In-house vs. agency experimentation: an honest comparison
With mobile phones now accounting for more than half of all the time we spend online, it’s never been more important to cater to the needs of mobile users.
We therefore thought it was about time that we wrote a detailed blog post sharing the most effective strategies that we’ve found for optimizing our clients’ mobile websites and ratcheting up their mobile conversion rates.
Here it is.
- What is mobile conversion rate optimization?
- Why is mobile CRO important?
- 12 strategies to boost your mobile conversion rates
What is mobile conversion rate optimization?
As you’ll probably know, conversion rate optimization (CRO) is the practice of increasing the rate at which website visitors perform a desired action – otherwise known as a conversion – on your website.
The only difference between standard CRO and mobile CRO is that mobile CRO focuses solely on improving the conversion rates of mobile users.
There are many different ways to optimize mobile conversion rates, but our approach – and the approach most likely to achieve the best results – involves a number of key steps:
- Conducting research to identify optimization opportunities
- Creating new web pages based on this research
- A/b testing these newly designed pages
- Analysing tests to identify insights that can be used for the next round of experiments
In this way, we can work out what motivates our clients’ mobile users and how they differ from users of other device types. With this information we can then progressively optimize our clients’ mobile websites so that they generate as many conversions as possible.
For those interested, here are a couple of useful resources that outline our general approach to CRO:
- Our testing framework
- Our hypothesis framework
- Our prioritization framework
- 9 principles of experimentation
Why is mobile CRO important?
As discussed above, more people are using their mobile phones to access the internet than ever before, and yet mobile user experiences are still often absolutely terrible.
Take a look at some of these mobile UX no-no’s.
Sadly, examples of poor mobile UX can be found everywhere, which is why engagement rates on mobile websites are often so much worse than they are on desktop sites.
On the one hand, this is a shame: it means that mobile users are forced to battle through a never-ending series of poorly designed popups, interstitials, and web pages before they can get anything done.
But on the other hand, this presents a great opportunity for those of us willing to dedicate time to building strong mobile experiences!
In fact, with all of this low-hanging fruit, focussing on your website’s mobile conversion rate is likely to offer you a better return on investment than almost any other approach to CRO.
So with that said, now let’s move on to the strategies and techniques that we as an agency are using to optimize our clients’ mobile websites and drive up their mobile conversion rates.
12 strategies to boost your mobile conversion rates
Because of our testing-focussed approach to optimization, we’ve chosen to make the first handful of strategies in this section about how to research, run, and analyze your a/b tests to improve your mobile conversion rates.
In the second half, we’ll then go into more ‘best practice’ type strategies, focussing on changes and interventions that you can make that are likely to have a positive impact on your mobile conversion rates.
Note, though: every website is different, and blindly applying best practices to your web pages is rarely a good idea. Instead, we suggest applying some of these recommended changes and running a/b tests on them to see how they impact your conversion rates.
And with that out of the way, let’s get into it.
1. Mobile-specific research
Every good CRO process begins with research.
Research allows you to better understand your users and helps you build strong hypotheses that you can validate or invalidate with your a/b tests.
But what many don’t realise is that research also plays a central role in creating web experiences that are optimized for specific device types.
For example, whenever we do user testing, we always ensure that we include people on both mobile and desktop devices. This allows us to identify issues that are specific to each device type, and we can then use this information to create new versions of each web page that are thoughtfully tailored to each device.
Likewise, surveys provide another valuable source of device-specific research.
Whenever we run a survey on one of our clients’ websites, we make sure to segment the results based on device type. This allows us to spot device-specific issues and concerns, and we can then use these findings to build better mobile experiences on our clients’ websites.
Analytics tools like google or adobe offer a really useful way of identifying potential device-specific issues that might be holding back your mobile conversion rates.
For example, maybe the drop-off rate on one of your pages is significantly higher on mobile than desktop; or maybe your cart completion rate is high on tablet but poor on mobile.
To illustrate this point, here’s a screenshot from a typical analytics account, showing the engagement metrics for a given landing page.
As you can see, the mobile bounce rate is almost 40% higher than desktop, and average session duration is almost three times greater on desktop than mobile.
Analytics data of this kind helps you identify device-specific issues on your website. From there, you’re one step closer to diagnosing your problem (through research) and, finally, fixing it.
3. Segmenting experiment results
Following on from the above point, you can also use your analytics tool to segment the data from your a/b tests.
For example, you may find that one of your variations performs particularly well on one device but poorly on another. By conducting research to understand why this was the case, you’ll start to understand the key differences between mobile and desktop users, which you can use to optimize the mobile version of your website and ramp up your conversion rates.
Cautionary note: sometimes we’ll run a test that performs significantly better on one device type than another.
Consider the results from one of our recent experiments, shown below, where the desktop version of the variation performed significantly better than the mobile version. What do we do in this instance?
We basically have three choices here:
- We can roll out the new version on desktop but keep the control in place on mobile
- We can roll out the new version on all device types
- Or, we can keep the control in place for all devices
In this instance, there’s no one-size-fits-all solution. Instead, you’ll have to ask yourself a couple questions:
- Will implementing this new web page on one device but not on the other harm UX continuity between the two device types?
- Is the expected increase in revenue from applying this variation on only one device type significant enough to justify this kind of UX discontinuity across devices?
- Are there other iterations of this web page that can be tested that produce the same uplift without harming UX continuity?
- What is the traffic breakdown between the two device types? For example, if it’s 99% desktop, 1% mobile, you’ll be better off focussing on how the variation impacts desktop.
By answering these questions, you should have a better idea as to the best course of action.
4. Understand your mobile users’ typical environment
Sometimes you’ll find that different device types are used in different environments, and that this can impact the intent and purchasing behavior of your users.
A particularly pronounced example of this comes from our client Ironmongery Direct, the UK’s largest supplier of specialist ironmongery.
When we surveyed their website visitors, we unearthed an interesting finding:
- Most of their desktop users were at home when they used the website, ordering supplies for an upcoming job
- But when their mobile users used the website, they were usually already on a building site and were looking to quickly purchase something that they were missing
This difference in setting meant that mobile users were much more responsive to ‘urgency’ as a motivator than desktop users, and we’ve been able to use this insight to inform our overarching strategy for this program.
Many of the issues and mistakes highlighted above could have been avoided with a bit of basic quality assurance (QA).
A good QA process allows you to see how your web pages look on different devices and lets you catch mobile UX mistakes before you’ve rolled out your pages on your live website.
Before we release any new web pages into the wild, our QA team spends hours making sure that everything works on different device types.
Browserstack is a solid emulator that you can use to see how your website appears on different devices and models.
Worth noting, though: there are a number of limitations to using emulators rather than the real devices themselves.
If possible, we’d therefore recommend that you dig through your analytics to work out which mobile devices your users are using, and then actually purchasing – or at least borrowing – these devices to see how they render your website.
6. Page load speed
Having a website that loads quickly is important for all kinds of reasons. Fast loading web pages:
- Rank better on google
- Have lower bounce rates
- Convert better
But with the huge variability of mobile internet speeds, page load times are especially important when it comes to the mobile version of your website.
7. Click to call
According to a report commissioned by google, 70% of mobile searchers report using click-to-call website features, and ‘calls remain an important channel of purchase for consumers’.
Couple of things to note:
- This report and others like it are quite dated (released in 2013), so mobile purchasing behavior may have changed since then
- The importance of click-to-call will vary a lot from one website to the next. For some, it will be indispensable; for others, it may have no impact.
Ultimately, there’s no real way to know for certain if adding a click-to-call feature is likely to improve your mobile conversion rates.
That’s why, if you don’t already have this functionality on your mobile website, we recommend that you run it as an a/b test to see if it improves your conversion rate.
8. Minimise the number of pop ups
Anyone who has ever used a mobile device to navigate the web (i’m hoping all of you!) will be familiar with the following example:
You land on a website and are immediately accosted by a cookie consent message.
This cookie message, along with the message at the top of the screen urging you to download the company’s app, occupies about 80% of your screen space.
You’re just about to consent to the cookie message (what other choice do you have?), when another ill-fitting and unsolicited popup arrives on your screen offering you 30% off a product that you have no interest in purchasing.
In order to get to the website you have to get through the cookie consent pop-up and in order to get to that you need to get through the ‘30% off’ popup.
Now, if you’re really determined to visit this website, you may be willing to battle through this obstacle course of poor UX.
But if you’re new to the website (or if you’re in a bad mood!), you may well decide to return to the search results and find another website that offers a better UX.
From a mobile CRO perspective, this is a nightmare and something that you should be doing everything in your power to defend against.
Couple of quick rules:
- There’s not really any subtle way to place pop-ups on mobile sites, which is why it’s usually better to avoid them where at all possible.
- If you absolutely have to, make sure your pop ups are specifically designed for mobile devices. Pop-ups designed for desktop and thoughtlessly shoehorned onto the mobile version of your site will ruin both your mobile UX and your conversion rates.
9. Payment and checkout
When people are on desktop devices, they typically have access to a large screen, a keyboard, and even a mouse, which makes navigating clunky payment processes slightly less of a nuisance – though still a nuisance! – than it is on mobile.
On mobile, a clunky payment process will completely ruin your UX and play havoc with your conversion rate.
Thankfully, there are a couple of fairly simple fixes that you can apply to solve this.
First, we recommend that you integrate your website with different digital wallets, so that people can pay with things like google and apple pay.
This functionality will allow your users to complete their purchases in only a couple of taps.
And following on from this, we also recommend that you offer a single sign-on option, like the one displayed below, so that people can use their google or facebook accounts to sign up for your website.
Again, this will reduce the number of steps that it takes your user to complete their purchase, which will generally improve your conversion rate by reducing friction and increasing trust.
10. Include filtering and sorting options
Scrolling through long, unorganised lists of items is tedious at the best of times.
On small screens – the kind found on mobile devices – this kind of scrolling becomes unendurable.
Filtering and sorting options offer a good solution to this problem, allowing mobile users to find what they’re looking for much more easily.
If you only offer a handful of products, this probably isn’t going to be necessary, but if you’ve got a big product catalogue with lots of different product types, then it could have a big impact, improving both your mobile UX and your conversion rates.
In one of our recent experiments, simply by making the filter options easier to use, we were able to increase the page’s conversion rate by more than 13%.
11. Avoid excessive wordiness
Avoiding excessive wordiness is always a good idea, but when it comes to writing copy for mobile devices, this need for economy goes further.
In our experience (we’ve got a handful of experiments that back this point up), mobile users are less patient with wordy copy and tend to respond more positively to language that’s precise, clear, and to-the-point.
12. Responsive design…and beyond
Responsive web design ensures that your website works properly on all screen sizes and resolutions and that it renders well on all device types.
This is important for both CRO and SEO, ensuring that your website looks great on all devices and that users aren’t having to pinch-and-zoom whenever they want to interact with your web pages.
There are numerous case studies out there showing how responsive design improves conversion rates, but here’s the catch:
Websites that employ a responsive design don’t automatically become optimized for mobile. In fact, this is only the first step in the mobile optimization process, to be used alongside many of the strategies and techniques outlined above.
As will hopefully be clear by now, there are various paths you can take to improve your website’s mobile conversion rates.
What’s important to remember is that every website is different and no best practice is going to work for them all.
Ultimately, CRO for mobile, like all CRO, should be a gradual, cumulative process that involves running tests, gathering learnings, and applying those learnings to your next batch of tests.
In the long-run, if you’re serious about optimizing your mobile website to drive the maximum number of conversions, this is the way to do it.