SEO KPIs: The Metrics Worth Tracking

TL;DR: The primary KPI for SEO is organic conversions, qualified leads, or revenue. Pair that lagging outcome with a small set of leading indicators from Google Search Console, especially rankings, impressions, and CTR. Treat indexed pages and Core Web Vitals as diagnostics, not goals. If a metric does not change a decision, remove it from the scorecard.

Most SEO KPI reports have the hierarchy backwards

I have seen dashboards lead with Domain Authority, total keyword count, pageviews, and a green arrow beside organic traffic. Conversions appear near the bottom, if they appear at all.

That is a report about activity, not value.

Ahrefs describes conversions from organic traffic as “arguably the only indisputable SEO KPI.” Its guide also warns that “more traffic won’t necessarily mean more revenue.” Both points are important: a purchase, qualified form submission, trial, or phone call is evidence that search produced a business outcome. A visit is only an opportunity to produce one.

You can grow organic traffic while making the economics worse. Publish dozens of pages for irrelevant high-volume queries and the traffic chart may look excellent. Content production, maintenance, and infrastructure still cost money. The visitors simply have no reason to buy.

I am confident about putting conversions at the top. I am much less confident about any company attributing every conversion perfectly to SEO (that uncertainty matters later). Imperfect attribution is not a reason to replace the outcome with an easier proxy.

Pair one lagging KPI with a few leading indicators

A useful SEO scorecard needs outcomes and early signals.

Lagging indicators describe what has already happened. Organic conversions, revenue, and total organic sessions belong in this group. They matter to the business, but they often move after the underlying search change.

Leading indicators move earlier. Rankings, impressions, indexing status, optimized pages, referring domains, and Core Web Vitals can change before conversions do. They are not proof of success. They tell you where to investigate or intervene.

We saw the distinction directly after migrating SEOJuice from seojuice.io to seojuice.com in January 2026. Search Console impressions and positions recovered weeks before organic conversions did. If Lida and I had watched conversions alone, we could have concluded that the migration was still failing. The leading indicators showed recovery was underway; the lagging outcome had not caught up yet.

The reverse mistake is just as common. A keyword moves from position 31 to 18, everyone celebrates, and the page continues generating no meaningful traffic. I have made that mistake too. Ranking movement feels productive because it appears quickly (revenue is less cooperative).

For most sites, my default scorecard contains:

• One primary outcome: organic conversions, qualified leads, or revenue.
• One volume metric: organic sessions or users.
• Three search indicators: target-query positions, impressions, and CTR.
• Two health checks: indexing exceptions and Core Web Vitals problems.

SEOJuice is a two-person team, so reporting time competes directly with product and customer work. That forces a useful rule: add a metric only when someone can name the decision it changes. “The client expects it” may explain a metric, but it does not make it a KPI.

The SEO metrics worth tracking

1. Organic conversions, qualified leads, and revenue

Start by defining the event that represents value: a paid order, activated trial, booked demo, qualified lead, phone call, or another action tied to the business model. Then measure how many of those events came through the Organic Search channel.

GA4 groups unpaid search visits under Organic Search in its default channel grouping. You can inspect the channel under Reports, Acquisition, and Traffic acquisition, then review key events and revenue by landing page.

Do not stop at one site-wide conversion total. Conversions by organic landing page reveal which pages attract useful demand and which only attract attention. Our guide to Google Analytics for SEO explains the GA4 setup in more detail.

The difficult part is not creating another event in analytics. It is choosing the right one. A free signup might be a strong buying signal for one product and a nearly meaningless action for another. When we moved SEOJuice’s entry pricing from $9 to $29, the commercial question mattered more than maximizing the cheapest possible signup count. Metrics have to follow the business model, not survive forever because somebody configured them two years ago.

For lead-generation sites, report qualified leads where possible. Twenty form submissions containing spam, job applications, and unsuitable enquiries should not beat eight leads that sales can actually pursue.

2. Organic traffic

Organic sessions or users measure the volume of unpaid search visits reaching the site. Track the trend, but never present it without conversions.

Traffic works well as a denominator. If sessions rise while conversions remain flat, investigate query intent, landing-page relevance, geography, device mix, and the conversion path. If traffic stays flat while conversions rise, the site may be attracting better visitors or converting existing demand more effectively. The chart suggests questions; it does not establish the cause.

Analyse traffic by landing page rather than relying on the site-wide total. A stable aggregate can conceal a commercial page losing 40 percent of its visits while several low-value articles compensate for the decline. The top-line graph looks calm. The revenue-producing section is not.

3. Rankings and average position

Rankings often move before traffic and conversions, which makes them useful leading indicators.

Google Search Console reports average position for queries and pages. According to Google Search Console Help, position is approximately where a result was seen relative to other results, with position one being topmost. A position is recorded only when the result receives an impression.

The word average does considerable work here. Position varies across queries, devices, locations, users, and search appearances. GSC is useful for an aggregated view based on actual impressions. A dedicated rank tracker is better for checking a fixed keyword from a defined location on a regular schedule.

Across sites we track on SEOJuice, pages that move into the top five almost always begin in positions 4–15. Below roughly position 20, title changes and internal-link fixes rarely create the jump on their own. That is not a universal law (I would distrust anyone claiming one), but it is a better prioritisation rule than working through pages alphabetically.

For a valuable query in position 8, inspect search intent, competing results, the title and snippet, content completeness, internal links, and cannibalisation. For a page in position 46, first ask whether Google considers it a plausible answer at all. The same “optimization checklist” should not be applied to both.

Avoid total keyword count. “We rank for 4,000 keywords” says nothing about relevance, position, demand, or conversions. Ten commercially relevant queries can be worth more than thousands of incidental terms sitting beyond page five.

4. Impressions, average position, and CTR

These Search Console metrics are substantially more useful when interpreted together.

Google defines an impression as how often someone saw a link to your site on Google. A click records how often someone clicked through, while CTR is clicks divided by impressions.

• Impressions rising, position improving, clicks rising: visibility is becoming traffic. Check whether conversions follow.
• Impressions rising, clicks flat: inspect position, title, snippet, search features, and intent. Weak CTR is not necessarily a title problem if the result is still near the bottom.
• Position improving, impressions flat: the query may have limited demand, or the page may need visibility across more relevant terms.
• Clicks falling while position is stable: check impressions and the SERP itself. Demand or result-page features may have changed.
• Impressions falling across many pages: segment by page, query, country, and device before editing anything.

This is where measurement becomes operational. A high-impression page around position eight with weak CTR creates a specific investigation queue. “Visibility increased by 6 percent” does not (though the green arrow is pleasant).

5. Organic visibility or share of voice

Share of voice estimates visibility across a tracked keyword portfolio, often relative to competitors. It is useful for monitoring a product category or market segment rather than diagnosing one page.

The output is only as credible as the keyword set. A portfolio filled with broad informational terms can show growth while commercial visibility declines. Keep the set relevant to the products, locations, or problems the business serves. Record major additions and removals too; otherwise a changed keyword set can masquerade as improved performance.

6. Indexed pages

An unindexed page cannot rank. Indexing is therefore an important health check, but a larger index is not inherently better.

Use Search Console’s Pages or Indexing report to verify that important pages are indexed. Investigate valuable URLs that are excluded, while allowing deliberate exclusions for duplicates, filters, staging URLs, and low-value CMS output.

The KPI is not “number of indexed pages.” The useful question is “what percentage of the pages we intend to rank are indexed?” That denominator prevents a growing archive of thin URLs from looking like progress.

7. Core Web Vitals

Core Web Vitals cover LCP, INP, and CLS. Google defines the “good” thresholds as LCP within 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1.

They are supporting diagnostics, not the headline KPI. Google Search Central states that “Core Web Vitals are used by our ranking systems,” but also says: “There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience.”

Fix serious performance problems, particularly on templates serving important organic landing pages. Do not promise a specific ranking or revenue increase merely because three red indicators became green. Relevance can outweigh imperfect page experience, and a technically fast page can still be the wrong answer.

Domain Authority is not a KPI for SEO

Domain Authority is a third-party score created by Moz. It is not a Google ranking factor and has no direct effect on rankings.

“We don’t use domain authority at all in our algorithms.”

That is Google Search Advocate John Mueller, as quoted by Search Engine Journal. The same source cites Moz’s own statement: “Domain Authority is not a Google ranking factor and has no effect on the SERPs.” Our Domain Authority guide explains the distinction in detail.

DA, Ahrefs’ Domain Rating, and Semrush’s Authority Score can still be relative yardsticks. Comparing sites through one provider’s consistent methodology may help with competitive review or link prospecting. Setting “increase DA from 30 to 40” as an SEO objective is different: it turns a vendor’s proxy into the goal.

Also de-emphasize raw pageviews, total ranking-keyword counts, and bounce rate as headline KPIs. GA4 emphasizes engagement rate, but neither engagement nor bounce is a business outcome. Use behavioral metrics to investigate pages, not to declare SEO successful.

Set up the free measurement stack

Google Search Console and GA4 divide the job at the search click.

GSC provides queries, impressions, clicks, CTR, average position, indexing status, and Core Web Vitals. GA4 provides organic sessions, landing-page behavior, key events, conversions, and revenue.

Link them through GA4 Admin, Product links, and Search Console links. This makes it easier to review search visibility beside landing-page outcomes, but it does not create a query-to-conversion join.

That boundary is permanent in the standard setup. GA4 has no organic keyword report. Search Console owns the query side; GA4 owns the post-click outcome side. If GA4’s model or privacy trade-offs do not suit your business, we have compared several GA4 alternatives. A different analytics product may be easier to operate, but it cannot reconstruct data Google does not expose.

SEO attribution is directional, not exact

A customer can discover a business through a non-branded organic query, leave, return directly two weeks later, and buy. A last-click view may credit the direct session even though search started the journey.

The original query is unavailable inside GA4, and SEO returns can arrive months after the work. Ahrefs’ SEO KPI guide describes SEO ROI as “incredibly difficult to measure properly.” That is the exact problem: ROI is valuable in principle, but the inputs contain attribution gaps and timing assumptions.

Trend outcomes over quarters, annotate migrations and major releases, and compare landing-page cohorts where practical. Use monthly changes for diagnosis, not causal theatre. A precise-looking percentage is still an estimate if nobody can defend the attribution model.

Turn the scorecard into an action queue

A KPI report should end with decisions: protect this page, rewrite that title, consolidate these URLs, add internal links to this near-ranking page, or investigate why this template is excluded from the index.

SEOJuice’s keyword tracking and SERP monitoring surface leading indicators such as rankings, impressions, CTR, and competitor visibility. Its role is not to replace GSC or GA4. Those tools remain the foundation for search data and business outcomes.

SEOJuice is the execution layer. It continuously applies on-site fixes including internal links, meta titles and descriptions, schema markup, and image alt text. The useful loop is simple: identify an opportunity from the indicators, choose the appropriate intervention, then watch whether the outcome follows.

If technical and on-page diagnostics keep falling to the bottom of the queue, run the free SEO audit. SEOJuice also has a free plan with no credit card required; paid plans start at $29 per month. Test it on a real site and judge it by the work completed, not by another dashboard count.

Frequently Asked Questions

What are the most important SEO KPIs?

The primary KPI is organic conversions, qualified leads, or revenue. Support it with organic traffic, target-query rankings, impressions, CTR, share of voice, indexing status, and Core Web Vitals. Keep the business outcome at the top and use the other metrics to explain why it is changing.

What is the difference between a leading and a lagging SEO indicator?

Lagging indicators measure outcomes after they occur, including conversions, revenue, and organic sessions. Leading indicators move earlier and guide action, including rankings, impressions, indexing status, and Core Web Vitals. Tracking only one group either reveals problems late or rewards movement that may never create value.

Is Domain Authority an SEO KPI worth tracking?

Not as a primary KPI or objective. Domain Authority is a third-party Moz score, not a Google ranking factor. Google’s John Mueller has said, “We don’t use domain authority at all in our algorithms.” Use DA only as a relative comparison produced by one provider, not as proof of ranking performance.

How do I measure SEO KPIs for free?

Use Google Search Console for queries, impressions, clicks, CTR, average position, indexing, and Core Web Vitals. Use GA4 for organic sessions, key events, and conversions by landing page. Linking them helps compare pre-click and post-click performance, but individual queries still cannot be joined to individual conversions.

Are keyword rankings a good SEO KPI?

Yes, as a leading indicator. Track positions for commercially relevant target queries, especially pages in positions 4–15 where a focused improvement may produce meaningful movement. Raw keyword counts are usually vanity metrics because they ignore relevance, position, demand, and conversion value.

Why is organic traffic difficult to attribute to revenue?

GA4 does not expose organic keywords, customer journeys often involve several channels, and the return from SEO work may arrive months later. Organic search may begin a journey that a later direct or branded visit receives credit for. Treat attribution as directional, compare longer trends, and avoid claiming more precision than the data supports.