
The SEO KPIs that matter are non-brand organic clicks, organic conversions and organic revenue, backed by leading indicators such as indexation health and visibility for target topics. Vanity metrics like total impressions, raw rankings and authority scores should not be headline numbers. Choose three to five KPIs tied to how your business earns money.
Pew Research Center’s 2025 analysis found that Google users clicked a traditional search result in 8% of visits when an AI summary appeared, compared with 15% of visits when there was no AI summary. Visibility and clicks can now move apart, so a KPI set needs both.
Most SEO reports fail in the same way: they list everything the tools can export and leave the reader to decide what matters. A long list of numbers feels thorough, but it hides the two or three that decide whether the work is paying off. A good KPI set is short, honest about its limits and connected to the way the business actually makes money.
This guide gives you a framework for choosing KPIs, a table of sensible defaults by business type, and a list of metrics I would stop putting at the top of reports. If you want to turn the chosen KPIs into a live view afterwards, the guide to building an SEO dashboard in Looker Studio is the natural next step.
Leading and lagging indicators
SEO is slow, so you need two kinds of measure. Lagging indicators are the outcomes you care about: organic enquiries, sales and revenue. They are the proof, but they arrive weeks or months after the work. Leading indicators are the earlier signals that the outcome is becoming more likely: pages getting indexed, impressions growing for target topics, more queries entering the top ten.
You need both because each fails alone. Lagging numbers tell you too late. Leading numbers can look healthy while nothing reaches the till. When I review a measurement setup, I ask for one lagging KPI per business goal and one or two leading KPIs that plausibly feed it, and I write down the reasoning that links them.
The core KPIs, in order of importance
- Organic revenue or qualified leads. The outcome. Count the events that matter to the business, such as a booked appointment, a quote request or a purchase, not every page view.
- Non-brand organic clicks. The best single measure of whether SEO work is winning new audiences. Brand clicks mostly reflect awareness you already have.
- Organic conversion rate on key landing pages. Tells you whether the traffic you win is the right traffic. A rise in clicks with a flat conversion rate often means the targeting drifted.
- Visibility for priority topics. Impressions and average position for a defined group of queries, tracked as a group rather than keyword by keyword.
- Indexation health. The share of pages you want indexed that actually are, and how many are stuck in an exclusion status.
A note on conversions: in GA4, the measures you mark as important are called key events. Google’s help page on key events in GA4 explains that any collected event can be marked as a key event. Choose carefully. If every click counts as a key event, your conversion KPI means nothing.
Rather have an expert do this for you?
Hand it over and get back to running your business.
KPIs by type of business
The right set depends on how the business sells. These are starting points I would adjust after a conversation, not a fixed template.
| Business type | Lagging KPI | Leading KPIs | Watch out for |
|---|---|---|---|
| Local service business | Calls, form enquiries, booked jobs from organic | Non-brand clicks for service and location pages; indexation of location pages | Phone calls missing from analytics |
| Online shop | Organic revenue and orders | Non-brand clicks to category pages; indexed product share | Brand search inflating results |
| B2B or lead generation | Qualified leads and pipeline from organic | Visibility for problem-led topics; conversions on guides | Long sales cycles delaying proof |
| Publisher or content site | Ad or subscription revenue per organic visit | Clicks and impressions by topic cluster; indexed article share | Zero-click answers lowering clicks |
| SaaS | Trials or sign-ups from organic | Visibility for comparison and use-case queries | Attribution of trial to later purchase |
Here is an illustrative example, not a real client. Imagine a dental clinic in Karachi whose monthly report leads with total impressions. Impressions rise for three months, so the report looks excellent, yet booked appointments from organic stay flat. Looking closer, most of the growth comes from informational queries far from the clinic’s services and from people outside the city. Replacing the headline with non-brand clicks to treatment pages and booked appointments from organic would have shown the problem in the first month, and pointed the next piece of work at local service pages instead of more general articles.
Indexation health as a leading KPI
A page that is not indexed cannot rank, so indexation is a precondition for everything else. I track two simple figures: the number of pages in the index that I intended to be there, and the number sitting in exclusion statuses that deserve attention. Two statuses are worth knowing by name. Crawled, currently not indexed usually points to a quality or duplication judgement, while discovered, currently not indexed is about crawl scheduling and priority. They need different responses, so a KPI that lumps them together is not much use.
Do not chase a total indexed count upward as a goal. Pushing thin or duplicate pages into the index can harm quality signals. The healthier KPI is the share of your valuable pages that are indexed, with the unwanted ones kept out deliberately. If you suspect the second problem, read about index bloat.
Where AI visibility fits
More searches now pass through AI summaries and chat assistants, and many businesses ask whether they appear in them. It is reasonable to track, but I would be careful about how you present it. There is no official, complete report that tells you how often an assistant mentions you.
A practical approach is a fixed set of questions your customers genuinely ask, checked on a regular schedule, with a simple log: was the brand mentioned, was it cited as a source, was the description accurate. Treat the result as directional evidence and label it as a manual sample. It belongs in the KPI set as a secondary, experimental line, not as a headline number that you report with false precision.
Vanity metrics to demote
These are not useless. They are just unfit for the top of a report, because they can rise without the business gaining anything.
- Total impressions. They rise when you appear for irrelevant queries or when Google changes how results are displayed. Useful as context, misleading as a goal.
- Average position as a single number. It blends unrelated queries and pages. Track position for a defined group of priority queries instead.
- Keyword count. Ranking for more keywords is not the same as reaching more customers.
- Authority or domain scores. These are third-party estimates, not Google metrics, and they say little about whether your pages answer anyone’s question.
- Total sessions. A rise in unqualified traffic looks good and pays nothing.
- Backlink totals. Counts ignore relevance and quality. If you do review links, the Search Console links report is a better place to look for patterns than a headline count.
Set targets without promising outcomes
A KPI without a target is a number, and a target based on a guess is a trap. I set targets in three steps. First, record a baseline for at least three months so seasonality shows. Second, set a range rather than a single figure, for example a band of growth the team agrees is realistic. Third, review the range at a fixed interval and change it only with a written reason.
Be wary of anyone who promises a ranking or a traffic figure by a date. Search results depend on competitors, algorithm changes and your own site’s constraints. A better commitment is to the work and the review process: these fixes will be done by this date, and we will look at these indicators on this schedule. When you want to link KPIs to money, the next step is working out the return, which I cover in how to calculate SEO ROI.
Report KPIs so people act on them
Present each KPI with three things: the current value, the comparison with the previous period or the same period last year, and one sentence on what you think is driving the change and what you will do. That sentence is the part clients and bosses value most, and it is the part automated reports usually leave out.
For a broader view of how measurement fits into ongoing SEO, see how to track SEO performance. A proper SEO audit also checks whether the tracking behind your KPIs is correct, and measurement setup is one of the areas covered in what a professional SEO audit includes.
Questions people ask about SEO KPIs
How many SEO KPIs should I track?
Three to five headline KPIs is enough for most businesses. Add secondary metrics for diagnosis, but keep them off the front page. If everyone in the room can name the KPIs from memory, the set is the right size.
What is the best KPI for SEO?
There is no single best one, but non-brand organic clicks combined with organic conversions come closest. The first shows whether you are reaching new people, and the second shows whether those people matter to the business.
Should rankings still be a KPI?
As a supporting metric for a defined set of priority queries, yes. As the headline, no. Rankings vary by location, device and personalisation, and a position gain does not guarantee more clicks.
How long before SEO KPIs show results?
It depends on the site, the competition and the scale of the changes, so I avoid giving a fixed timeline. Leading indicators such as indexation and impressions usually move before conversions do, which is exactly why they are worth tracking.
If you are not sure your current KPIs reflect the business, or the tracking behind them can be trusted, send a message through the contact page and I will review your setup as part of a free audit.
More in this series (SEO measurement and Search Console): start with What a Professional SEO Audit Includes (and Red Flags), then explore:
- Google Search Console Performance Report Explained
- Crawled Currently Not Indexed: Causes and Fixes
- Discovered Currently Not Indexed: Meaning and Fixes
- How to Use the URL Inspection Tool in Search Console
- Search Console Regex Filters: 12 Practical Examples
- SEO Dashboard Looker Studio: How to Build One in 7 Steps
Request received
Thank you! Saad will contact you within the next 24 hours.
Keep an eye on your phone and inbox.
