What SEO Reports Should Explain Before They Show a Percentage

A percentage is a summary, not an explanation. Here’s how to judge whether an SEO report connects movement to cause, commercial meaning and action.

A dashboard says organic traffic is up 18%. Another says average position has improved by 2.4 places. A third shows qualified enquiries down 11%.

Which one tells you whether SEO is working?

None of them on its own. A percentage summarises a defined measurement. It does not explain what changed, why it changed, whether the movement matters to the business or what anyone should do next.

That distinction matters because SEO reports often bring together measurements from different systems and different stages of the customer journey. Search Console reports impressions, clicks, click-through rate and average position. Analytics may report sessions and key events. A CRM may hold the qualified enquiries the business actually cares about. These are useful pieces of evidence, but they are not interchangeable outcomes. Google’s Search Console documentation defines these search-performance metrics separately, while Google explains why Search Console and Analytics figures will not necessarily match.

A good report connects the number to a decision. A weak one decorates it with a green or red arrow and leaves the reader to invent the explanation.

A percentage is the starting point, not the conclusion

Imagine an illustrative monthly report for a software company. It shows:

• organic sessions up 22%;
• Search Console impressions up 31%;
• average position improved from 18.6 to 15.9;
• organic clicks up 14%; and
• qualified enquiries down 9%.

The headline might say: “SEO performance improved strongly this month, with traffic and rankings rising.”

That statement may be true as far as it goes. It is also incomplete. The report has not told us whether the additional visibility came from valuable commercial searches, whether the traffic reached the right pages, whether a change in search results affected clicks or whether the enquiry figure is being compared on a consistent basis.

The opposite would be just as problematic. A report saying “SEO has failed because enquiries are down 9%” is making a causal claim from one movement. The fall could relate to query mix, landing-page changes, sales capacity, tracking, attribution or demand. It may still be an important problem. The percentage alone cannot tell us which one.

A better question is:

What does this movement tell us, what does it not tell us and what decision should follow?

The context a percentage needs

A practical test for report quality is whether the explanation covers six elements:

1. Movement: what changed?
2. Scope: where did it change?
3. Comparison: compared with what?
4. Likely explanation: which causes fit the evidence?
5. Commercial significance: does it matter to the business?
6. Next action: what should happen now?

This is not a demand for a 40-page appendix every reporting period. A concise headline can work if it identifies the relevant population, comparison, uncertainty and decision.

1. Movement: what actually changed?

Start with the measure, not the verdict. “Visibility improved” could mean more impressions, a higher average position, more pages appearing in search or greater exposure for a small group of queries. Those are different movements.

Average position needs particular care. It is an aggregate approximation based on the topmost position occupied by a site or page across impressions, rather than a fixed rank that every user sees. Google describes how average position is calculated and advises looking at impressions and clicks alongside it, rather than treating position as the whole story.

A report might therefore say:

“Average position improved by 2.7 places, driven mainly by a wider set of informational queries. Clicks rose more slowly, while the commercial product pages in scope showed no material ranking improvement.”

That is more useful than “rankings up”. It tells the reader what kind of visibility changed and where the headline number may be misleading.

2. Scope: where did it change?

An aggregate percentage can hide very different movements across queries, pages, countries, devices and search appearances. This is a reporting inference from the dimensions available in Search Console: when a total combines several populations, it cannot show whether each population behaved in the same way. Search Console’s performance report documentation sets out these dimensions and metrics.

In the software-company example, the extra sessions might come from:

• brand searches for the company’s name;
• new blog pages answering broad educational questions;
• mobile traffic from a country where the company does not sell; and
• existing product pages receiving no additional qualified visits.

“Organic traffic up 22%” is therefore a true statement that may still be a poor description of commercial SEO performance.

Brand and non-brand groupings can help make this visible, as can page groups and country or device splits. They are useful working categories, though. A non-brand query is not automatically valuable, and a brand query is not automatically worthless. These labels do not directly measure intent or lead quality.

3. Comparison: compared with what?

Month-on-month comparisons are easy to produce and often easy to misread. A travel company reporting July against June may be measuring a seasonal shift as much as an SEO shift. A financial services site may see changes in search interest around tax deadlines. A software company may be affected by a product launch, a competitor campaign or a change in the wider market.

Google recommends using appropriate longer-period comparisons, including year-on-year comparisons where relevant, and considering wider search-interest evidence. That does not mean year-on-year is always the correct baseline. The comparison should fit the question.

Google Trends can help show directional changes in search interest, but it reports normalised relative interest rather than absolute search volume. Google’s explanation of Trends makes that limitation clear.

A report might therefore say:

“Sessions rose 22% month-on-month, but the same period last year also showed a similar increase in category searches. The current result looks positive, although the available evidence does not isolate the contribution of the latest SEO work.”

That is not talking down good performance. It is putting the performance in a fair frame.

4. Likely explanation: which causes fit the evidence?

SEO movements can reflect several different factors:

• Demand: more or fewer people are searching for the topic.
• Visibility: the site is appearing for more queries or in different positions.
• SERP behaviour: the search-results page has changed, affecting click behaviour.
• Website changes: a release, migration, template change or URL change altered the pages or links search engines can access.
• Measurement: tracking, attribution or processing changed what is recorded.
• Commercial context: the landing page, offer, sales process or lead qualification changed.

These categories organise an investigation. They do not provide six convenient excuses.

For example, a site release on 10 June is relevant if rankings changed on 12 June for the affected templates. It is not proof that the release caused the change. The affected URLs, timing and plausible mechanism still need to line up. Google notes that technical changes, indexing issues and migrations can affect search performance, sometimes with a delay while pages are crawled and reprocessed.

The same discipline applies to seasonality. “It is seasonal” is a hypothesis unless historical patterns, relevant search behaviour or category evidence support it.

5. Commercial significance: does the movement matter?

The software company in our example gained traffic but lost qualified enquiries. That does not automatically mean the traffic was useless. It does mean the report should investigate what the additional visits represented.

Perhaps broad educational content accounted for most of the growth. Perhaps brand searches increased because of an offline campaign. Perhaps product pages lost visibility while the blog gained it. Perhaps the enquiry form changed. Perhaps leads were stable but the CRM qualification process changed.

Clicks and sessions are evidence of visits or search interactions. They are not the same thing as commercial value. Attribution models allocate credit across touchpoints; they do not prove that organic search exclusively caused a conversion. Google’s Analytics documentation explains attribution as the allocation of credit across interactions.

Analytics conversion figures also depend on how a conversion is defined, which attribution model is used and what data is available. Some figures may be processed or modelled and can change as more information becomes available. Google documents these reporting and modelling considerations.

The practical point is not to dismiss conversions because attribution is imperfect. Consistent definitions can still support useful comparisons. The report should simply avoid presenting a conversion percentage as a simple, exclusive verdict on SEO.

6. Next action: what should happen now?

A report becomes useful when its interpretation changes a decision. That might mean investigating a page group, holding off on a major recommendation, validating tracking, reviewing a release or doing nothing until more evidence is available.

For our illustrative report, the next action could be:

“Review the 40 product and comparison pages separately from the blog. Compare brand and non-brand clicks with qualified enquiries for the same period, check whether the enquiry definition or form changed and review the June release against affected URLs. Do not treat the overall traffic increase as evidence of improved lead generation until this split is complete.”

That is more valuable than asking the team to “keep an eye on conversions”. It names the population, the comparisons and the checks that could change the conclusion.

Rankings, clicks and traffic need different explanations

It is tempting to treat the SEO journey as a neat sequence: better rankings produce more clicks, which produce more traffic, which produce more conversions. Sometimes that is broadly what happens. It is not a reporting law.

Search-result position affects click probability because people respond to where results appear as well as what they say. Research on position bias in web search supports that general principle, although its strength varies by query, device, result layout and user behaviour.

Search-result pages also contain different features and layouts. These can alter organic click behaviour beyond the traditional blue-link ranking. The evidence is directional rather than a permanent CTR rule: findings depend on the market, device, query and period studied. Recent research on SERP features illustrates why those changes need to be considered carefully.

A ranking improvement can be encouraging without proving that qualified demand will rise. A report should ask whether the improvement affected relevant queries, whether search demand was stable, what the result page looked like and whether the landing pages could convert the resulting visits.

If clicks fall while rankings appear stable, the right response is not necessarily to declare a ranking problem. Our earlier guide, When rankings hold but clicks fall, looks at how query mix and SERP behaviour can help explain that pattern.

Useful caveats narrow uncertainty

Context is valuable. It becomes evasive when it lists every possible explanation and avoids a conclusion.

Compare these two statements:

• Weak caveat: “Performance may have been affected by seasonality, competitors, SERP changes, tracking, the algorithm or wider market conditions.”
• Useful caveat: “Traffic rose mainly in brand queries, while non-brand product clicks were flat. Category search interest also increased during the period. We can support a positive visibility movement, but not yet a claim that SEO generated more qualified demand. We will review product-page enquiries after the next reporting period.”

The second statement does not pretend to know everything. It tells the reader what is known, what remains unresolved and what will happen next.

Some movements are simply too small, sparse or noisy to justify a confident conclusion. A 100% increase from one enquiry to two is mathematically correct and commercially easy to overstate. There is no universal percentage threshold that tells every business when to act. Absolute volume, variance, business value and the cost of acting all matter.

Correlation may be enough to justify a low-risk investigation. It is not automatically enough to claim causation or approve a high-risk change.

What a better report would say

Returning to the synthetic example, the first version said:

“SEO performance improved strongly. Traffic is up 22% and rankings are up.”

A stronger version might say:

“Organic sessions rose 22% month-on-month, with most of the increase coming from brand and informational queries. Search Console impressions rose 31%, while the average-position improvement was concentrated outside the core product-page set. Qualified enquiries fell 9%; this is not yet explained by the search data alone. We will separate product and content page groups, check the enquiry definition and review the recent release before attributing the commercial decline to SEO.”

It still contains the positive numbers. It simply refuses to make them carry more meaning than the evidence supports.

For a deeper look at the difference between more organic visits and more business value, see When organic traffic grows but leads do not. When a report follows a specific SEO change, the timing of evidence matters too: After an SEO change, when should you expect evidence?

The practical test for your next SEO report

Before accepting a green or red percentage, ask six questions:

1. What exactly moved?
2. Which pages, queries, markets or devices are included?
3. What is the fairest comparison?
4. Which explanations fit the evidence and which are still hypotheses?
5. Does the movement affect qualified commercial outcomes?
6. What decision or next check follows from it?

If the report answers those questions, it does not need to show every underlying row of data. If it cannot answer them, adding more charts probably will not help.

A good SEO report reduces uncertainty and supports a decision. It does not merely decorate a number with an arrow.

The discipline is simple: separate demand, visibility, search behaviour, website change, measurement and commercial context, then state what the evidence supports and what remains to be checked. The percentage still has a place. It just no longer has to carry the whole explanation.

Where the diagnosis spans several systems, page groups and implementation changes, specialist support can help connect the evidence and prioritise the work. Liquid Silver can help diagnose the movement, assess its commercial significance and work with the relevant teams on the next check or implementation step.