A cadence for desktop Google ranks: daily, weekly, or monthly
How often a small website should check google rankings is a budget question before it is a tooling question. A check that nobody reads is a cost. A check that is too rare hides a ranking-URL switch until the monthly report. The practical split for a small site is three clocks that do not have to match: how often the cabinet requests Google, how often a person looks, and how often a client or a founder sees a summary.
The google rank tracker in the SpeedyIndex cabinet records desktop Google only. It does not check Yandex, Bing, mobile results, or YouTube. The region is a country or a city. Depth is Top 10, Top 30, Top 50, or Top 100. Pages after the first are loaded and paid only until the site, or a competitor saved in the project, is found. Charts of movement appear after the second check, so a single run is a photograph, not a trend.
What actually changes between two checks
Read the row, not the average. Between one saved check and the next, five things can move, and they do not move together.
- The organic position of the site on desktop Google.
- The URL that is ranking. A different page of the same site can take the query without the position number looking dramatic.
- The change versus the previous check. One step up or down is ordinary. A gap that is still there on the following check is the one worth a look.
- Whether an AI Overview cites the site, and as which source. The tracker counts only the organic result as the position. A link inside the AI Overview is shown separately and is not a promise that the citation will stay.
- Who else is on the first page. The project holds up to 10 competitors, and a separate tab lists sites that show up for the project’s keywords.
Each saved row keeps the Google results page with the site highlighted, so a number can be checked by eye. That snapshot matters more than the integer when a layout change, not a “drop”, explains the new position.
Three cadences, and the site that fits each one
| Cadence | Fits | Look at | Leave alone |
|---|---|---|---|
| Daily, short list | Launch, migration, or the days after a confirmed Google ranking update | Money queries and the ranking URL | The long tail, and any one-day wobble |
| Weekly, full list | Most small sites and most client retainers | Position, ranking URL, Top 3, Top 10, visibility | Rewriting a page because it moved two places |
| Monthly | A stable brochure site with a short list and rare edits | The month-to-month line and new SERP competitors | Treating the month as a verdict on last week’s edit |
Weekly is the default because a small team does not rewrite pages every morning. Four checks in a month are enough for a line. Daily on the whole list multiplies the token spend about sevenfold, because the price is per keyword per run.
Monthly is enough when the site rarely changes and the queries are not local-service terms that shift with the map pack and the city. It is a poor fit in the first weeks after a redesign, because the second and third checks are what make the chart exist.
A small site, worked as a budget
Price is per keyword, and the keyword has to stay in the sentence. The full rate starts at 4 tokens per keyword for Top 10 depth: 4 tokens for each results page that is actually loaded, and 8 tokens when the first page includes an AI Overview. Until 2 December 2026 the starting rate is 2 tokens per keyword, and 4 tokens when that first page includes an AI Overview. Deeper pages are still only loaded until the site or a project competitor is found. From 4 tokens per keyword; until 2 December 2026 — 2 tokens. Tokens are spent only when a check is run. They do not expire. There is no monthly minimum and no required subscription. A subscription, if you want one, adds 10% tokens. New accounts get 100 tokens. Token packs start at $5.
The arithmetic the cabinet is built around: 100 keywords × 4 checks in a month = 1,600 tokens at the full Top 10 rate, and 800 tokens at the starting rate. That is a weekly pass over one country or city, with no AI Overview premium and with the site found on the first page. Daily on the same 100 keywords is on the order of 30 runs, so about 12,000 tokens at the full Top 10 rate, or 6,000 during the starting window. A small site rarely needs that. Put a daily schedule on the few queries that pay the bills, and leave the rest on a weekly schedule. Tags exist so those two groups do not have to share one button.
If some keywords miss the first page, Top 30 or Top 100 can load more pages and the bill rises only for the pages that were loaded. The window before a run shows the charge. Do not multiply “depth” by the full page count and call it the invoice: loading stops when the site or a competitor in the project is found.
Agencies: one cadence does not fit the roster
An agency account is several projects, not one giant list. Each client site is its own project, with its own country or city, its own tags, and its own schedule. Export the check when the client week needs a sheet. The same tokens pay for every project, and a month with no runs costs nothing, because nothing is submitted.
Match the schedule to the client, not to the tool’s maximum:
- A local service business with city-level queries: weekly, one city in the project, a second city only if that city is a real market. Two cities are two checks.
- A new site in the first month after pages are indexed: weekly, plus one extra run after a large template change. Daily only for the handful of brand and money queries.
- A stable content site that publishes twice a month: weekly on the queries tied to the new URLs, monthly on the archive.
- A link-building sprint: a check before the links are requested, and another after those URLs are known to be indexed. The row shows whether the position and the ranking URL changed. It does not prove that a link caused the move.
What a browser tab will not settle
Google states that results can differ by time, location, language, and device type, and that personalization is a further layer on top of that context. A signed-in search on a phone in another city is not the row in the report. The report is a desktop Google result for the country or city saved on the project. Compare the report with the saved results page, not with a colleague’s laptop. The official explanation of why two people see different results is on Google’s help page about differing search results.
Search Console’s average position mixes real impressions and devices, and it lags by a couple of days. Use it for demand. Use the rank check for a desktop snapshot you can reopen.
A week that is worth repeating
- Keep the keyword list to queries someone will act on. Tag brand, money, and local separately.
- Set the region once and do not change it between runs you want to compare.
- Set depth to Top 10 unless you have a reason to see past the first page. Deeper pages cost more only when they are loaded.
- Schedule the weekly run. Add a daily schedule only to the short tag.
- After the second run, read the chart, the ranking URL, and the AI Overview column. Ignore a one-step move that reverses next week.
- Export when a person needs the sheet. The export is the handoff, not a second product.
The per-keyword pricing, including the starting window through 2 December 2026, is written out in the per-keyword pricing guide.
FAQ
How many checks make a trend?
Two. The cabinet draws charts after the second check. A third check tells you whether the move held. One check cannot show a change, because there is no previous row.
Does a daily check rank the site higher?
No. A check reads Google. It does not submit the site, and it does not promise a better position or a citation in an AI Overview. Google decides what to show. Eligibility to be cited is not the same as being cited.
Should the long tail be daily?
No. Daily belongs on a short tag: brand queries, a launch, or the days around a confirmed ranking update. The rest of a small site’s list is a weekly job. Monthly is for a list that almost never changes.
What if the position and the browser disagree?
Check the saved results page in the row, the city on the project, and the device. The tracker is desktop Google for that region. A phone, another city, or a signed-in account can show a different page. The highlighted snapshot is the record of what was measured.