Rank tracking
How to track keyword rankings
Last reviewed 24 August 2026 · 9 steps
The short answer
Pick the 50–200 keywords that map to pages you can actually influence, then record each one daily against a fixed location, language and device from a clean, logged-out IP. Store the ranking URL alongside the position — a position that moves because Google swapped which of your pages ranks is a different problem from one that moves because a competitor overtook you.
Check a ranking free
Check Google's top 100 in the United States. Cached results may be available without an account; fresh checks require sign-up.
Setting up tracked keywords across engines and locations is what the 14-day trial adds.
The method, step by step
Choose keywords tied to a page you can change
A tracked keyword is only useful if you can act on it. Start from the queries that already bring impressions in Search Console at positions 5–30, because those are the ones where a content or title change moves revenue. Tracking 5,000 keywords you will never optimise produces a chart, not a decision.
Fix one location, language and device per keyword
Rankings are location- and device-specific, so "position 4" is meaningless without saying where and on what. Track the same keyword separately for the markets you sell in rather than averaging them. Mobile and desktop SERPs differ enough — different layouts, different local packs — that a mobile-first audience needs mobile as the tracked device.
Sample daily from a clean, logged-out IP
Checking your own rankings in a normal browser returns a personalised result shaped by your history, your location and your logged-in account. Use a logged-out incognito session with a VPN in the target country for spot checks, or a tracker that samples from clean IPs for anything you plan to report on.
Record the ranking URL, not just the number
Store which URL ranked for each keyword each day. When two of your pages take turns ranking, the position looks volatile while the real problem is keyword cannibalisation, and no amount of on-page work on either page fixes it until you consolidate.
Read movement against what changed on the SERP
A drop from 6 to 9 can mean you lost ground, or that Google inserted an AI Overview and a video carousel above the organic block. Log SERP features alongside position so you can separate "we got worse" from "the page got pushed down", which need opposite responses.
Review weekly, not hourly
Daily data with weekly review is the right cadence for most sites. Google runs continuous live tests, so single-day swings of two or three positions are usually noise. Compare seven-day averages and act on trends that persist past a week.
Automate it once you pass ten keywords
Manual checks stop being viable at roughly ten keywords, and they never produce history. Automating means a scheduled daily check per keyword, per location, per device, with the result stored — either a tracker that does it for you, or a script calling a SERP API on a cron. The scheduling is the easy part; storing the ranking URL and the SERP layout with every reading is what makes the history worth having.
Track over time by storing every reading, not the latest one
A ranking is only useful next to the readings before it. Keep one row per keyword per day — position, ranking URL, location, device, SERP features — so you can plot 30, 90 and 365-day trends and align drops with the day you changed a title or Google shipped an update. Overwriting yesterday leaves you with a number and no explanation.
Track several engines and several domains separately
Bing, YouTube, Amazon and AI answer engines each return a different result set, so a keyword tracked "everywhere" is really several keywords; keep one series per engine rather than blending them. The same applies to domains: track competitors and your own sites as separate projects so budgets, alerts and reports stay per-domain, then compare them in one share-of-voice view.
Honest limits
What this method will not tell you
- Rank tracking tells you position, not traffic. A number-three position on a query with an AI Overview and four ads can send fewer clicks than position six on a clean SERP — pair positions with Search Console clicks.
- Every tracker samples the SERP; none of them sees what each individual user saw. Treat a tracked position as one clean sample per day per location, not an absolute truth.
- Tracked positions cannot explain why something moved. That needs the ranking URL, the SERP layout and your own change log next to the number.
When positions fall
Diagnosing a ranking drop
Rank tracking shows that a position moved, not why. Before rewriting content, rule out technical causes on the site itself.
- Ongoing technical checks. LemWatch SEO monitoring (opens in a new tab) runs scheduled checks on titles, meta descriptions, robots.txt, sitemaps and broken links, so you can see whether something changed around the date a position moved.
- A one-off independent audit. LemAudit free site audit (opens in a new tab) scores a page across performance, technical, on-page, mobile and security checks and lists prioritised fixes.
LemWatch and LemAudit are separate Pearl Lemon products. Neither check proves why a ranking changed; treat their findings as evidence to compare against your rank history.
Questions
Frequently asked
How often should I check keyword rankings?
Collect daily and review weekly. Daily collection catches real drops early; weekly review stops you reacting to normal test-driven noise of two to three positions.
How many keywords should I track?
For most sites, 50–200 keywords that map to pages you actively work on. Add the head terms you want, the long-tail variants you already rank 5–30 for, and your brand terms.
Can I track keyword rankings in Google Search Console?
Partly. Search Console gives average position for queries you already get impressions on, which is real first-party data, but it averages across locations and devices and omits queries you do not yet rank for. Use it together with a tracker rather than instead of one.
Why do my rankings differ from what I see in my browser?
Your browser result is personalised by location, search history and your logged-in account. A tracker samples from a clean, logged-out IP in a fixed location, which is why the two disagree.
How do I track keyword rankings automatically?
Schedule a daily check per keyword against a fixed location and device and store each result. A rank tracker does this out of the box; the DIY version is a cron job calling a SERP API and writing position, ranking URL and date to a table.
How do I track keyword rankings over time?
Keep one stored reading per keyword per day rather than overwriting the latest position, then compare 7, 30 and 90-day averages. Annotate the dates you changed pages so you can attribute movement instead of guessing at it.
Can I track the same keyword across several search engines?
Yes, but track each engine as its own series. Google, Bing, YouTube, Amazon and AI answer engines return different results, so a blended "average position" across them describes nothing that exists.
How do I track rankings for multiple domains?
Set up one project per domain so keywords, limits, alerts and reports stay separate, then compare them in a single share-of-voice view. Tracking two domains inside one keyword list makes per-site reporting impossible later.
How often should keyword rankings be updated?
Daily is right for most sites. Hourly only pays off during a migration, a launch or a suspected penalty; weekly or monthly is enough for low-volatility, low-volume terms and costs far less.
Keep reading