SEO KPIs: Leading, Lagging and AI Visibility
SEO KPIs (key performance indicators) are the few metrics that show whether search is producing business results: typically organic conversions, non-branded organic clicks, impressions for the queries you target, and in 2026, how often AI assistants name and cite you. Everything else in an SEO report is a diagnostic that explains why those KPIs moved.
This guide is for marketing leads and owners who want a short, defensible SEO scorecard. It separates leading from lagging KPIs, explains exactly what the key Search Console and GA4 metrics measure, adds AI visibility KPIs, and ends with a scorecard you can copy. For the deeper method behind AI metrics, see the AI visibility metrics that matter.
On this page
- What makes a good SEO KPI?
- Leading vs lagging SEO KPIs
- Search Console metrics: what they actually measure
- GA4 metrics that matter for SEO
- Which AI visibility KPIs should you add?
- Technical health KPIs
- Which SEO KPIs should you stop reporting?
- An SEO KPI scorecard you can copy
What makes a good SEO KPI?
A good SEO KPI is tied to a business outcome, measured the same way every period, and something your SEO work can actually move. Average position across every query on your site fails the first test. Total organic sessions fails the third, because branded traffic mostly reflects everything else the business does.
Four tests to apply to every candidate KPI:
- Outcome link. Can you explain in one sentence how this number connects to revenue, leads or cost?
- Stable definition. Will it be measured the same way next quarter, from the same source?
- Controllable. Does SEO work change it, as opposed to seasonality or ad spend?
- Target-able. Can you set a direction and a rough target for it?
Most businesses end up with five to eight KPIs. More than that, and the report stops driving decisions.
Leading vs lagging SEO KPIs
Leading KPIs move early and predict results; lagging KPIs confirm the results later. SEO needs both, because the lagging ones (conversions, revenue) can take months to respond, and without leading indicators you cannot tell whether the work is on track. The chain below shows how they connect.
Scroll the diagram sideways to see all of it.
Leading KPIs are where you set short-term targets: new pages indexed, impressions for a target query group, non-branded clicks. Lagging KPIs are where you set annual targets: organic conversions and the revenue or pipeline they produce. If leading KPIs rise for three months and lagging ones do not, the problem is usually intent (you are ranking for the wrong queries) or conversion (the pages do not persuade), not SEO volume.
Search Console metrics: what they actually measure
Search Console gives four core metrics, and each has a precise definition that matters for how you read it. Google’s help page on the Performance report metrics defines them:
- Impressions: how often someone saw a link to your site on Google; depending on the result type, the link might need to be scrolled or expanded into view.
- Clicks: how often someone clicked a link from Google to your site.
- CTR: clicks divided by impressions.
- Average position: a relative ranking where 1 is the topmost position. When a site appears several times for one query, only its topmost position counts, and that value is averaged across queries.
Scroll the diagram sideways to see all of it.
Two consequences for KPIs. First, never use site-wide average position as a KPI; losing a batch of low-ranking queries makes it “improve.” Second, report clicks and impressions for defined query groups (for example, non-branded commercial queries for your core service), which you set up with regex filters in the Performance report.
AI features are counted in these numbers too. Google’s documentation on AI features in Search says traffic from AI Overviews and AI Mode is included in the Performance report under the Web search type. Separately, the generative AI performance report shows impressions specifically in AI Overviews and AI Mode, by page, country, device and date, though Google notes it is still rolling out to properties. The full walkthrough of the Search Console interface is in the Google Search Console guide.
GA4 metrics that matter for SEO
GA4 tells you what organic visitors do after they arrive, and three of its metrics belong in an SEO scorecard. Google’s help page on engagement metrics defines an engaged session as one that lasts longer than 10 seconds, has a key event, or has two or more page or screen views. Engagement rate is the percentage of sessions that were engaged, and bounce rate is its opposite.
For SEO KPIs, use:
- Organic key events (conversions), split into branded and non-branded landing pages where you can.
- Organic conversion rate, key events divided by organic sessions, per landing page group.
- Engagement rate for organic landing pages, as a diagnostic for intent match. A page that ranks but rarely engages is answering the wrong question.
Keep attribution settings fixed once you start reporting. GA4 lets you change the lookback window and attribution model, and changing either rewrites how much credit organic receives; the details are in SEO ROI.
Which AI visibility KPIs should you add?
Add two or three AI visibility KPIs, measured on a fixed prompt set, rather than trying to track every AI surface. When someone asks ChatGPT, Gemini, Perplexity or Google’s AI features a question in your category, the answer names a few brands and cites a few sources. Whether you are among them is now a real share of search visibility, and none of it shows up in classic rank tracking.
The KPIs we recommend adding:
| AI KPI | Definition | Source |
|---|---|---|
| Presence rate | Share of prompts in your set where your brand is named | Fixed prompt set, run monthly |
| Citation share | Share of answers that link to your pages | Same prompt set |
| Share of model | Your mentions relative to competitors’ across the set | Same prompt set |
| AI Overviews and AI Mode impressions | Impressions in Google’s AI features | Search Console generative AI report |
| AI referral conversions | Key events from AI assistant referrers | GA4 |
The first three depend on a well-designed prompt set: the questions your buyers actually ask, grouped by intent, run the same way every month, and scored consistently. How to build and size that set is covered in share of model and the AI visibility metrics that matter. Setting up the GA4 side is in tracking AI traffic in GA4 and Search Console.
Why citation share deserves a KPI slot: a Semrush study of 230,000 prompts and more than 100 million citations found AI engines lean heavily on a small set of domains, with Reddit and LinkedIn among the top five on ChatGPT, Google AI Mode and Perplexity. If your pages are not cited, someone else’s are, and that someone else is shaping how buyers see your category.
Technical health KPIs
Technical health belongs in the scorecard as a guardrail, not a goal. Two or three indicators are enough:
- Indexed pages versus pages you want indexed, from Search Console’s page indexing report. A falling ratio usually means a template or release problem.
- Core Web Vitals pass rate. Per web.dev, the “good” thresholds are LCP within 2.5 seconds, INP of 200 milliseconds or less and CLS of 0.1 or less, measured at the 75th percentile of page loads. Report the share of URLs passing, by template.
- Crawl errors on important templates, such as server errors or unexpected redirects.
Set alerts for these rather than reporting them monthly. A broken robots.txt should be caught in hours, not at the end of the month.
Which SEO KPIs should you stop reporting?
Stop reporting metrics as KPIs when they are easy to move without helping the business. The usual suspects:
- Total organic sessions. It mixes branded demand (driven by everything else you do) with the non-branded demand SEO actually wins. Split it or drop it.
- Number of keywords ranked. Tools count any position in the top 100, including irrelevant phrases. It rises when you publish anything.
- Domain Authority and similar third-party scores. Useful for comparing link profiles, but they are tool vendors’ estimates, not Google metrics, and nobody buys from you because your score went up.
- Bounce rate on its own. In GA4 it is simply the share of sessions that were not engaged. A contact page that answers the question in five seconds can “bounce” and still do its job.
- Pages published. Output, not outcome. It belongs in a work log.
- AI “mentions” with no fixed prompt set. A screenshot of one good answer is an anecdote. Without the same prompts run the same way each month, the number cannot trend.
Moving these into an appendix, or out of the report entirely, is often the single biggest improvement to an SEO report.
An SEO KPI scorecard you can copy
A good SEO scorecard fits on one page and answers three questions: is demand reaching us, is it converting, and are we visible where buyers now ask? This is the scorecard we suggest as a default, to be trimmed or extended for your business.
Scroll the diagram sideways to see all of it.
A few rules make the scorecard work in practice:
- Set a target and a direction for every KPI, even a rough one. A KPI with no target is just a metric.
- Show the same period last year next to every number, because most businesses are seasonal.
- Write one sentence per KPI explaining the change, and one decision for next month. The layout is in the SEO report template.
- Keep rankings in the appendix as diagnostics.
Choosing SEO KPIs that drive decisions
The right SEO KPIs are few, tied to money, measured the same way every month, and split into leading indicators you can act on quickly and lagging results you report to the business. In 2026 that set has to include AI visibility, because a growing share of discovery happens in answers that never produce a click.
If you want a scorecard like this set up for your site, including the AI prompt set and the tracking behind it, our AI SEO service includes it from the first month.
Frequently asked questions
What are the 5 main SEO KPIs?
For most businesses: non-branded organic clicks, organic conversions (leads or orders), organic conversion rate, impressions for target queries, and AI visibility on a fixed prompt set. Together they cover demand, traffic, business outcome and the AI answers that increasingly sit in front of search results.
What is the difference between an SEO metric and an SEO KPI?
A metric is anything you can measure, such as average position or crawl requests. A KPI is the small set of metrics tied to a business goal and reviewed on a schedule, with a target. Most SEO reports fail by showing dozens of metrics and no KPIs.
Is keyword ranking still a good SEO KPI?
Rankings are a useful diagnostic, not a good headline KPI. Positions vary by location, device and personalization, and a ranking for a phrase nobody searches means nothing. Report clicks and conversions from groups of queries instead, and use rankings to explain changes.
What are the four pillars of SEO?
They are usually described as technical SEO, on-page content, off-page authority such as links and mentions, and user experience. For KPIs, the useful point is that each pillar has leading indicators you can track long before revenue moves.
How do you measure AI visibility as an SEO KPI?
Run a fixed set of prompts your buyers would ask across ChatGPT, Gemini, Perplexity and Google's AI features on a schedule, and record whether your brand is named, whether your pages are cited, and whether the description is accurate. Add Search Console's generative AI impressions for Google's surfaces.
How often should SEO KPIs be reported?
Monthly for most KPIs, with a quarterly review against targets. Weekly checks are useful only for technical health alerts. SEO moves slowly, and weekly reporting of lagging KPIs mostly shows noise.