alone. Add a value column.
Wrap-up
Volume says where the searchers are, CPC hints at what they're worth, and tiers keep you honest. The value score spreadsheet is linked in the description. Next episode: generating keyword ideas and related terms with the DataForSEO API. See you there.