An SEO KPI is a metric tied to a business outcome. Most numbers in a monthly report are not KPIs. They are diagnostics: useful for explaining why something moved, useless as evidence that the work is paying for itself. The distinction decides whether a report informs a decision or decorates one.
Both kinds belong in a report. The problem starts when a diagnostic is presented as proof, because diagnostics can improve while the outcome that matters does not move at all.
A KPI is a metric tied to an outcome, the rest are diagnostics
A number qualifies as a KPI when it passes three tests. Failing any one of them does not make a metric worthless. It makes it a diagnostic, which is a different job: diagnostics explain movement, KPIs justify spending.
- It connects to something the business counts: inquiries, bookings, calls, orders, or revenue.
- It can move because of the work: a metric nobody can influence is a weather report, not a target.
- It survives a direct question: if the number doubled, could anyone say what the business gained?
The test is simple to apply once the metrics in front of a reader are named.
What each metric in a report actually measures
Most SEO reports draw on two sources: Google Search Console, which reports what happened in search results, and an analytics tool, which reports what happened on the site. Google defines the four Search Console figures precisely (source: Google Search Console Help).
| Metric | What it measures | Where the detail is set out |
|---|---|---|
| Impressions | How often someone saw a link to the site on Google | what the core metrics mean |
| Clicks | How often someone clicked a link from Google to the site | The metric definitions guide, linked above |
| Click-through rate | Clicks divided by impressions | The metric definitions guide, linked above |
| Average position | A relative ranking of the link’s position, averaged across queries | how rank tracking works |
| Organic sessions | Visits that arrived from unpaid search | reading Search Console data |
| Conversions | Inquiries, calls, or form submissions from those visits | tracking leads, calls and forms |
| Return on spend | Revenue or pipeline attributable to organic search | calculating return on SEO spend |
Only the last two are KPIs under the test above. The rest are diagnostics, and one of them is misread more often than all the others combined.
Average position is an average, not a rank
Average position is the single most misinterpreted number in SEO reporting. It is not the position a page holds. Google defines it as the topmost position occupied by a link to the property, averaged across all queries in which the property appeared (source: Google Search Console Help).
Google’s own worked example makes the consequence clear. A site appears for one query at positions 2, 4 and 6, which counts as position 2. It appears for a second query at positions 3, 5 and 9, counting as position 3. The reported average position is 2.5, and no page ever held position 2.5 for anything.
- It moves when the query mix changes: ranking for new, harder terms pulls the average down while visibility grows.
- It hides the distribution: an average of 8 can mean every query sits at 8, or that a handful sit at 2 and the rest at 20.
- It rewards narrowness: a site ranking for fewer terms can show a better average than one ranking for many.
A falling average position alongside rising clicks is usually a site succeeding, not failing. The same misreading applies to when a metric is expected to move.
Leading and lagging indicators
SEO metrics move in sequence, not together. Early metrics respond first because they sit closest to the change that was made, and later metrics respond last because they depend on everything upstream. Reading them in the wrong order produces both false alarms and false confidence.
| Indicator | When it moves | What it licenses |
|---|---|---|
| Pages indexed | Days to weeks | Confirms the work exists and is visible to Google. Proves nothing about demand. |
| Impressions | Weeks | Shows pages entering results for queries. Early evidence of relevance. |
| Clicks and CTR | Weeks to months | Shows the listing is chosen once seen. First sign of real interest. |
| Organic sessions | Months | Aggregate traffic. Still upstream of any business outcome. |
| Conversions | Months | The first genuine KPI. Traffic is reaching people who act. |
| Revenue and return | Months to a year | The lagging proof. What the engagement is ultimately judged on. |
Impressions rising in month two is the expected result, not a strong one. Conversions flat in month twelve is a problem regardless of how good the upstream numbers look. Across nine Singapore engagements, six-month growth in organic traffic ranged from 1.3× to 5×, while multi-year engagements ranged from 13× to 40×, which is why an early judgment tends to be the wrong one.
How traffic becomes a business outcome
The metrics form a chain, and each link in it is a place where progress can stall without the number above it showing any sign. Reading the chain in order is what turns a report into a diagnosis.
- Impressions to clicks: if impressions climb and clicks do not, the pages appear but the titles and descriptions are not persuading anyone to choose them.
- Clicks to sessions: a gap here is usually a tracking fault rather than a real one.
- Sessions to conversions: traffic arriving and not acting points at intent mismatch or a page that does not ask for anything.
- Conversions to revenue: inquiries that never close point at lead quality, which is a targeting problem rather than a visibility one.
Each break has a different fix, which is why a report showing only the top or only the bottom of the chain cannot direct anything. What a full report should contain is set out in what a good SEO report contains.
The numbers that look like progress and are not
Certain metrics appear in reports because they reliably move upward, not because their movement means anything. They are not fabrications. They are diagnostics promoted to proof, and the promotion is what makes them misleading.
| Metric | Why it looks like progress | What it hides |
|---|---|---|
| Keyword rankings alone | Positions visibly improve | Whether anyone searches those terms, or clicks the result |
| Impressions alone | The number grows steadily | Whether the queries are relevant or the listing is ever chosen |
| Total keywords ranked | The count only goes up | That most additions are irrelevant terms at low positions |
| Domain authority scores | A single tidy number that rises | That it is a third-party estimate, not a figure Google publishes or uses |
| Traffic without intent | Sessions increase | Whether visitors wanted what the business sells |
| Links acquired | A countable monthly deliverable | Whether the sources are relevant, and whether rankings moved |
The related question, which benchmark each metric should hit, has no honest general answer. A click-through rate that is strong for a branded query is weak for an informational one, and a conversion rate that is healthy for a specialist service would be poor for a low-value product. Published benchmark tables rarely state the category, the intent mix, or the competing set they were drawn from, which makes them decoration rather than a target. The only defensible comparison is a site’s own prior period, and the competitors it actually competes against. Judging the trend honestly is covered in judging whether SEO is working.
None of this holds if the underlying measurement is wrong.
What has to be true before any of it is trustworthy
Measurement problems are easy to mistake for performance problems, and the two call for opposite responses. A report built on broken tracking will show a decline that never happened, or hide a gain that did, and no amount of additional work will correct a number that was never counted properly.
- Conversion tracking fires correctly: forms, calls and bookings are all counted, and counted once.
- Organic traffic is properly separated: paid, referral and direct visits are not mixed into the organic figure.
- The property covers the whole site: subdomains and alternate versions are not reporting separately or not at all.
- Internal visits are excluded: staff and agency traffic do not inflate the numbers.
- The comparison period is stated: month against month, or against the same month last year, and not switched between reports.
How often these figures should be reviewed, and in what format, is covered in how reporting cadence is set.
Frequently asked questions
How long before SEO metrics show anything?
Indexing and impressions can move within weeks, because they sit closest to the work. Conversions and revenue take months, because they depend on everything upstream. Across nine Singapore engagements, six-month growth in organic traffic ranged from 1.3× to 5×, while multi-year engagements ranged from 13× to 40×.
Are keyword rankings still a useful KPI?
Rankings are a useful diagnostic and a poor KPI. They explain why traffic moved, which matters, but a position is not an outcome and Google states that no one can guarantee one. Rankings belong in the working layer of a report, beneath the numbers the business actually counts.
Why does Search Console show different numbers from Analytics?
The two tools measure different events. Search Console records what happened in search results, counting impressions and clicks at Google. An analytics tool records what happened on the site, counting sessions after a page loads. Clicks that never become sessions, blocked tracking, and different time zones all widen the gap, so the figures are not expected to match.
What should a monthly SEO report contain?
At minimum: the business outcomes for the period, the diagnostics that explain them, what was done, and what happens next. A report that lists activity without connecting it to the chain above describes effort rather than progress.
Written by the AuthorityLighthouse editorial team, reviewed by Sunny Tay, Editor. Updated September 2026. Primary sources: Google Search Console Help; Google Search Central.
This article shares general guidance on search visibility for Singapore businesses, drawn from AuthorityLighthouse’s client work. Results shown are from real, named client engagements; every business starts from a different baseline, and no outcome is guaranteed.
