Measurement
What changed after your change, and how sure Lastmod is.
Lastmod measures your SEO changes: it compares clicks per day before and after a change and says how confident the comparison is. When there isn't enough data, it says that instead. It never claims the change caused the difference.
Two periods, one comparison
Lastmod doesn't forecast. It sets the days after a change next to the days before, in the same unit. Where it can, it also sets that change against the change on comparable pages of your site that had no task in the same period.
Baseline
The 28 days before you mark the task done. Lastmod averages the clicks per day.
Marked done
You make the change and mark the task done. That day is left out of both periods.
Measurement window
The days after, as many as that kind of change needs: 21 for a title, 56 for internal links.
Settling
After the window closes, Lastmod waits 3 more days for Search Console data to arrive before it calculates anything.
Result
Average clicks per day after, against the average before, with how reliable the measurement is.
Marked done on 10 September 2026. Expect a result around 4 October 2026.
Clicks per day changed by +26% in the 21 days after you marked this task done, compared with the 28 days before that day.
HighResult confidence: High, based on 2,140 impressions.
Clicks per day
28 days before0.93 a day
21 days after1.17 a day
A change in clicks doesn't prove causation
Different changes, different windows
A title or meta description change is measured over a shorter window than a change to internal links, so each kind of change is measured for its own number of days. The result follows 3 days after the window closes.
| Change | Measured for | First result after done |
|---|---|---|
| Title | 21 days | day 24 |
| Meta description | 21 days | day 24 |
| Canonical | 28 days | day 31 |
| Internal link | 56 days | day 59 |
| Content updated | 56 days | day 59 |
| Redirect set up | 28 days | day 31 |
| Page performance | 35 days | day 38 |
- Site-wide tasks
- A task that covers the whole site has no single page to compare. Lastmod records the work but doesn't measure it.
- Page performance
- Lastmod measures the clicks, but it can't check that a speed change went live on your page.
- Starting values
- These windows are starting values and may change as real data comes in.
Enough data, or no number
Before Lastmod reports a figure, it checks that there is something to compare: at least 400 impressions in the 28 days before, and data for at least 70% of the days in both periods. If not, the result is Inconclusive and there is no percentage.
- Impressions are not the result
- They only decide whether a comparison is possible and how confident it is. The result is always clicks per day.
- Inconclusive is a result, not an error
- On a low-traffic site it will often be the honest answer, and Lastmod prefers it to a percentage built on too little data. These are the current starting thresholds.
Marked done on 10 September 2026.
Why there is no figure
There wasn't enough search data to say anything about the effect of this change.
Result confidence
Every result states how well the data supports the comparison. The band follows from how much traffic there was and how large the change was. It doesn't say whether the change is good.
- High
Measured difference backed by a lot of search data.
Clicks per day +26%, based on 2,100 impressions.
- Medium
Measured difference backed by a fair amount of search data; read it as an indication.
Clicks per day +14%, based on 900 impressions.
- Low
Measured difference backed by little search data; read it as a tentative indication.
Clicks per day +4%, based on 450 impressions.
- Inconclusive
Too little search data to point to a difference.
180 impressions in the baseline, below the minimum of 400: no figure.
Confidence is not causation
A High band means the data backs the measured difference well. It doesn't show that the change caused it.
When something else could explain it
Lastmod flags a result as confounded when another factor makes the comparison harder to interpret. The result is still shown, with the reason, and it is left out of the impact total.
- Another task on the same page
- Another change was made on the same page during the measurement period.
- The task was reopened
- This task was moved back from done during the measurement period.
- Clicks fell while rankings held
- Impressions and position held steady while the click-through rate dropped. A search result feature may be taking the clicks.
- The whole site moved
- The whole site moved in a similar way over the same period, so this page's change may not come from your edit.
- What Lastmod doesn't correct for
- Lastmod doesn't correct for a page's own seasonality, and it can't see everything else that moves clicks.
- The impact total
- It counts only High and Medium results that aren't confounded. Declines count too. Lastmod shows the total as a band, with the estimated sum of incremental clicks beside it.
Clicks per day changed by +14% in the 56 days after you marked this task done, compared with the 28 days before that day.
MediumResult confidence: Medium, based on 900 impressions.
Clicks per day
28 days before0.39 a day
56 days after0.45 a day
What a result means
A result describes what the numbers did between two periods. It doesn't explain why.
Example result
Clicks per day changed by +26% in the 21 days after you marked this task done, compared with the 28 days before that day.
A result tells you
- What changed between the two periods, in clicks per day.
- How confident Lastmod is in that comparison.
- Whether something else could explain the difference.
A result doesn't tell you
- That the change caused the difference. Seasonality, demand and other factors also move clicks. Comparable pages absorb part of that, not all of it.
- That rankings or traffic will follow.
- Anything a statistical test would. There are no p-values, and the comparable pages are not a randomised control group.
The thresholds and windows are starting values, not yet calibrated against customer data. On lower-traffic sites, “not enough data” will often be the honest answer.
See it measured on your own site
Start the 30-day free trial. No credit card needed.
30-day free trial · no credit card · database and hosting in the EU