A few months ago I was given a perfectly sensible piece of marketing advice.
I agreed with it… but I tested it anyway.
I'm very glad I did, because the results weren't what either of us expected.
I've been working with Premier Veins for over two years as founder Martin Claridge's Fractional CMO. Premier Veins provides private varicose vein treatment through clinics across the UK, and Google is a hugely important source of new patients, both through organic search and paid advertising.
Since I’m not an SEO expert, I’ve been partnering with a local digital marketing agency to improve Premier Veins' SEO. One of their recommendations was to add significantly more useful content to the individual clinic pages.
This made complete sense. If someone searches for "varicose vein treatment Birmingham", we want Premier Veins' Birmingham clinic to rank prominently. Likewise Leicester, Solihull, Cardiff and the other clinic locations.
More useful, locally relevant content gave us more opportunities to mention the location naturally, answer the questions potential patients might have, demonstrate Premier Veins' expertise and build the relevance and authority of those pages for Google and AI search.
The agency rewrote them. The content was good. The SEO rationale was good.
The clinic pages aren't just there to attract organic search traffic. They're also landing pages for our Google Search Ads.
And Premier Veins spends a significant amount of money sending people to them.
So I said something along the lines of: "I think it's unlikely the copy will negatively affect our conversion rates, but we spend a lot of money driving traffic to the site so it'd be remiss of me to assume conversions won't be affected. Let's do an A/B split test to be sure."
That's the engineer in me.
Marketing has too many interconnected variables to assume that changing one thing won't affect something else. So rather than replace the existing Birmingham clinic page with the longer, SEO-optimised version, we split the traffic between them and measured what happened.
I expected the results to be fairly similar.
They weren't.
The existing Birmingham clinic page received 426 visits and generated 73 conversions, giving us a conversion rate of 17.14%.
The new, much longer version received 430 visits and generated 53 conversions. Its conversion rate was 12.33%.
Almost exactly the same amount of traffic. Twenty fewer consultations booked.
I was genuinely surprised.
And when you're spending a significant amount of money every month driving people to those pages, that's not a difference I'm prepared to shrug off because another marketing metric might be improving.
But this is where the story gets more interesting.
Because if you're already thinking, "Well, obviously short landing pages convert better", I've got another A/B test for you.
And it says almost the exact opposite.
A few months before the SEO test, we'd been trying to solve a different problem with Premier Veins' Google Ads.
We use Performance Max, usually shortened to PMax, alongside our Search campaigns. PMax allows Google to show Premier Veins ads across its different channels, so the circumstances in which somebody encounters an ad can be quite different from a conventional Google search.
That's important because someone typing "varicose vein treatment Birmingham" into Google is actively looking for a solution. They're already thinking about their veins. They've decided to investigate treatment and they're deliberately searching for information.
Someone who's encountered Premier Veins through an ad that interrupted them when they were thinking about something completely different may need much more convincing.
We were getting a very low conversion rate from our PMax traffic, so we created a dedicated landing page for it.
And it's REALLY long.
It's got tonnes of information, videos, FAQs, treatment explanations and reassurance. It's designed to answer the questions somebody might have when they're interested enough to click, but haven't necessarily spent the past ten minutes actively researching varicose vein treatment.
The result?
Our conversion rate increased from 0.08% to 0.61%.
Yes, 0.61% still sounds pretty low. But it's more than seven times better than it was. Seven times as many consultations booked for the same amount of ad spend.
And we didn't change the ads at all.
That's an important point for anyone who's ever concluded, "We tried Google Ads and they didn't work."
Maybe the ads didn't work.
Or maybe they did their job perfectly well and the problem was what happened after somebody clicked.
We're still looking at ways to improve the PMax conversion rate further, but the experiment had already taught us something useful. For that particular audience, at that particular point in their buyer journey, giving them a LOT more information worked.
Then along came our Birmingham test and the much longer page performed substantially worse.
Helpful.
Welcome to marketing.
You already know now that this is the wrong question.
Our PMax test doesn't prove long landing pages convert better. And our Birmingham test doesn't prove short landing pages convert better.
They tell us something much more useful.
Context matters.
Someone actively searching for varicose vein treatment is in a different headspace from someone whose browsing has just been interrupted by an advert.
They may be looking at the same treatment from the same company, but they're at different points in their buyer journey. They have different levels of intent, different questions and different amounts of convincing still to do.
So why would we automatically assume the same landing page should work equally well for both?
This is also why I'm sceptical about marketing "best practice". Benchmarks and other people's test results can give you sensible hypotheses, but they're not proof of what your customers will do.
Your customers have an irritating habit of not reading marketing textbooks.
There's a bigger point here than landing-page length.
The agency’s SEO recommendation was good. I want to be really clear about that, because this isn't a story about an SEO specialist making a mistake. The objective was to increase the relevance and authority of individual clinic pages so they could perform better for geographically relevant searches. Adding more useful, locally relevant content was a perfectly sensible way to pursue that objective.
But the clinic page doesn't exist in an SEO bubble.
It sits inside a much bigger marketing system.
I wrote recently about why spike tests don't work in marketing, and used the analogy of a spider's web. Tweak one strand and the rest of the web quivers.
We weren't trying to change our Google Ads performance. We were trying to improve our SEO. But the pages we wanted to optimise for organic search were also the pages receiving paid traffic.
One change. More than one consequence.
And this is the awkward bit. Hire an SEO specialist and they'll quite rightly report on SEO metrics: rankings, impressions, clicks, organic traffic. Hire a social media specialist and they'll report on social media metrics. Every specialist is measured on their own part of the machine, incentivised to improve their own part of the machine, and will honestly tell you that their part of the machine is performing better.
Nobody in that arrangement is measured on whether the business made more money.
So you can end up in a situation where everyone is doing a good job, every report looks encouraging, and sales haven't moved. Nobody's lying to you. Nobody's incompetent. The gap is simply that nobody's job is to look at the whole thing.
Your SEO can improve without your marketing improving.
That's not a criticism of SEO. The same is true of every marketing specialism.
It's also one of the reasons I love networking with good marketing specialists. They're often the first people to say, "My numbers look great and the client still isn't getting more sales." That's a clue. And it's usually the moment to stop looking at one individual specialism and start looking at how the whole machine fits together.
I've written before about the difference between having an overall marketing strategy and having strategies for individual marketing channels. An SEO strategy, Google Ads strategy or social media strategy can all be excellent, but somebody still needs to make sure they're serving the wider business objective. That's the point behind my 3-question test for marketing strategy.
The question here was never simply, "Will this help our SEO?"
It was, "What else might happen if we do this?"
We're not throwing the SEO work in the bin. That would be a rather spectacular way to miss the point.
Instead, we're wondering whether we've been asking one webpage to do two different jobs.
From an SEO perspective, richer clinic content gives us more opportunity to build relevance and authority around searches such as "varicose vein treatment Birmingham". It may help more people discover Premier Veins organically and give those visitors useful information when they arrive.
But our A/B test suggests that the people we're paying to send to that Birmingham page through Search Ads are more likely to convert on the shorter version.
So perhaps the answer isn't deciding which page is "better".
Perhaps it's having both. Richer clinic pages designed to attract and inform organic search visitors, alongside more succinct landing pages built specifically for high-intent paid Search traffic.
There are pros and cons to separating them, and we're weighing those up now.
We've already learnt from PMax that different traffic can respond very differently to different landing pages. But that doesn't mean I'm going to assume separate Search landing pages will automatically work better.
I think I've learnt my lesson about that.
We'll test them.
I went into the Birmingham A/B test thinking the extra content probably wouldn't affect our conversion rate.
I was wrong. I don't mind admitting that.
Because my job isn't to correctly predict the result of every marketing experiment. It's to make sure we find out what actually happens before we roll a change out across the whole machine and bet a substantial amount of money on the assumption that we're right.
Good marketing isn't about blindly following best practice. And it isn't about collecting specialists and hoping that if everyone makes their own numbers go up, the overall result will somehow be more sales.
It's about understanding how the pieces fit together, measuring what matters and testing your assumptions.
Sometimes a test proves your hypothesis. Sometimes it proves you wrong. And sometimes a perfectly sensible SEO improvement knocks 28% off the conversion rate of a page you spend a lot of money sending people to.
Excellent.
Now we know.
And finding that out in an A/B test is considerably cheaper than discovering it six months later in the sales figures.
If you take one thing from this case study, I’d like it to be this question:
Who in your business is accountable for whether the whole machine got better, rather than one part of it?
If the honest answer is "nobody", that's not a failure. It's the single most common gap I find in businesses that have grown beyond founder-led sales but haven't yet built a proper marketing function. Everyone's doing their job. The job that's missing is the one that joins them up.
If you'd like a hand working out whether your marketing is actually working, get in touch and let's have a conversation about it.
And if you're not ready for that yet, start with the numbers. My Marketing Metrics Guide explains which measurements actually tell you whether your marketing is moving the business forward, and which ones just look reassuring on a slide.