Please note, for speed, I have used AI to generate this post based on research I have performed for my brightonSEO April 2026 talk.
SEO reporting is still acting like users politely walk from ranking to click to conversion in a straight line.
Which is adorable.
Completely wrong, but adorable.
For years, the basic model was:
Rank → Click → Session → Conversion
Nice. Clean. Comforting.
Also increasingly useless as a complete explanation of what’s actually happening.
Because modern search doesn’t work like that anymore.
Users now see brands in AI answers, featured snippets, comparison boxes, Reddit threads, YouTube results, review sites, social posts, and whatever else Google has decided to wedge between the query and your website this week.
They might see you today, search for you tomorrow, come back direct next week, and convert through another channel entirely.
And then someone looks at the SEO report and says:
“Traffic is down. Is SEO working?”
Great question.
Wrong question.
The problem isn’t that SEO stopped working.
The problem is we’re still measuring it like everyone is patiently clicking blue links like civilised little spreadsheet entries.
They’re not.
So here’s what SEO teams can actually do next week, and what they should build next quarter.
What to do next week
Don’t start with a transformation programme.
Don’t build a 47-tab dashboard called “Organic Search Intelligence Hub”.
Start with three things.
1. Separate reporting by layer
Most SEO reports squash everything together:
- rankings
- impressions
- clicks
- traffic
- conversions
- revenue
Then we act surprised when the story makes no sense.
Instead, split reporting into three layers.
Presence
Did we show up?
This includes:
- impressions
- rankings
- query visibility
- page coverage
- SERP features
- AI visibility where measurable
- local or product visibility
This is the visibility layer.
It tells you whether you were part of the search environment.
Not whether someone loved you.
Not whether they bought.
Just whether you showed up.
Important distinction.
Preference
Did people lean towards us?
This is the bit most SEO reports skip.
Useful signals include:
- branded search growth
- CTR changes
- returning users
- direct visits
- assisted conversions
- engagement
- self-reported discovery, such as “How did you hear about us?”
None of these are perfect.
But neither is pretending traffic explains human behaviour.
Preference is where SEO influence often starts to show up before the final conversion.
Someone sees you.
They remember you.
They search for you later.
They come back.
They compare you.
They choose you.
That is not always captured in a neat organic session.
Annoying, yes.
But also reality.
Performance
Did value follow?
This is the business layer.
Depending on the site, this could be:
- leads
- sales
- quote starts
- demo requests
- bookings
- calls
- pipeline
- revenue
- subscriptions
This is where SEO has to connect to commercial impact.
But the key point is this:
Performance should not be the only layer you report.
Because if you only report the final outcome, you miss how SEO helped create the conditions for that outcome.
2. Add preference signals into reporting
Traffic has been massively over-promoted.
It still matters.
But it is not the whole story.
If someone sees your brand in an AI answer, doesn’t click, then searches your brand two days later, your old SEO report probably shrugs and says:
“Nothing to see here.”
Useful.
So next week, add a few preference signals.
Start with:
Branded search
Is more demand showing up for your brand?
Use:
- Google Search Console
- Google Trends
- SEO tools if available
This helps answer:
Are people starting to look for us specifically?
Returning users
Are people coming back?
Use:
- GA4
- Adobe Analytics
- whatever analytics platform you have
This helps show whether SEO is creating remembered interest, not just one-off visits.
Direct traffic
Direct traffic is messy.
It includes people who typed the URL, bookmarks, lost attribution, weird tracking gaps, and probably a few ghosts.
But directionally, it can still be useful.
If visibility and branded search are rising, and direct traffic also rises, that is worth paying attention to.
Not as proof.
As a signal.
Assisted conversions
SEO does not always close the journey.
Sometimes it starts it.
Sometimes it supports it.
Sometimes it sits quietly in the background doing useful work while another channel takes the credit, like the office introvert of marketing.
So add assisted conversions or path analysis where you can.
Use:
- GA4 attribution paths
- Adobe journey reports
- CRM attribution
- HubSpot/Salesforce where available
The aim is not to prove everything perfectly.
The aim is to stop reporting SEO like it only matters when it gets the last click.
Self-reported discovery
Ask people:
“How did you first hear about us?”
Or:
“What made you look for us today?”
This is not perfect data.
People forget.
People simplify.
People lie to forms because forms deserve it.
But it can reveal things click data misses:
- AI tools
- word of mouth
- social posts
- podcasts
- events
- communities
- previous search exposure
Use it carefully.
But use it.
3. Stop treating traffic as the whole story
Traffic is not useless.
Let’s not do the usual marketing thing where one metric disappoints us and we immediately declare it dead.
Traffic still matters.
But traffic is no longer the scoreboard.
It is one signal.
In modern search, you can:
- improve rankings
- grow impressions
- appear in richer SERP features
- get included in AI answers
- influence more people
…and still get fewer clicks.
That is not always failure.
Sometimes it means the search result answered more of the query before the user reached your site.
Sometimes it means the user came back later another way.
Sometimes it means SEO created awareness, but not the tracked session.
So instead of only asking:
“Did traffic go up?”
Ask:
“Where did we show up?”
“What behaviour changed?”
“What outcomes followed?”
That is a much better conversation.
And much harder to reduce to one sad line graph.
What to build next quarter
Once you’ve made those changes, build something slightly more joined-up.
Not a perfect attribution model.
Perfect attribution is the corporate version of Bigfoot.
Lots of people claim they’ve seen it.
No one has convincing evidence.
Build a simple influence view instead.
Build a simple influence view
The influence view should answer three questions.
1. Where did we show up?
Track your presence.
This could include:
- impressions
- rankings
- priority keyword visibility
- page-one coverage
- SERP features
- AI mentions
- AI citations
- Google Business Profile visibility
- product visibility
- marketplace visibility
The exact metrics depend on your business.
For a SaaS company, it might be category queries, comparison terms and AI mentions.
For ecommerce, it might be product visibility, merchant listings and category rankings.
For local SEO, it might be map visibility, calls, directions and profile interactions.
For lead gen, it might be non-brand visibility across problem, solution and comparison queries.
The point is simple:
Before you ask whether SEO converted, ask whether you were even present in the places decisions were forming.
2. What behaviour changed?
This is where reporting gets more useful.
Look for changes in:
- branded search
- CTR
- returning users
- direct traffic
- engagement
- assisted conversions
- repeat visits
- form starts
- quote starts
- demo page visits
- comparison page visits
This layer helps you spot demand forming.
Because influence often appears as behaviour before it appears as revenue.
Someone searches.
Someone compares.
Someone comes back.
Someone searches your brand.
Someone visits the pricing page.
That matters.
Even if it doesn’t fit neatly into last-click reporting.
3. What outcomes followed?
Then look at outcomes.
Depending on the business, this might be:
- leads
- sales
- revenue
- quote starts
- demo requests
- bookings
- calls
- subscriptions
- pipeline
- qualified opportunities
This is where SEO has to earn its keep.
But the point is not to pretend every outcome can be tied neatly to one organic click.
The point is to show patterns.
For example:
- visibility increased across priority non-brand queries
- branded search rose afterwards
- returning visits increased
- assisted conversions improved
- qualified leads increased
That is not courtroom evidence.
It is business evidence.
And business evidence is usually what decisions are actually made on.
Track patterns, not perfect causality
This is the big shift.
Old SEO reporting tried to say:
“This ranking caused this traffic, which caused this conversion.”
Nice when it works.
Increasingly rare.
Modern reporting needs to say:
“Here is where we showed up. Here is how behaviour changed. Here is what happened commercially afterwards.”
That is more honest.
It is also more useful.
Because in a messy journey, the goal is not fake certainty.
The goal is directional truth.
If the same pattern repeats across pages, journeys, campaigns and time periods, you probably have something worth acting on.
If it only appears once in a dashboard after someone has filtered the data 19 different ways, maybe calm down.
What about AI search?
AI visibility is worth tracking.
But please don’t turn it into another nonsense scoreboard.
Right now, you can credibly track:
- whether you are mentioned
- whether you are cited
- whether competitors appear more often
- whether AI tools send referral traffic
- whether AI visibility overlaps with branded search or assisted demand
That is useful.
But it is still early.
What is not credible is pretending we can already measure precise AI ROI from every prompt, answer and citation.
We can’t.
Prompt results vary.
Tools disagree.
Volumes are unclear.
Personalisation changes outputs.
Attribution is messy.
So treat AI visibility as a presence signal.
Useful.
Directional.
Not magic.
A simple reporting structure
If I were rebuilding an SEO report now, I’d use this structure:
1. Executive summary
One paragraph:
What changed, why it matters, what we recommend next.
Not 12 charts and a nervous appendix.
2. Presence
Where did we show up?
Include:
- impressions
- rankings
- visibility
- SERP features
- AI mentions/citations where relevant
- local/product visibility if relevant
3. Preference
Did behaviour shift?
Include:
- branded search
- CTR
- returning users
- direct visits
- engagement
- assisted conversions
- self-reported discovery
4. Performance
Did value follow?
Include:
- leads
- sales
- revenue
- pipeline
- bookings
- calls
- quote starts
- demo requests
5. Interpretation
This is the bit people skip.
Do not just show the numbers.
Explain the pattern.
For example:
“Non-brand visibility increased, but traffic stayed flat. That suggests visibility is improving, but click capture is weaker. However, branded search and assisted conversions both increased, so SEO may be contributing more to demand and consideration than the traffic line alone suggests.”
That is a useful SEO report.
A traffic graph with a red arrow is not.
What to avoid
Avoid reporting that says:
“Traffic is down, therefore SEO is down.”
Too simplistic.
Avoid reporting that says:
“AI visibility is up, therefore revenue will follow.”
Too optimistic.
Avoid reporting that says:
“This exact blog post generated this exact pipeline figure.”
Maybe.
But probably not as cleanly as the slide suggests.
Avoid building a giant influence score where no one understands the formula.
That’s not strategy.
That’s spreadsheet cosplay.
The real shift
The shift is not:
“Stop measuring traffic.”
The shift is:
“Stop asking traffic to explain everything.”
Traffic can tell you that a visit happened.
It cannot always tell you:
- where the decision started
- where preference was formed
- where the brand was remembered
- what role search played before the click
- what happened when the user didn’t click
That is why SEO reporting needs to evolve.
Not into something more complicated.
Into something more honest.
The takeaway
Next week, do three things:
- separate reporting by layer
- add preference signals
- stop treating traffic as the whole story
Next quarter, build a simple influence view:
- where you showed up
- what behaviour changed
- what outcomes followed
That’s it.
Not perfect attribution.
Not a miracle dashboard.
Just a better way of reporting SEO in a world where users are no longer behaving like obedient little funnels.
Which, frankly, was rude of them.