Most SEO ROI reports still lean on one formula: incremental organic revenue minus SEO cost, divided by SEO cost, times 100. It made sense for years. It’s also starting to fall apart, because a growing share of what SEO does now happens somewhere this formula can’t see.
NP Digital’s Rob Tindula wrote a piece for Search Engine Land recently, making this case worth taking seriously, whether you’re an in-house analyst or running an SEO agency that has to explain performance to a client every month. He lays out three ways to patch the gaps in the old model. None of them is perfect. They’re better than what most teams are using, which is honestly the bar that matters here.
Why the old formula stopped telling the full story
60% of searches now end without a click. That number keeps climbing. Add AI Overviews and chatbot answers that often don’t bother passing referrer data back to your analytics, and you’ve got a real problem: a lot of SEO’s actual work has become invisible to the tools used to measure it.
This shows up in an annoying, familiar way. Visibility climbs, impressions climb, maybe you’re getting cited in AI Overviews or showing up in chatbot answers, and the organic traffic line in GA4 looks flat or even down. Anyone reporting on this knows the awkward conversation that follows. The fix isn’t a new tool; it’s measuring three things most reports skip entirely: defended traffic, assisted conversions, and content marketing that earns its keep outside the organic channel.
1. Stop Only Counting The New Wins, Count What You’re Defending Too
Here’s an analogy Tindula doesn’t use but this works better than the official framing: judging SEO only on incremental gains is like judging a goalkeeper purely on goals scored. Most of the job demands keeping things from getting worse, not making them better, and that’s genuinely hard to put a number on.
So the starting point for a better model isn’t incremental organic revenue. It’s total organic revenue, and SEO plays a critical role in protecting it. A flat line in a market that’s otherwise declining is a win. Most ROI models can’t see that win at all, which is a real flaw, not a minor one.
Except that claiming all of that revenue runs into an obvious objection, and a fair one. If most of your organic traffic is branded, someone typing your company name into Google, that’s not really SEO’s win to claim. That demand existed already, built by PR, ads, word of mouth, whatever got your name into someone’s head in the first place. SEO just happened to be standing there when they searched for it.
So you split the traffic. Pull branded versus non-branded from Search Console, since GA4 won’t hand you this cleanly on its own, then apply different weights to each. Tindula’s example uses 10% credit on branded clicks and 100% on non-branded, which is a starting assumption, not a law of physics. Some teams will argue for a different split, and that argument is fine to have. The point is having a defensible number instead of an indefensible all-or-nothing claim.
A worked version, using his numbers: branded traffic sits at 70% of clicks and is sliding, non-branded sits at 30% and is climbing.
- Branded: 70% x 10% credit = 7%
- Non-branded: 30% x 100% credit = 30%
- Blended weight: 37%
Apply 37% to a site doing $100,000 a month in organic revenue, and SEO gets credited with $37,000. Not the whole figure, not nothing. It’s a discount you’re applying to yourself before anyone else has to, which oddly enough tends to make the number more believable, not less.
2. Claim the Conversions You Assisted, Even When Something Else Closed the Deal
Last-click attribution has been unfair to SEO for years, and it’s gotten worse, not better. A lot of the time, organic is the very first thing someone encounters in their research, sometimes just an impression they never even clicked. Last-click throws all of that away and hands the credit to whatever channel happened to be there at the finish line.
GA4 doesn’t make fixing this easy, annoyingly. Universal Analytics used to show assisted-conversion value without much fuss. GA4 wants you to export path data into BigQuery if you want a proper fractional value per channel, which is more setup than most teams have patience for.
There’s a workaround that’s good enough for most reporting, though. GA4’s data-driven attribution model already splits conversion credit across channels based on actual influence, so the early- and mid-touch portions of that credit work as a stand-in for the assist value last-click throws out. It’s not exact. It’s close enough to be useful, and useful beats exact in a report someone’s actually going to read.
A sample pull might look like this:
- Early-touch credit: 1,345.69
- Mid-touch credit: 687.34
- Total: 2,033.03
- At $100 per conversion: Roughly, $203,303 in assisted value SEO can reasonably point to
Late-touch credit gets dropped from this on purpose. That’s usually where branded search clusters, the same demand-capture issue from the first method, so including it would just double up the same overclaim from a different angle.
3. Don’t Ignore What Seo Content Marketing Does Once It Leaves the Organic Channel
A lot of content marketing that started as an SEO play doesn’t stay there. Paid search runs ads pointing to landing pages SEO wrote. Social pulls quotes and angles from blog posts that were never meant for social in the first place. Email sequences lean on resources that started life as an organic article nobody on the email team had anything to do with writing. None of that shows up anywhere if the report only looks at the organic line.
Tindula’s fix takes three steps: figure out which SEO-built pages get reused elsewhere, work out what share of conversions in those other channels touch those pages, then apply that share to the channel’s conversion value.
A small example: paid search closes 500 conversions worth $100 each, and 5% of those conversions touch a page SEO originally built. That’s $2,500 in downstream credit. Not huge by itself, I’ll admit, but it adds up once you’re running this across every channel and every quarter, and it’s exactly the kind of content marketing crossover that quietly disappears from a report if nobody bothers to track it.
Putting the 3 Methods Together
None of these three methods is airtight, and Tindula doesn’t pretend otherwise. What they need to be is consistent and defensible, applied the same way every reporting period, so the trend means something even if any single month’s number involves a bit of estimation.
A few things worth keeping in mind if you’re building this out for real:
- Claim the full organic picture, but don’t pretend branded search is something SEO invented out of nothing.
- Look past the organic silo entirely. Assisted-conversion models exist because last-click genuinely underrates whatever channel starts the journey, and SEO usually starts more journeys than it gets credit for.
- Track where SEO content earns money outside its own channel, instead of letting that work disappear from the report just because it happened somewhere else.
- Accept that some of this math is going to be a bit rough, and that’s fine. An outdated formula is a worse outcome than an estimate someone can defend in a meeting.
This kind of reporting is genuinely more work than pulling last-click numbers once a month, which is exactly why a lot of in-house teams hand it off to an SEO agency rather than trying to build it themselves. It matters even more anywhere the attribution gets genuinely messy: multiple brands, multiple regions, the kind of complexity that shows up once you’re running enterprise SEO services across an organisation that’s big enough to have five different teams arguing over the same dashboard.
Conclusion
The old SEO ROI formula was never wrong exactly, just built for a simpler internet. Defended traffic, assisted conversions, and content doing quiet work in other channels are all real value; a click-only model was never going to catch. As more search happens without a click at all, whoever’s measuring SEO needs models that catch up to that, or the people signing off on the budget keep getting handed a smaller, less accurate picture than the one that’s actually true.
Techosoft Solutions helps Australian businesses build SEO strategies backed by reporting that actually reflects the value being created, not just what last-click attribution happens to show. From enterprise-scale technical SEO to content built to perform across multiple channels, our team focuses on results stakeholders can trust. Get in touch with Techosoft Solutions to see how we can help your business build a clearer, more complete picture of SEO ROI.
FAQs
Why doesn’t the traditional SEO ROI formula work anymore?
It was built for a clickable web. With AI Overviews and chatbot answers increasingly resolving queries without a click, and referrer data often missing entirely, a formula based only on incremental click-driven revenue misses a growing share of SEO’s actual impact.
How do I separate branded and non-branded traffic if Google Analytics doesn’t split it cleanly?
Pull the branded-versus-non-branded breakdown from Google Search Console instead, then apply that ratio to your total organic revenue using a weighted model, giving SEO partial credit for branded clicks and full credit for non-branded ones.
Is it dishonest to claim credit for assisted conversions that SEO didn’t directly close?
No, as long as the credit is fractional and clearly labelled as assisted rather than presented as a direct, last-click win. GA4’s data-driven attribution already calculates this kind of fractional credit, so the approach reflects how the platform itself already values each channel’s contribution.
Do small businesses need this level of ROI modelling, or is it only relevant for large organisations?
The underlying principles, defending existing traffic, crediting assists, and tracking cross-channel reuse, scale down fine for smaller sites. The more advanced steps, like BigQuery exports for precisely assisted conversion values, tend to matter more once a business reaches the size and complexity, where enterprise SEO services come into play.
