How do you know if a keyword is worth targeting?

To know if a keyword is worth targeting, multiply four numbers: monthly search volume, the share of searchers who click your result, your conversion rate, and what one conversion is worth. Search volume on its own tells you almost nothing. A 200-search keyword often beats a 20,000-search one.

That is not how anybody picks topics, though. You open a keyword tool, you get a volume column and a difficulty score, and neither of them is money. So the biggest number wins, and three months later a page ranks beautifully for something that was never going to pay for the afternoon it took to write.

What is the formula for a keyword’s value?

A keyword is worth whatever its traffic converts to, which is four numbers multiplied together: monthly search volume, the share of those searchers who click your result, the share of those visitors who convert, and the value of one conversion. Everything else is commentary.

Work it once with real figures. Say a keyword gets 500 searches a month, you rank third and take a tenth of the clicks, which is 50 visits. Two of those fifty fill in your contact form. One lead in four turns into a job worth $400, so that keyword is worth around $200 a month to you.

Two hundred a month. That is the number that tells you whether six hours of writing and three months of waiting were worth spending, and no keyword research tool will ever hand it to you, because three of the four inputs are facts about your site rather than facts about Google. The SEO ROI Calculator runs the same chain across a set of keywords at once, showing the traffic, the conversions and the revenue at each step.

One thing to hold on to: what comes out is revenue, not profit. You still have to set it against what the work cost, whether that was a freelancer’s invoice or your own weekends.

Why is search volume the wrong number to judge a keyword on?

Because volume is the input everyone can see, and it has the weakest connection to money of the four. Compare two keywords. One gets 20,000 searches a month and is a definition query, where most people read the answer on the results page and never leave it. The other gets 200 searches, and every one of them comes from somebody with a broken boiler and a phone in their hand.

The small keyword usually wins. Two hundred searches, a third of them clicking a result that answers exactly what they typed, five in a hundred of those calling you: three or four jobs a month. The twenty-thousand keyword can send ten times the visits and still be worth less, because those visits wanted a sentence, not a plumber.

This is the part that stings when you have been picking topics off a volume column for a year. The column was never wrong. It was just answering a different question from the one you were asking.

How many of those searchers will actually click your result?

Fewer than the position-by-position tables suggest, and the gap keeps widening. Those tables are built from third-party click data averaged across every kind of query, which is to say they describe nobody in particular. Google has never published a click-through rate for any position, and the absence is not an oversight.

What has changed underneath them is the results page. According to a Pew Research Center analysis published in July 2025, people clicked a traditional search result on 8% of visits where an AI summary appeared, against 15% of visits where none did, and clicked a link inside the summary on just 1% of visits. That analysis tracked the browsing of 900 US adults across 68,879 Google queries during March 2025, and Pew published its method alongside the figures, which is more than most click data offers.

So the honest position is that nobody can hand you a reliable click share for your keyword. Query type, brand recognition, whether a summary fires, how many ads sit above you: each of those moves the number further than your rank does. Which leaves one click-through rate worth putting into the sum, and it is yours.

Where do you get your own click-through rate for free?

Google Search Console, in the Performance report, for any query your site already appears for. It gives you clicks, impressions, CTR and average position query by query, and the CTR there is simply clicks divided by impressions.

Two details are worth knowing before you lean on it. Google’s own documentation for the report defines average position as the topmost position your link held, averaged across every query it appeared in, so a query where you sat at 3 and again at 9 is recorded as 3. An impression is counted when your link appeared on the results page, which for some result types means it was never scrolled into view.

Take the rate from a query you already rank around where you hope to rank, in the same corner of your site, and use that. It is a rough proxy. It is also built on your snippets, your brand and the results page your readers are genuinely looking at, which is a claim no published table can make.

What conversion rate should you assume when you have no data?

None, ideally. If you run a form or a checkout, your analytics already knows the answer for the traffic you have, and that number beats any industry benchmark, because a benchmark is an average across businesses that have nothing to do with yours.

When you truly have nothing, do not guess once. Run the sum twice, at one percent and at four percent, then watch what happens to the decision rather than to the output. If the keyword is worth writing at one, write it. If it only pays at four, you are not making a keyword decision any more, you are betting on a conversion rate you have never measured.

That habit survives being wrong, which is the whole point of it. The Conversion Rate Calculator takes your current rate, your monthly traffic and your average order value and shows what a movement of a point or two does to revenue, which is the same sensitivity check pointed the other way.

What is one conversion actually worth to you?

For a shop, average order value, and you almost certainly know it already. For anything lead-based it is two numbers rather than one: the proportion of enquiries that turn into customers, and what a customer is worth once they do. A form fill is not a sale.

Get this wrong and nothing else in the chain matters. A plumber who counts every enquiry as a completed job overstates the value of every keyword on the site by a factor of three or four, and that error quietly compounds across a year of content decisions. If you sell a subscription, use what a customer is worth over their life with you, not what they pay in month one.

Does the keyword still deserve the effort if the maths works?

Not always. Look at what is ranking for it now. If page one is wall-to-wall large sites with hundreds of pages on the subject, or the whole question is answered by a summary above them, the click share you plugged into the sum is optimistic even at position one.

Then look at intent. A query phrased as a question generally wants an answer rather than a purchase, and it converts like it. That does not make it worthless. It makes it worth less, which is a different thing and an easy one to model: halve the conversion rate, run the sum again, see whether the keyword still earns its slot.

How do you turn this into a decision about what to write next?

Put your candidates in a list and give each one a monthly value from the four numbers, then sort by that value instead of by volume. Ten keywords is enough. The ordering usually looks nothing like the one your keyword tool produced, and the reshuffle is the entire benefit.

Then apply the effort side by hand. There are 3 things worth weighing against the money:

• How hard the page looks - A comparison piece you can write from experience in an afternoon is a different bet from a guide that needs original research.
• Whether you already rank near it - A query sitting at position 12 in Search Console needs a rewrite, not a new page, and it will pay back in weeks rather than months.
• What else the page can carry - A post that answers one high-value question often picks up a dozen related ones, and none of that shows up in the sum for the keyword you targeted.

Final Thoughts

Most keyword decisions get made on the largest number on the screen, and search volume has the least to say about whether the work pays for itself. Four multiplications is not much maths. It is enough to stop you spending three months ranking for something that was never worth ranking for.