How to measure SEO performance?
Save rankings and clicks before an edit, then compare them afterward for the same page and query.
Previously: Does CTR affect SEO? Turn those clicks and impressions into a record you can use for decisions.
Read the change with its position
Compare equal 28-day periods for the same page and query.
Before adding the relevant internal link
What to noticeSave this report before the edit. A low CTR at position 18 gives you a ranking problem to investigate.
A later report shows a top-three position
What to noticeRead the better CTR alongside the better position. This example does not establish that one edit caused all the movement.
Check whether your page ranks for relevant queries, earns organic clicks and brings enquiries or sales. A site-wide traffic total can hide which topics are growing and which are losing clicks.
Start with one page. In Google Search Console, open the Performance report, filter by that page’s URL and select the query you want to check.
How search history feeds into later rankings
Ranking lets people see a page in search. Some of those people click. The model connects that history of rankings and clicks to information the search system can use when comparing pages later.
You can use the same history to choose what to work on. Find relevant queries where pages repeatedly rank in the top three or five and earn organic clicks. Those pages give you a starting point for new pages and internal links.
Read CTR alongside position. A page moving from position 20 to position three should have more opportunity to earn clicks. Comparing the percentages alone hides that important change.
Save the numbers before making a change
Record the page’s URL, target query, title and main heading. Save its impressions, clicks and average position, along with the date range, country and device filters.
When you change a link, save the linking page, the linked page, the anchor text and the date the edit went live. Keep the previous wording so you can restore it if needed.
If you estimate what the edit will achieve, write that down before you see the results. This lets you compare your expectation with what actually happened.
Compare equal-length periods
Use the same country, device and query filters before and after the edit. For example, compare 28 complete days before a change with 28 complete days after it. The page needs to have been live throughout the period you are checking.
Early data can help you spot indexing problems or the first clicks. Check whether recent dates are still incomplete before using them in a comparison. A page being a month old doesn’t mean you have a complete month of data.
Read the numbers together
| Measure | What it tells you | What to check alongside it |
|---|---|---|
| Impressions | How often your result appeared in search | Which queries showed it |
| Clicks | How many clicks your result received | Whether those visits brought enquiries or sales |
| CTR | Clicks divided by impressions | Average position and results-page layout |
| Average position | The average position reported for those impressions | The exact query, country and device |
| Conversions | Actions such as bookings, enquiries or purchases | Which pages and visits brought them |
A query missing from an export wasn’t reported in that file. Keep it labeled missing rather than filling in a zero you haven’t measured.
Use the results to choose the next edit
Which relevant queries brought more clicks? Which pages improved their average position? Which visits produced bookings or sales? Use those answers to choose the next page to publish or link to.
Search Console may also show queries you hadn’t planned for. Search those phrases and inspect the ranking pages before deciding whether to create another page.
Save rankings and clicks before an edit, then compare them afterward for the same page and query.