← All articles
AI-era SEO · 29 April 2026 · 9 min read

What to Measure Beyond Search Rankings

Rankings are still useful. They are just no longer enough. That sentence annoys two groups of people. The first group wants to declare rankings dead, because declaring things dead…

Analyst''s desk with several distinct evidence sources arranged around one real enquiry, not a single chart

Rankings are still useful.

They are just no longer enough.

That sentence annoys two groups of people.

The first group wants to declare rankings dead, because declaring things dead is a reliable way to look serious for six minutes.

The second group wants SEO to stay inside the old reporting frame: keywords up, keywords down, traffic up, traffic down, meeting over. Nice and tidy. Occasionally useful. Increasingly incomplete.

Search has become messier than that.

AI overviews, richer result pages, zero-click answers, local packs, citations, snippets, comparison journeys, and branded follow-up searches all change how visibility behaves. A page can still rank and receive fewer clicks. A business can be cited without winning the same click it used to win. Traffic can fall while enquiry quality improves. A bland ranking report can make everything look worse or better than it is.

That is not a reason to panic.

It is a reason to measure more carefully.

Rankings still have a job

We have not stopped looking at rankings.

That would be performative.

Rankings still tell us whether a site is visible for searches that matter. They help show whether a page is moving in the right direction, whether a competitor has overtaken it, and whether a technical or content change has affected organic visibility.

For commercial searches, rankings still matter a lot.

If someone searches for a service, product, location, or problem with clear buying intent, the position of the site still affects whether it gets considered. Pretending otherwise is silly. A buyer cannot enquire from a page they never find.

But rankings are now the start of the conversation, not the whole report.

The old shortcut was:

"We rank well, therefore SEO is working."

The better version is:

"We are visible in the right places, for the right searches, in ways that help the business create useful demand."

Less tidy. More honest.

Presence in AI search results

One of the first things to watch now is whether AI overviews appear for priority searches.

Not every query triggers one. Not every overview behaves the same way. Some are broad summaries. Some cite sources prominently. Some answer enough of the query that the user may not need to click. Some sit above results that used to carry the attention.

That changes the meaning of a ranking.

Position two under an AI overview is not the same as position two on a simpler results page.

So we look at the actual result page, not just the rank number.

That means checking:

  • whether an AI overview appears
  • which sources are cited
  • whether competitors are included
  • whether the summary answers the query well enough to reduce clicks
  • whether the remaining click is likely to be more qualified

This is not glamorous work. It involves looking at search results like a human being. Deeply unfashionable, apparently. Still useful.

Track which pages AI tools quote

In AI-era search, being cited can matter even when the click is less direct.

A citation can put a brand inside the answer. It can make a page part of the evidence layer. It can help a cautious buyer recognise a name before they search again later.

That does not mean citations are magic.

They are not a replacement for traffic, leads, or revenue. We should not invent a new vanity metric just because the old vanity metric is looking tired.

But citation presence is worth tracking for important query groups.

We want to know where the site appears as a source, where competitors appear, which pages are being used, and which topics seem to earn trust from the result page.

The useful question is not "can we force the AI overview to cite us?"

The useful question is "what page is trusted enough to be cited here, and do we have anything that deserves that role?"

That keeps the work grounded.

Track searches for your business name

Branded search has become more interesting.

If people encounter the business through an AI overview, a comparison page, a social post, a referral, or a sales conversation, they may not convert immediately. They may search the brand later.

That branded search is a signal.

It can show whether the market is remembering the business, whether content is creating follow-up interest, and whether people are moving from general discovery to direct evaluation.

For a smaller business, this can be especially useful.

You may not always win huge volumes of broad informational traffic. You may not need to. If more of the right people search for the business by name after encountering useful content or proof, that matters.

We watch branded search because it often tells a more commercial story than broad traffic alone.

The boring graph may be the useful one. This happens more often than dashboard designers would like.

Watch how click-through rates change

Click-through rate needs more careful interpretation now.

If impressions stay steady but clicks fall, that may mean the result page is answering more of the query before the click. It may mean an AI overview appeared. It may mean a competitor has a stronger title. It may mean the query itself has shifted. It may mean the page is still visible but less compelling.

The numbers alone do not explain which one happened.

That is why we connect click-through rate changes to the shape of the results page.

For a priority query, we want to know:

  • what appeared above the organic result
  • whether the title and description still make a strong case
  • whether the query has become more informational than commercial
  • whether the lost clicks were likely to be valuable
  • whether the page still supports later conversion

Sometimes a falling click-through rate is a real problem.

Sometimes it is a sign that low-value clicks are being filtered out earlier.

The job is to know which.

Track enquiries influenced by search

Last-click reporting has always been a bit rude to SEO.

Someone reads an article, compares a service page, looks at a case study, leaves, comes back through the brand name, and submits an enquiry. The final click gets the credit. The earlier pages sit there like unpaid interns.

That was already a problem.

AI-era search makes it more important.

If organic pages help buyers understand the offer, build trust, answer objections, or remember the brand, they may be doing useful work before the final conversion.

So we look for assisted value.

Which organic landing pages appear in converting journeys?

Which topics lead people towards service pages?

Which case studies support later enquiry?

Which articles get referenced in calls, emails, or sales conversations?

Not all of that lives neatly in analytics. Some of it comes from CRM notes, call feedback, email replies, and the deeply technical act of asking people what helped them decide.

Messy. Useful. Like most grown-up measurement.

Lead quality by page

Traffic volume is a blunt instrument.

Lead quality is sharper.

A page that brings fewer visitors but better enquiries may be more valuable than a page that attracts thousands of people with no buying intent.

This matters for SEO because the easiest traffic is not always the best traffic.

Broad informational posts can inflate reports. Commercially specific pages can look modest but do the real work. A services page that produces three serious enquiries may be worth more than an article that gets hundreds of visits from people who will never buy anything.

So we look at which pages create fit.

Do enquiries from this page understand the offer?

Do they match the kind of work the business wants?

Do they arrive with clearer expectations?

Do they move faster through sales?

That is the sort of SEO measurement that makes commercial sense.

It is less likely to impress someone who only wants a graph going up. That is fine. Some graphs are just decorative stress.

Watch how customer questions change

Search intent is not fixed.

The same keyword can become more informational, more commercial, more local, more comparison-led, or more answer-led over time.

AI overviews make this more visible, but the principle is older than AI.

If the result page changes, the strategy may need to change with it.

A page that used to work for a broad query may need to become more specific. A service page may need clearer proof. A content piece may need stronger internal links. A comparison page may now deserve more attention than a generic explainer.

Monitoring intent means asking what the search result is actually rewarding now.

Not what we wish it rewarded.

Not what it rewarded when the original content plan was written and everyone was feeling optimistic.

What it rewards now.

Technical health still matters

It would be convenient if AI-era SEO meant we could ignore technical SEO and simply think deep thoughts about citations.

No such luck.

Technical health still matters.

Crawlability, indexing, page speed, mobile usability, structured data, internal links, canonical signals, broken pages, redirects, and thin duplicate content still affect whether search engines can understand and trust the site.

The basics remain basic because they are foundational, not because they are old-fashioned.

If the site is technically messy, measurement gets messy too. You cannot sensibly interpret performance if the underlying site keeps tripping over itself.

That is not a glamorous point. It is, regrettably, true.

The measures we trust

For most businesses, the useful monitoring view is not wildly complicated.

It usually combines:

  • ranking movement for commercial terms
  • AI overview presence for priority queries
  • citation tracking across important topics
  • Search Console impressions, clicks, and click-through rate
  • branded search trends
  • organic landing pages in assisted journeys
  • lead quality by page or topic
  • technical health checks
  • periodic human review of actual result pages

The important bit is not the tool stack.

It is the interpretation.

A report should help the business decide what to do next. Improve a page. Build proof. Rewrite a title. Add internal links. Fix a technical issue. Stop producing filler. Create something citation-worthy. Invest more in a topic that produces fit.

If the report does not improve decisions, it is not reporting. It is weather.

Search reporting needs a wider view

Rankings still matter.

Traffic still matters.

But they need company.

AI-era SEO needs a wider view because search is now a wider environment. The answer may appear before the click. The brand may be cited before it is visited. The enquiry may happen days after the first organic touch. The commercially useful page may not be the one with the largest audience.

That makes measurement harder.

It also makes it better.

The goal was never to win a ranking report. The goal was to be found, trusted, remembered, and chosen by the right people.

That is what we monitor now.

If you want to understand what your search visibility is doing, ask us to review it. We will look beyond rankings and show what is moving traffic, trust, citations, and useful enquiries.

Occasional updates

One email,
when we publish.

One plain email when a new article is up, nothing else. We'll send one email first to check the address is yours. Every email includes an unsubscribe link.

Free 20-minute call

Recognise your own problem in a piece? Talk it through with us.

Analytics

Can we measure your visit?

Google Analytics helps us understand which pages are useful. It stays off unless you allow it. See our cookie policy for details.