Those who know me will be well aware of my belief in testing and analysis as the basis for targeted investment in improved retail performance – and the content question should, I believe, get the same treatment.
So, I’ve taken a look behind the scenes and dug into the data from some of the UK’s biggest high-street fashion and lifestyle brands – essentially to answer the question: “Does content really improve conversion?”
The answer, as it turns out is “Yes and no”. But before I expand on that, some context…
It’s undeniably true that the shift to mobile has hit conversion for most brands.
That is driven by three inter-related trends that are right there in the data.
First, there has been a dramatic shift to mobile over the last few years – the data tells us that tablet and mobile use (combined) moved from 40% of all sessions in 2013 to 68% in 2015.
Source: Biglight benchmark data (aggregated 2013 – 2015)
But what’s interesting here is the lack of any real session growth. Quite simply, people aren’t shopping more because of mobile; they are shopping differently.
Then there is the issue of mobile conversion. In general, conversion on tablet is lower than desktop for most brands and conversion on mobile falls to between 14% and 64% of that achieved on desktop.
Source: Biglight benchmark data, based on UK sales (anonymous retailers 2015)
Once again, the data reveals the impact of those trends.
Conversion rates were hit hard in 2014, before recovering in 2015, largely due to the implementation of mobile and responsive sites and subsequent optimisation efforts.
Source: Biglight benchmark data, based on UK sales (anonymous retailers 2013-15)
As this relentless shift to mobile continues though, retailers will face an uphill struggle to improve conversion – no surprise then, that conversion rate optimisation (and on mobile in particular) has become a high priority for most brands.
So there’s the context; but what does the data tell us about the role of content in that conversion optimisation struggle?
The good news is, there is a positive correlation between content and conversion
Back in 2014, L2 published research that sought to demonstrate a correlation between content quality and conversion.
I’ll refer to this as “rich content” for ease.
Source: L2 Inc – Content and Commerce, 2014.
L2 concluded that improvements in rich content were responsible for 50% of retailers’ conversion improvements and, for every five point increase in content score, conversion increased by 1%. You can read the full report here.
Looking at my data, I’ve also been able to demonstrate a correlation between engagement and conversion – Note: We did check that session durations correlated with page-views to rule out site performance /checkout issues.
Source: Biglight benchmark data, based on UK sites (anonymous retailers 2015)
What’s more, we’ve run a series of A/B tests to really understand the impact of rich content on conversion and AOV. In our most significant test to date, we found that:
- Users who interacted with rich-content were 20% more likely to purchase than those who didn’t.
- AOV was 22% higher for those who interacted with rich content before proceeding to purchase.
Clearly, these are very encouraging results, on closer inspection, they must carry two very important caveats:
- Rich content was presented to customers during their journey and within the same session, so it was highly visible.
- It related to the items being purchased, so it was highly relevant.
So the question is how can both of these objectives be achieved on a typical ecommerce website?
To understand that we must look at how effective the use of content is currently; crucially, whether it passes the visibility and relevance test – and, in all too many cases, the answer is ‘no’.
The truth is traditional content destinations simply do not engage users. For instance, only 25-35% of users see the homepage during their ecommerce journeys, so rich content featured or merchandised here is invisible to the majority of users.
In fact, we’ve found that engagement with rich content when it’s featured on the homepage (and category landing pages) is generally very low – conversion rates are actually improved if rich content is relegated or removed from these pages (but the brand people don’t like it).
But what about blogs, the historical home of rich content on ecommerce sites? Back in 2014 the L2 Research compared blog traffic volumes to that of the sites they support and found that engagement with them is poor.
We’ve found that traffic to blogs is significantly lower that to other areas of the site that should be comparable and they suffer from exit rates that are up to three times the site average, even where blogs feature in the main navigation and are merchandised on the homepage.
That’s not exactly what we are all trying to achieve and not popular with trading teams.
So, if we know that making relevant content visible to users during their journeys to drive engagement and conversion and we can pretty much rule out the homepage, key landing pages and the blog, where should we be looking?
Well the answer is in the data; 50% of users now start their ecommerce journeys on a product listings page or product details page – that’s where the opportunities lie.
Source: Biglight benchmark data (aggregated 2015)
This all has some fairly big implications for site design and content optimisation
The headline here is that content can positively affect conversion and AOV – but blindly throwing money at content without really understanding where it is best used is to trust to luck.
Chance dictates that some content will be in the right place, but those fortunate retailers will not know why and will still be wasting time and money on content that barely anyone sees.
The solution is two fold.
First, retailers need to move away from rich content destinations and content merchandising to create content elements that are featured or “threaded” through the pages that make up the user journey – principally the product listings pages, but also the product details pages.
Second, they must test, test and test again with real users and A/B testing tools – then let the data tell them what works, and do more of it. That means looking at every type of content to understand its impact in different contexts:
1. Brand heritage content that’s true for ever
2. Seasonal content that drives core merchandising messages
3. Short term “now” content that creates relevance
4. Social validation content – user-curation, ratings and review
In short then, content really can be at least part of the solution to lacklustre conversion rates – but only content that is delivered with purpose and focus; content that is both visible and relevant and whose performance is understood and optimised through exhaustive testing.