How to Track SEO Improvements Over Time: A Simple System

To track SEO improvements over time, record a baseline before you change anything, follow a short list of metrics on a fixed schedule, and save your Search Console data before its 16-month window drops it. Log every change you make, and compare periods like-for-like. Fewer, steadier numbers will show progress more clearly than a crowded dashboard.

This guide covers the long-term system. To judge whether one particular fix worked, see how to know if SEO changes are working. To catch sudden breakages, see SEO monitoring.

What does tracking SEO improvements over time actually mean?

It means you can answer one question at any point: is the site healthier and more visible in search than it was 3, 6 or 12 months ago, and why? Answering it takes four things kept together: a baseline, a consistent set of metrics, stored history and a record of what changed.

Most teams have some of these but not all. They have traffic charts but no list of what they shipped, so they can't explain a rise. Or they have an audit from last spring but nothing to compare it with. Or they rely on Search Console alone, and then the data they wanted to compare against has expired. Tracking SEO over time is less about tools and more about habits: the same numbers, pulled the same way, on the same dates, with notes. When those habits are in place, a quarterly review takes an hour, and you can say with evidence which work paid off and which didn't. Without them, every review starts from scratch and opinions win over data.

Step 1: How do you set an SEO baseline?

Take a dated snapshot before your next round of fixes. Save it in a spreadsheet or folder you control, not just inside a tool. A useful baseline has three layers:

1. Search performance. Clicks, impressions, CTR and average position for the last 3 and 12 months, for the whole site and for your key page groups (for example, product pages, service pages, blog posts).
2. Technical health. Number of indexed and non-indexed pages by reason, Core Web Vitals status, crawl errors and the issue list from a full audit.
3. Business outcomes. Organic leads, sign-ups or sales from analytics, so SEO numbers can be tied to revenue.

Write down how each number was pulled: date range, filters, property type. If next quarter's export uses a different filter, the comparison means nothing.

Which SEO metrics should you track over time?

Track a small set of metrics that answer different questions, each from one consistent source. The table below is a practical starting point for most small and mid-size sites.

For Core Web Vitals, use Google's targets: LCP within 2.5 seconds, INP of 200 ms or less and CLS of 0.1 or less, measured at the 75th percentile of page loads on mobile and desktop separately.

Average position on its own misleads. It falls when you start ranking for many new, long-tail queries, which is good news. Always read it together with impressions and clicks.

How do you keep SEO data for longer than 16 months?

Export it regularly. According to Google, "Search Console keeps data for the last 16 months", and the Search Console reports linked into Google Analytics have the same limit. After that, the older months are gone, and so is your year-two baseline.

You have three options, from simplest to most robust:

• Monthly spreadsheet snapshot. On the first working day of each month, export last month's totals and top pages from the Performance report into one sheet. Ten minutes, and it never expires.
• Search Analytics API. Google's own guide to traffic drops suggests using the Search Analytics API or bulk data exports to store data beyond the standard interface.
• Bulk data export to BigQuery. Search Console can schedule a daily export of your performance data to BigQuery, which includes all performance data except anonymized queries. It suits sites with a developer or analyst who can query it.

Start the export now, even if you're not planning to analyze it yet. Data you didn't save can't be recovered later.

How do you record the changes you make?

Keep a change log. Every time something ships that could affect search (a redirect batch, a template change, new content, a migration, a robots.txt edit), write down the date, what changed, which URLs and who did it.

Search Console now supports this directly. Since November 2025 you can add custom annotations to Performance report charts, so site updates, campaigns, migrations and algorithm updates appear on the chart itself. Two limits matter for long-term tracking: annotations are deleted automatically after 500 days, and everyone with access to the property can see them. Keep a master copy of the log in your own spreadsheet.

Add external events to the log too: Google core updates, seasonal peaks, price changes, PR coverage and paid campaigns. Half of explaining a trend is ruling out causes that had nothing to do with SEO.

How often should you review SEO progress?

Use three rhythms, each with its own job:

• Weekly (5–10 minutes): look for breakages only, such as indexing drops, server errors or a sudden fall in clicks. This is monitoring, not progress tracking.
• Monthly (about an hour): update the metrics table, compare with the same month last year, and note anything unusual.
• Quarterly: run a full audit, compare issue lists with the previous one, and update the plan for the next 90 days.

Don't judge progress on weekly numbers. Weekly figures swing with weekdays, holidays and news, and most SEO fixes take weeks to show up in search results.

How do you compare periods fairly?

Compare like with like. Year-over-year comparisons (October against last October) remove most seasonality. Month-over-month comparisons are fine for spotting direction, but they mislead around holidays and sales seasons. Google's own debugging guide suggests viewing the last 16 months to check whether a change is a pattern that repeats every year.

A few rules keep comparisons honest:

• Segment before you conclude. A flat total can hide a growing blog and a shrinking product section. Compare page groups, countries and devices separately.
• Separate branded and non-branded queries. Brand searches rise with advertising and word of mouth. Non-branded growth is the clearer SEO signal.
• Check Google's data anomalies page. Search Console keeps a list of known data issues. One matters a lot for year-over-year work: "A logging error prevented Search Console from accurately reporting impressions from May 13, 2025 until April 27, 2026." Google says impressions, CTR and average position were affected but clicks were not. Until those dates pass out of your comparison window, use clicks for year-over-year trends. The same page notes that February 28 and March 1, 2026 are missing from bulk exports for some properties and won't be recovered.
• Use the same filters every time. The same property, the same search type, the same country filter.

How do you track whether audit issues are actually getting fixed?

Traffic shows the outcome. Issue tracking shows whether the work is getting done. Compare two audits issue by issue and put each finding into one of four groups: fixed, improved, unresolved or new. A single SEO score is not enough, because a score can stay flat while you fix ten critical problems and three minor new ones appear.

To make the comparison fair, run the second audit with the same scope and settings as the first, 30–60 days after the fixes, so Google has had time to recrawl. The before-and-after audit comparison guide covers the method in detail, and the sample comparison report shows what an issue-by-issue recheck looks like. To get a starting point for your own site, a free SEORecheck audit preview shows your SEO score and real issues found on your pages, covering indexing, technical health, Core Web Vitals and AI search access. You can then repeat it later as a recheck against that baseline.

Track the "new" group closely. New issues that keep appearing usually point to a process problem, such as a CMS template, a plugin or a publishing habit, rather than a one-off mistake.

How do you track AI search visibility over time?

For Google, use the Generative AI performance report in Search Console. It shows how many times links to your site were shown in AI Overviews and AI Mode, broken down by page, country, device and date, and Google has rolled it out to all websites worldwide. It reports impressions. The help page doesn't mention click data, so track AI visibility next to your regular clicks, not as a replacement for them.

If the report doesn't appear, Google says your site may not have enough impressions yet, or it may be excluded from Search generative AI features. Check that it is included. For other assistants such as ChatGPT or Perplexity, there's no equivalent Search Console report. Track referral sessions from those tools in your analytics, and re-check your AI crawler rules in robots.txt every quarter. Our guide to SEO vs GEO performance measurement covers which numbers exist for each platform.

Common mistakes when tracking SEO progress

• Tracking too many metrics. Twenty KPIs means none get a proper look. Seven is plenty.
• Relying on one score. Tool scores change when the tool's rules change, not only when your site does.
• No baseline. Without a dated snapshot, "we improved" is a guess.
• Changing many things at once with no log. Even when the trend goes up, you can't tell which change did it.
• Letting data expire. Starting exports after month 16 means losing the year you needed.
• Calling a result too early. Two weeks after a fix is usually too soon to judge.

Key takeaways

• Take a dated baseline of search performance, technical health and business outcomes before you fix anything.
• Follow about seven metrics from consistent sources, and read average position together with impressions and clicks.
• Export Search Console data monthly or set up bulk export, because the interface keeps only 16 months.
• Keep your own change log; Search Console annotations help, but they expire after 500 days.
• Compare year over year, segment by page group, and check Google's data anomalies page before drawing conclusions.
• Track audit issues as fixed, improved, unresolved and new, not as a single score.
• Measure Google AI feature impressions in the Generative AI performance report, alongside regular clicks.

Prices for audits and rechecks are listed on the pricing page.