Before making changes, save your current phrases, groups, and the date of the latest measurement in your own spreadsheet. After updating the list, compare the old and new results for the queries that stayed the same. Show new services separately.
For example, you removed a pump model that is no longer for sale and added a new one. The average position for the entire list may change because of this replacement. It won't tell you what happened to the pages for the other products until you open their rows.
Record the reason for each change
Remove discontinued services, add current offerings, and clarify ambiguous phrases. Record the reason for each change. “Replaced the old model with a new one” is clear; “updated the semantics” does not let you determine which observations now relate to a different product.
Don't delete a query just because it shows an unfavorable position. That hides the problem instead of solving it. If the phrase remains important to the business, keep it and assign a clear check for the service page. Changes to the list should reflect changes in what needs to be tracked, not a desire to improve the average figure.
If you're still exploring a new area, keep exploratory phrases separate. Their absence from the old set doesn't mean the old site wasn't visible for them. It means those queries weren't tracked in that series before.
Save the old list
Before editing, record all phrases, groups, region, search engines, device, and date. For important queries, save the expected service pages. If the list was agreed with a client, note that version in the working document and let them know about the planned changes.
Don't keep the only copy in an employee's memory. A few weeks later, it will be difficult to remember which rows were removed and why. The external log should let you reconstruct the conditions without relying on verbal explanations from the person who configured the settings.
Also save the date of the latest available measurement for the old set. The edit date and the result date may differ. The measured data points—not the moment you opened the dashboard—make up the observation history.
Distinguish between adding, replacing, and changing frequency
Adding a new phrase expands the set, removing one reduces it, and changing its wording creates a different query. Moving a phrase between groups changes how often it is tracked. Changing the region changes the search conditions; a manual search on a phone also uses different conditions. Don't combine these operations under the single word “update” when comparing results.
| Example operation | What to save | How to read the next result |
|---|---|---|
| Added a service | The new phrase and service page | The start of tracking for this area |
| Removed a discontinued product | The reason for removal and previous data | A summary with a changed composition |
| Added a city to the wording | The old and new phrases | Two different queries |
| Moved a phrase to a daily group | The date and previous frequency | A new frequency of data points |
| Changed the project region | The boundary of the conditions | A separate comparison series |
Compare queries that appear in both lists
To assess the previous pages, select the queries that remain unchanged under the same conditions. Treat these as the overlap between the two versions. Show new rows separately until repeat measurements are available with the same settings.
The average for the old and new sets may differ because the phrases have changed. Don't attribute the entire difference to site quality. If you added many new areas with no positions found, that changes what the summary means even if the previous queries remained stable.
You can find a practical approach to selecting phrases in the keyword query guide. For an internal meeting, it's useful to label both versions and show which part of the comparison is shared.
Check the new schedule
The project allows up to 50 daily queries, up to another 150 weekly queries, and up to another 800 monthly queries. Yandex is tracked for these groups; Google is tracked for the daily group. Moving a phrase to a less frequent group may affect the availability of regular Google measurements in the current setup.
Choose the frequency based on the task. A new, important service is useful to discuss often; you can track an exploratory area less frequently. Don't fill every slot with similar phrases just to increase the volume.
After saving, check the actual settings and the date of the next available result. A planned schedule doesn't prove that a specific collection has completed. If there is no data yet, leave that status as it is instead of entering a hypothetical position.
Explain the change in the report
In the summary, state what changed and from which measurement the new version applies. Name the old and new number of rows, the areas added, and the conditions for comparing the overlap. You don't need a long account of every minor edit if the main boundaries are clear.
If the client is comparing the new report with an old image, show the list version alongside it. This makes it possible to explain the difference without excuses and without attributing nonexistent changes to the site. Mark the new group as a new observation.
For changes to geographic settings, include a separate note about the region. The Yandex help page. A settings change alone doesn't prove that previous results improved.
Read more: Website semantic core: how to build and use it for SEO.
Check results in the site monitoring dashboard.