← Blog

How to estimate traffic gains with a CTR curve

The most common question in SEO planning is this: what do we gain if we reach 3rd for this keyword? A plan whose answer is "more traffic" does not get budget. A plan whose answer is "roughly 140 clicks a month, upper bound" can be compared with other work.

The way to produce that number is a click-through rate curve: a table showing how many clicks each position gets on average. This post covers the formula, where to get the curve and where the estimate misleads.

The formula

Estimated extra clicks = monthly search volume × (CTR at target position − CTR at current position).

An example. Say you derived rates from your own data and found 10 percent for 3rd and 3 percent for 7th. For a keyword searched 2,000 times a month, moving from 7th to 3rd means 2,000 × (0.10 − 0.03) = 140 extra clicks. The rates in this example are assumptions chosen to show the method.

The formula is good for ranking candidates against each other. A high-volume keyword at 9th can promise more than a low-volume keyword at 4th, and the formula shows that right away.

A general curve: the Backlinko study

If you do not have your own data, a published study is a starting point. One of the most cited is Backlinko's Google CTR study by Brian Dean. It is based on data from multiple Search Console accounts provided by Semrush and covers about 4 million search results.

The findings stated explicitly in the text:

  • The #1 result on Google has an average click-through rate of 27.6 percent.
  • The #1 result gets about 10 times more clicks than the #10 result.
  • In the study's words, moving up one position increases absolute CTR by an average of 2.8%.
  • Only 0.63 percent of searchers click a result on the second page.

A general curve is not your market

These rates are an average across many sites. In your industry the results page may look different: four ads on top, a map pack, a video carousel. The same 3rd place sits in the top half of the screen for one query and is invisible without scrolling twice for another.

So treat the general curve as a starting point and replace it with your own at the first chance.

Build your own curve from Search Console

Export the last three months of query data from the Search Console Performance report. Then follow these steps:

  • Separate branded queries. Queries containing your brand name get very high click-through rates and distort the curve.
  • Split queries into buckets by average position: 1, 2, 3, 4-5, 6-10 and so on.
  • In each bucket, divide total clicks by total impressions. Do not take a simple average of query rates; low-impression queries swing the average.
  • Do not use a bucket with few impressions. A rate from a few hundred impressions is noise.

Test your curve against the past

Before trusting the curve, try it on a few keywords that changed position in the past. Find a keyword that moved from 8th to 4th last quarter. Compare the gain the formula predicts for it with the click difference actually visible in Search Console between the two periods. If the estimate is systematically high across five or ten keywords, your curve is optimistic; scale results down accordingly.

This test is affected by seasonality. Search demand itself may have risen or fallen between the two periods, so look at click-through rate rather than clicks.

What breaks the estimate

AI Overviews change how clicks are distributed. According to a Pew Research Center analysis published in July 2025, among the searches of 900 US adults in March 2025, users clicked a traditional result in 8 percent of visits to pages with an AI summary, versus 15 percent on pages without one. Google's Search Console help page also states that an AI Overview occupies a single position in results and all links inside it share that position. If your query shows an AI Overview, a general curve will likely overstate the gain.

Ads, shopping boxes and the related questions section have the same effect. Volume is also an estimate: the monthly volume tools report is often rounded or a range, and for a seasonal keyword the yearly average does not reflect the campaign month.

How to write the result

Do not present the estimate as an exact number. "Roughly 140 clicks a month, upper bound" is more honest than "140 clicks". Write the curve and the volume source next to the number.

Do not write 0 as the gain for a keyword with unknown volume. Zero means no gain. Unknown volume means the gain is unknown too; it goes into the table as an empty cell.

To choose which keywords go into this calculation, see the previous post: Keywords close to the top 3: which ones to push?

With Adversee

Adversee's SEO module uses this formula to calculate the estimated monthly click gain of reaching 3rd for keywords close to the top 3, with rates taken from the Backlinko study. The result is always shown with an estimate label and a method note. For a keyword with unknown volume no gain is calculated; the field stays empty.

Start tracking competitor moves automatically.

How to estimate traffic gains with a CTR curve · Adversee