Record these six things
1. Page and query group
Save the exact URL and the search intent or query cluster that motivated the work. A page URL alone often loses the context.
2. Baseline window
Write down the dates and the clicks, impressions, CTR and position you will compare. Keep the same property, country, device and search type where possible.
3. Hypothesis
State why the edit might help: for example, the title is too generic for a relevant comparison query. Make it specific enough to be wrong.
4. Exact change
Save the old and new title, section, internal link or other edited element. Note unrelated template and technical changes too.
5. Publish date
Record when the change actually reached the public site, not only when it was drafted. Search engines may recrawl at a different time.
6. Review and decision
Set a follow-up date and record whether the evidence supports keeping, revising, reverting or waiting. “Inconclusive” is a valid conclusion.
Compare like with like
Choose date ranges that account for weekday patterns and seasonality. Compare the same Search Console property and search type, then inspect the page and query rows rather than relying on a sitewide average. When possible, consider a control page or unaffected query group, but remember that small sites rarely provide a perfect experiment.
Search Console data is delayed and aggregated. A single page may receive too few impressions to distinguish a real change from noise. Avoid calling a result a win solely because the average position moved for a short period; query mix and result features can change that average.
Separate observation from cause
Write “CTR increased during the review window” when that is all the data establishes. Do not automatically write “the new title caused CTR to increase.” Other pages may have changed, demand may have shifted, the search result may have changed, or Google may have tested a different query mix. Your log becomes more trustworthy when it records those caveats.
Use a simple change-log template
Page: exact URL
Query / intent: phrase or cluster
Baseline: dates, clicks, impressions, CTR, position
Hypothesis: one sentence
Change shipped: exact edit and public date
Other changes: template, links, promotion or technical work
Review date: planned checkpoint
Observation: what changed in comparable data
Conclusion: win, loss, mixed, or not enough evidence
For a first candidate, use the free Search Console opportunity finder to sort an exported Pages report locally. For the ongoing version, Rank keeps Search Console review, local page auditing, proposed changes and 7/14/28-day checkpoints in one Mac workspace. Those checkpoints are reminders; use a longer window when the page has not collected enough data.
When this workflow is useful
This approach suits indie founders, consultants and small teams that make occasional but consequential SEO updates. If you need large-scale crawling, backlink intelligence or portfolio-wide enterprise reporting, specialist platforms may fit better. Rank focuses on a narrower job: decide what to improve on pages you already own and preserve the learning from each attempt.
Rank is a $39 one-time Mac purchase for people who want their Search Console evidence, changes and follow-up in one repeatable workflow. See the product details.
Related: find low-CTR pages · run a weekly review · use the free SEO tools.
What counts as a change worth logging
- Title or meta description edits
- Rewrites of the first paragraph or main heading
- New internal links to a page
- URL changes and redirects
- Template changes affecting many pages
- Site moves, redesigns and platform changes
Small copy tweaks do not need an entry. Anything that might plausibly move search results does.
Questions people ask
How long should I wait before judging a change?
At least 14 days, and 28 for a fair view.
What if two changes overlap?
Note both and treat the result as shared. Next time, space them out.