Guest post by Rohan Ayyar
If you believed all the CRO posts you saw, and took them at face value, you’d think CRO was all you needed. Just look at marketers tweaking their landing pages and getting 1000% increases in conversions! Like any tool, though, CRO doesn’t work in isolation and it has to be used right to get the right results. So if you’re doing CRO and it’s not paying off for you, maybe you’re falling into one of these CRO traps?
It’s not about your landing page
Your landing page isn’t the last chance your visitors have to convert, ever. You haven’t lost them for all time if you don’t get them now. It’s not a rom-com, your landing page isn’t the airport. Conversion journeys are increasingly nonlinear; take a look at the myriad touchpoints Forrester Research uncovered along a typical digital purchase funnel:
Conversion optimization needs to be too, which means every touch is the trigger and informant for retargeting.
That has implications for both strategy and data. If your data shows super high bounce and it’s from your landing page you’ll likely have poor conversion figures that mislead you about the efficacy of your marketing efforts. And your strategy needs to widen in scope to take account of multiple entrances – you don’t have to push customers through your landing page, force them to sign up, mug them with exit intent popups, offer popups and screaming CTAs.
Most of what you do will be off-site. Says Gerrid Smith, CEO of digital agency Black Fin, and a specialist in in niche marketing:
“The truth is, things we do on a client’s website will typically only account for about 30% of the SEO effects we’ll generate. The other 70% happens offsite.”
Users come to your landing page, click away, and a week later that same user is on your homepage or reading your blog. Having an effective, widespread offsite presence is vital to funnel users back to your site – when they’re ready. Don’t try to stop them leaving.
Depending on your industry, your landing page might not even be the place conversions really manifest. You need a strategy that addresses the whole conversion journey, not a quick fix for something that might not even be the problem.
So if your landing page isn’t the source of all your conversion rate issues, what is? How do you set about finding the root cause?
It’s not about your keywords either
Of course, you want to drive more qualified traffic to your assets – landing pages, blog posts, anything where you can get a conversion. That starts with keyword intent – what are your audience looking for?
Depending on the type of conversion you’re after, identify the appropriate type of keyword and intent that you’re targeting: informational, navigational, or transactional/commercial. If you want more local mobile searches to find your store, using navigational language might work better for you. If you want to encourage more content downloads, answer the appropriate informational queries. Most marketers will be seeking to increase sales, in which case they should target transactional search terms.
You can’t A/B test your way out of everything
A/B testing isn’t the same thing as optimization, and there are plenty of things it sucks at detecting. As many as seven out of eight A/B tests are unreliable; best practice advice varies from space to space (and from expert to expert). So what do you do? Track conversions, and conversion behaviors, using techniques that cover more variables and more of the funnel.
When should you A/B test?
A/B testing works best when you’re comparing fairly big changes. If you focus on making small changes to small things, you need more traffic to get a result that shows up against the background noise of the numbers.
The bigger the change that you make, the more quickly you’ll see results. If there’s a positive and a negative, A/B testing will tell you which one’s which. If there’s nothing to choose between two options, A/B testing will give you a tiny, statistically insignificant false positive.
So what else is there?
Many businesses A/B test because they want data and they know that A/B tests work. But that’s like banging in screws with a hammer because you’ve used it before and it worked fine. There are other testing methods out there than A/B tests.
Multi-arm bandit testing is named after “one-armed bandits” – slot machines where you pull a lever to gamble. Ironically though, MAB is less of a gamble than A/B tests. Without getting into too deep a dive about exactly how it works (though if that’s what you want, here it is), bandit testing allows you to run ongoing tests on small variables.
If you want to know which of two options greatly outperforms the other, use A/B tests.
Sites such as BuzzFeed and The Washington Post often work with a custom MAB setup where the algorithm automatically presents the articles in the order of most likely to be clicked, read and shared by readers. No wonder, then, that the Content Experiments statistical engine in Google Analytics runs on MAB.
Ignorance is bliss but doesn’t get sales
You know you want conversions. But can you correctly identify meaningful conversions? Often, conversion optimization sucks because it’s optimized for the wrong conversion metrics.
What are your goals?
Measuring conversions might mean measuring social shares. It might mean measuring email signups. It might mean content downloads. There’s no one action, and not every conversion is a sale. CRO only makes sense in the context of a marketing process where you’re seeking to optimize specified conversion types at each stage, according to a wider plan. If you’re not exactly sure what you mean by a conversion, what your goal is for this funnel stage, and how that fits into your overall strategy, it might be time to zoom out and re-examine the whole process before you get hung up on 0.1% here or moving buttons around there.
Start small and grow out of it.
If you’ve never done any CRO before and your whole marketing process is a mess, just finding a solid how-to on the basics and implementing it will yield impressive results. But after that, you need a base in data. That means before you say, “Go team, get us more conversions,” you have to know what a conversion means to you – the value it adds to your bottom line.
You might be looking for leads in all the wrong spaces… If you are hoping to get a big increase in conversion rate, make sure you focus on big changes.
Neil Patel, in an evergreen post on Quick Sprout about A/B testing pitfalls, wrote:
“If you really want to move your conversion rates, don’t focus on small changes. Instead, focus on drastic changes as they are the ones that boost your revenue.”
Some traffic is more valuable than others
Traffic is what you convert. If your traffic is junk, you can’t convert it, no matter how optimized your conversions strategy. If it’s great, CRO works incredibly well. Making sure your traffic is convertible can have a powerful effect on everything else down the funnel, so it makes sense to tick this box first.
Tracking traffic sources and acquisition channels in analytics and correlating with your desired conversions lets you see which traffic sources are feeding you useful traffic. Some sources will always be noisier than others – for instance, your organic search traffic might be lower-converting than email because it’s not prequalified.
But other sources will be underperforming, either by lower numbers or lower conversions than industry benchmarks. That lets you identify traffic acquisition and qualification opportunities that will increase the power of your CRO efforts down the line.
If you use a marketing automation software (such as Marketo) that correctly identifies and attributes traffic to various channels along your custom-defined conversion funnels, you might end up with a table that looks like the one below. Alternatively, you can use the Goal Flow report in Google Analytics to track the efficiency of your funnel over time by comparing traffic sources to conversions.
Over to you
CRO goes wrong when businesses act like it’s the solution to all their marketing problems. It’s not a panacea and it’s not a black box where magic happens. You can get people to convert if you can reach them, so if your traffic is junk, your CRO will suck too. You can only tell what works and what doesn’t if your testing is set up right.
And you can only hit your goals if you define them and delineate them in terms of wider strategy. Nail that and CRO will deliver in spades.
Rohan Ayyar is the go-to guy for scalable SEO services at E2M, India’s fastest growing digital agency. Rohan is a pro at helping startups gain quick visibility online with creative content strategies. Search Engine Watch lists him as a must-follow SEO expert on Twitter – you can do that at @searchrook.