How the data works
Where the numbers come from, what average position means, how fresh the data is and how much history you get.
Every number in Serplyze comes from the Google Search Console API for the property of your project. Serplyze does not scrape Google results pages and does not simulate searches. This page explains what the numbers mean and where their limits are.
Clicks, impressions, CTR and position
| Metric | Meaning |
|---|---|
| Clicks | Clicks from Google web search results to your site, as counted by Search Console. |
| Impressions | How often a result of your site was shown, as counted by Search Console. |
| CTR | Clicks divided by impressions for the same rows and days. |
| Average position | The position Search Console reports for your site, averaged over the searches where it appeared. |
Serplyze reads the Web search type. Each tracked keyword is requested by its exact query text, so a keyword is complete however large the site is.
Average position is not a scraped rank
A scraping tracker checks one results page from one location at one moment. Search Console reports the average position real searchers saw, across every location and device. So a keyword can show 4.6 instead of a whole number, and it has a position only on days when your site was actually shown.
Whenever positions are combined (several days, several keywords, a group), Serplyze weights them by impressions:
average position = Σ(position × impressions) ÷ Σ(impressions)A plain average of daily positions would give a day with two impressions the same weight as a day with two thousand.
There is no exact position for a single place and moment, no search volume, and no data for sites other than your own verified property.
No impressions, N/A and rare queries
- No impressions: Search Console reported nothing for the keyword on those days. It is not the same as not ranking: the page may rank where nobody scrolled.
- N/A: Search Console has never reported the keyword in the stored history. It is checked again on every daily sync.
- checking…: the keyword was added a moment ago and its history is still loading.
Google hides very rare queries for privacy (anonymised queries). They are missing from the Search Console API, so a keyword that is searched very rarely can show no data even when the page ranks.
Data delay and provisional days
Search Console dates are days in Pacific Time. Each sync reads data up to the previous day and reads the last four days again, because Google keeps revising recent days.
The newest days are provisional: Search Console is still collecting them, so they under-report. Search Console names the first incomplete day and Serplyze treats everything from that day on as provisional (at most five days). When Search Console gives no such day, the last two days are provisional.
- Charts draw provisional days and label them Provisional.
- KPI cards, changes, comparison periods, movers and the keywords table stop at the last final day.
- Alerts and alert rules use final days only.
How much history you get
| Data | History |
|---|---|
| Site totals (whole property, by device) | Up to 16 months on every plan, as far back as Search Console holds data. |
| Keyword-level data on the Free plan | 30 days |
| Keyword-level data on paid plans | 3 months |
| Keyword-level data with the history add-on | 16 months |
A keyword you add later gets the same history as your other keywords; loading the older days takes a few minutes. After an upgrade the extra history is loaded from Search Console in the background.
After a downgrade, keyword days outside the new plan's window are hidden at once and deleted 14 days later. Site totals are not affected.
Daily sync
Every project syncs once a day. Settings → Project shows the time of the daily sync and the recent syncs, and owners and admins can start one with Sync now. After each sync Serplyze checks alerts, looks for rising queries and refreshes the suggestions.
For each tracked keyword a sync stores clicks, impressions and position per day for all devices together and for desktop, mobile and tablet, plus the pages the keyword ranked with. The rank tracker offers All devices, Mobile and Desktop.
A failed sync is retried automatically. Only lost Google access needs you: see Google access.
Why numbers can differ from Search Console
- Country filter. If the project has a country, every number is filtered to it. Compare with the same country filter in Search Console.
- Search type. Serplyze shows Web search. Image, video, news and Discover data are not included.
- Provisional days. Totals and changes leave the newest days out, while Search Console's interface may include them.
- Rank in the keywords table. The Rank column is the keyword's position at the end of the selected period, not the average of the whole period. See Visibility and rank spread.
- Query text. Keywords are matched to Search Console queries in lower case with single spaces, as Search Console reports them.
- Site totals versus keywords. Site totals include anonymised queries; the sum of your tracked keywords never will.
- Brand split. Branded numbers use your own brand terms, not Google's branded-queries filter.
Read next
- Visibility and rank spreadThe exact definitions behind the visibility score, top-10 share, rank spread, group average position and change values.
- Google accessWhich Google permission Serplyze asks for, what it can and cannot do with it, and how to reconnect or revoke it.
- KeywordsAdd keywords from Search Console, a pasted list or a CSV file, understand the keyword limit, and work with the rank tracker table.
- Plans and billingHow pricing works per tracked keyword, what Free and paid plans include, and how to buy, change, pay for and cancel a subscription.