Leverage your analytics platforms more effectively

Dan says: “Start understanding your analytics platforms a lot better, and start leveraging them – not just for attribution and ROI, but to look at your organic strategies and the second-order effects of what those strategies are bringing to the business.”

Have SEOs focussed less on analytics since the release of GA4?

“We didn’t put our heads in the sand, but it is a very different model. We'd grown accustomed to something. We'd grown used to reporting, and then bringing in user events and other event types really mixed that up.

A lot of opportunity was lost in terms of what could be set up with the analytics models itself, and it gave us greater scope. We probably did want to divert attention to data analysts and away from SEO in the actual setup of it, because it was like relearning a completely new skill.

Realistically, you don't always get paid for analytics; it's seen as a given in terms of using those platforms, reporting, and understanding the data. It's a heavy investment into a skill set that you don't necessarily get a direct return on.”

With users consuming content elsewhere and reduced visibility on that, what does this mean for attribution?

“This has always been a bit of a long-running joke in marketing channels. Marketoonist did a really good cartoon recently that showed a relay race where ‘long-term brand building’ is handing the baton over to ‘last click attribution’ to take it over the line. That's always been a problem in SEO.

Search is becoming more multimodal. People are consuming information, engaging with brands, and researching across not just Google and LLMs but TikTok, Instagram, etc. We're consuming in different ways, the journey is definitely not linear, and we're not stacking queries in the same way.

That's being compounded by an increased awareness of online security. People are more aware of what cookies are, and they're more aware of clicking banners and elements like that. It's becoming more privacy-focussed.

Overall, you want to be looking at more than just the direct attribution of channels, in organic specifically, because people may have multiple touchpoints with your brand. They'll be researching, they'll come back, and they'll come back directly or through different elements.

What's important is that you understand how you can use analytics effectively, which is why I mentioned things like second-order effects. If your content and efforts are surfacing you in organic, or you've been surfaced through an LLM, and then that user's coming to you through organic, then they're coming back to you a few times, in Universal Analytics, that was assisted attribution, and you could see that quite nicely.

It's a lot more complicated to set that up with Explorations reports in GA4, but you want to understand, if this user eventually converted through a social display ad, did they only see that social display ad because they've interacted with the site five times previously, and that's how they got the cookies?

It's more of an ecosystem now, as opposed to looking at individual channels and the direct ROI of those channels. The user journey is becoming messier, so channel interactions are becoming messier.”

Are stakeholders more likely to gravitate towards paid search because it's becoming harder to measure the value of organic?

“That's an element of it. With paid search, if you put £1 in, you get X return, but that's how paid search has always been sold. You've always looked at ROAS and very specific financial metrics like that, whereas, especially right now, there's a lot of uncertainty with organic.

There's a lot of hype from non-specialist publications, and there are even a lot of differences of opinion within the SEO industry itself, with different thought leaders echoing and amplifying different messaging. That level of uncertainty is affecting investment because people want to be sure with their investment. Llms.txt is a great example. As an industry, we can't even agree among ourselves what we're going to do with it, yet we're expecting stakeholders who don't have the level of expertise that we do to make committed investments.

By understanding GA4 better, and understanding how you can use audience insights and second-order effect attribution to paint better data pictures, you can create that confidence in what a modern organic program looks like.”

In an analytics platform like GA4, what are the key reports that SEOs need to be diving into?

“The Explorations section of GA4 is one of the most underutilised. We're still quite eager to put things into Looker Studio and manipulate the raw data. Whereas, through Explorations, you can get some very powerful insights into your business.

Each business will have different needs, depending on its objectives, but you can manipulate that data into whatever visualisations you want. You can get some really good insights into second-order attribution, how many touchpoints users are making before converting, performance timelines based on the drop-off from different pages, etc., and then overlay that with your other data.

If you can understand why people are going to page X, then from page X to page Y, you can understand that journey. If they're then dropping off from page Y to page Z, is there a reason for that? Maybe the objective isn't to get everyone to go from Y to Z, but you might want to take them back to X and convert them there.

It gives you a better understanding of the user journey. We've always looked at it as a funnel, and a funnel is very linear. Now, we're multimodal, so it's messy. We've gone from a straight line to spaghetti, and we need to amend our data analytics to be better at understanding spaghetti.”

Should an SEO be expected to do this kind of analysis by themselves?

“It comes down to expectation management with clients. There are a lot of pre-built tutorials and pre-built reports out there that you can easily implement in GA4. If you're working by yourself or you're working in a silo, it's about picking your battles and choosing what's doable.

If you want to try and scale the entire mountain of analytics and do all this on your own, even before the complexity of multimodal and the spaghetti got cooked, it would be very difficult from a time management perspective.

Think about what your clients care about and how you can influence that. You could even go back and use older, more traditional business methods like PESTEL (or PESTLE) analysis. What outside factors are going to affect this that you can bring to the forefront and educate your clients on?

Start to create that narrative of what the next 6-12 months looks like, and use the mechanisms that are available online. There are a lot of really good GA4 tutorials and guides to draw this experience from, so that you can do this in a way that’s manageable for you. Realistically, doing everything at once is impossible. Even for SEO teams, everything has to be done in a phased manner.

You need to understand how you can phase that out and prioritise it, for both your client and your workflow.”

By analysing these journeys, how can you show organic’s role in driving assisted conversions, nurturing awareness, and amplifying the effectiveness of other marketing channels?

“With organic, a lot of the content we produce will be aimed at top-of-funnel, middle-of-funnel, or bottom-of-funnel. If you're quite prominent top-of-funnel – that’s the content you're producing, and it's surfacing in the traditional blue links, it's being surfaced and cited in AI Mode and AI overviews, and it's getting that engagement – that's a brand touchpoint.

If that's how you're first introduced to the user, you can see this through GA4 Explorations reports, as long as everything's set up correctly beforehand (for things like cookie expiration, etc.). Then, they come back to you directly a couple of times. They may then decide to follow the brand on social. They may also engage by email and sign up to the newsletter because, when they came to the site, there was a nice pop-up. They may then convert through email two weeks later, from a different campaign.

In that case, the last click direct attribution goes to email, but the actual reason why they're on that email newsletter send list in the first place is because they discovered the brand through the first touchpoint in organic search. Whilst organic hasn't technically driven that conversion, it's enabled that conversion.

You could also argue that the user might have had a brand touchpoint through many other things. They could have seen a TV advert or an out-of-home advertisement with different elements, but we don't have the attribution levels or the data to show that.

John Wanamaker once said, ‘Half the money I spend on advertising is wasted; the trouble is I don't know which half.’ We're in that messiness, but if we've got the data to say that they definitely came through this blog post, this guide, or this XYZ, and then they converted five weeks later through an email campaign, and it's that same user and the same journey, that gives us an idea of what people are still discovering through this channel.

It's the upstream value of the conversion of organic. There may be other efforts that are influencing LLMs, bottom-of-funnel and middle-of-funnel conversations, and different elements like that. It's a lot messier right now. It's a lot more nuanced and qualitative, and it requires more data analysis beyond tables of how many sessions you had last month, how many new users, how many conversions, how much revenue, etc. It's just much messier.”

What are the big mistakes that SEOs are making with GA4 implementation at the moment?

“From a lot of the analysis I've done and studies I’ve read (especially around things like LLM conversions), the biggest mistake is key event bloat, where people are setting up everything as a key event.

By default, you have a certain number of key events in GA4. I've looked at datasets that were sent anonymously through Looker Studio, where 103% of the sessions had key events. That's because somebody set up a scroll back as a key event as opposed to a user event, or people haven't made use of the key events properly.

Then, when stakeholders are asking questions about a website, you don't know because, during the initial setup, no one did an audit of what you need that data to say, what that data could potentially answer, and how you track it.

In Universal Analytics, you had some custom events and you had enhanced e-commerce, and it was relatively simple. There's so much more manipulation and data you can see from GA4 that hasn't been invested in. Analytics has always been viewed as free, and any investments are reluctant investments, and it’s not necessarily valued in that way.”

Dan, what's the key takeaway from the tips you shared today?

“Go back to your analytics. Get it set up. Understand what questions you need to answer and what data points you need to answer them.

Then, start using the Explorations reports to create a better data narrative and storytelling, and do it in a scalable way that works for your businesses and clients.”

Dan Taylor is Partner and Head of Innovation for SEO and AI at SALT Agency. Find out more over at SALT.Agency.