SEO Goals and KPIs
SEO goals and KPIs live on four levels: the business, marketing, search, and the team's week. Confusing the levels is behind almost every conversation that begins "we rank, so why no enquiries". This lesson covers how to phrase a goal, which metric belongs where, and how to read what a change in one means.
The measuring happens in the performance report and in analytics; the reading of changes follows Google's guide to debugging search traffic drops, which lays out an order of investigation that saves weeks of guessing.
SEO Goals Start With Business
SEO does not exist for rankings. It should help the business capture demand from search: sales, leads, registrations, demos, calls, store visits, subscriptions, or lower dependency on paid traffic.
| Level | Question | Example metric |
|---|---|---|
| Business | What should grow? | Revenue, leads, sales, organic share of acquisition. |
| Marketing | How will search support the funnel? | Organic sessions, conversions, CAC, comparison with PPC. |
| SEO | Which search levers are we moving? | Impressions, CTR, rankings, indexability, top-3/top-10 share. |
| Work | What does the team do weekly? | Fixed issues, published pages, updated snippets, new clusters. |
The levels must not be mixed in one report. The business asks about money, SEO answers with impressions, and both leave feeling the other missed the question. The SEO funnel and the search-to-result bridge connect them: they trace the path from impression to conversion on your own data.
What counts as a conversion and how to set one up is in the GA4 help on events and key events. Without that setup every KPI conversation ends at traffic, because there is nothing else to measure.
KPI vs Metrics
A KPI tells whether SEO is moving toward the goal. Not every SEO metric should be a primary KPI. Rankings, CTR, and impressions are useful diagnostics; businesses usually care about conversions and revenue.
| Metric type | Examples | Purpose |
|---|---|---|
| Outcome KPIs | SEO leads, sales, revenue, qualified leads, organic assisted conversions | Show business impact. |
| Intermediate KPIs | Non-brand traffic, CTR, impressions, top-3/top-10 share by cluster | Show whether visibility is growing. |
| Operational metrics | Indexable pages, Core Web Vitals, 404/5xx errors, publishing velocity | Show whether execution is happening. |
| Diagnostic metrics | Individual keyword rankings, crawl stats, cannibalization, duplicates | Help find the cause of a problem. |
A Good Goal Formula
A strong SEO goal answers five questions: what grows, by how much, by when, in which segment, and how it is measured.
Increase [metric] by [amount] by [deadline]
for [page/query segment]
measured in [GSC/GA4/CRM/BI].| Weak goal | Strong goal |
|---|---|
| Be at the top | Increase top-3 share for 40 commercial queries from 12% to 25% in 6 months. |
| Get more traffic | Increase non-brand organic sessions to /pricing/ by 35% by the end of Q3. |
| Improve the blog | Generate 300 additional organic leads from informational clusters over 2 quarters. |
| Do SEO | Reduce paid traffic share of lead acquisition from 70% to 55% through organic growth in 9 months. |
Leading and Lagging Indicators
SEO is delayed: work happens now, impact often appears weeks or months later. Use two kinds of indicators.
| Type | What it shows | Examples |
|---|---|---|
| Leading indicators | The work is moving in the right direction. | Pages indexed, CTR improved, snippets updated, clusters published, blockers fixed. |
| Lagging indicators | The business has received impact. | Leads, sales, revenue, lower CAC, higher organic share of pipeline. |
If you only watch lagging indicators, you discover problems too late. If you only watch leading indicators, you may celebrate work that never produces money.
Nobody knows in advance how far SEO lags: it depends on the competition, the state of the site and the amount of work. In its starter guide Google says plainly that changes do not show up immediately and may take months, and names no timeline. So a plan carries not a result date but a review date: when we look, and what should have moved by then.
How to read a change in a KPI
A KPI you cannot interpret is not a KPI. Traffic fell: is that an algorithm, a season, a breakage, or a reporting glitch? Google set out the order of investigation in its guide to debugging search traffic drops, and the same order works for growth.
| Step | What you do | What it rules out |
|---|---|---|
| 1 | Stretch the range to 16 months | Seasonality. If the dip repeats last year's in the same weeks, the cause is not you |
| 2 | Compare the period with the previous one and with the same period a year earlier | Random noise and the short-window effect |
| 3 | Look at search types separately: web, images, video, news | The case where one type dipped and you rebuild the whole site |
| 4 | Check Search Console's data anomalies page | A collection glitch on Google's side — it happens, and then there is nothing to fix |
| 5 | Only now look for a cause on your side: an algorithmic update, a technical issue, a manual action, changing interests, a site move | Testing hypotheses in random order |
The shape of the graph hints too: a sharp site-wide collapse is usually an update or a security problem, a slow slide is seasonality or fading interest, a step on one page template is a technical breakage. Wins and losses and the GSC export compare periods by page and query.
Brand and non-brand traffic deserve separating: brand grows with advertising and PR, and looked at together the search work appears better or worse than it is. The brand/non-brand split does that.
Goals Depend on Project Type
| Project | Main goal | Good KPIs |
|---|---|---|
| Ecommerce | Organic revenue and category sales | Revenue, transactions, category visibility, product page CTR, product rich results. |
| B2B SaaS | Qualified leads and demos | MQL/SQL from organic, demo requests, trial signups, non-brand BOFU/MOFU queries. |
| Local business | Calls, directions, bookings | GBP actions, local rankings, calls, form leads, reviews, "near me" visibility. |
| Media / blog | Audience and monetization | Non-brand traffic, returning users, subscriptions, ad revenue, Discover/News visibility. |
| Marketplace | Growth of indexable landing pages and demand capture | Template indexability, long-tail traffic, category conversions, cannibalization. |
Bad KPIs that break a strategy
- Position on one query. A single query does not describe a market, and the reported position is the topmost link to your site averaged across impressions: that is its definition. Making it the headline KPI means measuring something other than what a person sees.
- More traffic at any cost. Informational traffic may never buy. Which queries bring conversions is shown by the query-to-conversion view.
- The number of articles published. That is volume of work, not a result. Easy to test: how many of them got impressions — visible in top pages.
- Growth in DA or DR. Vendor metrics: Google publishes no such numbers. The glossary covers the term: Domain Authority.
- A perfect audit score. A hundred out of a hundred guarantees no demand if the intent, content and offer are not there. The glossary covers the terms: search intent and index coverage.
All five share a trait: you can improve them without getting closer to the goal. That is what makes a KPI bad — you can report progress while changing nothing.
Explaining SEO to Leadership
A strong explanation sounds like this: "We invest in pages and technical foundations that capture demand without paying for every click. In the first months we track indexation, impressions, CTR, and cluster growth. Then we measure leads, revenue, and reduced dependency on paid traffic."
Practice: write one goal all the way through
Five steps on your own project:
- Ask the client, or yourself: what should grow in money or enquiries? That is the business level, and "positions" is not an accepted answer here.
- Find the current value of an intermediate KPI in the performance report: non-brand clicks on the section in question over 28 days.
- Stretch the range to 16 months and check for seasonality. A goal set without that can "succeed" on its own or fail through no fault of yours.
- Write the goal to the formula: metric, value, deadline, segment, data source. If any element will not fill in, the goal is not ready.
- Set a review date and decide in advance what will count as "we are going the wrong way": which number should have moved, and by how much.
The glossary covers the terms: CTR, search intent, index coverage, Domain Authority.
How to tell the goal is set properly
The sign is simple: you can fail it. If the phrasing is such that in six months it cannot be declared missed, it is an intention, not a goal.
Three questions to check yourself. Does the goal name a metric, a number, a deadline and a segment? Do you know where that number comes from, and do you and the client use the same source? Have you decided in advance what happens if nothing moved by the review date?
What not to expect: that the business will accept an intermediate KPI as a result. Impressions and CTR are for your diagnosis; in a meeting they are the explanation, not the achievement. And the reverse — demanding revenue in month two is pointless, because search does not work that way.
What comes next
The next lesson is competitive analysis: how to see what competitors take demand with, and what of it applies to you. If you arrived here without the fundamentals, start with the SEO Foundations course — the robots.txt lesson, for instance, because no goal is reached on a site that is blocked from crawling.