5 ecommerce personalization insights for 2021
We studied 1.5 billion customer journeys to help businesses unlock opportunities that improve conversions rates on their retail sites. In this post, we’ll share the top five insights we’ve uncovered to inform your ecommerce personalization strategy this year. You’ll learn why you’re competing for your visitors’ attention, how irrelevant most homepages are and why you should use your long-tail products more effectively.
What we learned from studying 3.3 billion sessions, 1.5 billion customer journeys, 1 million products,, 30,000 categories and 173,000 purchases.
1. Attention is the new form of currency.
Consumers are bombarded with information from every angle. it’s been found that humans now have attention spans shorter than goldfish. While not as drastic, Netflix, a business synonymous with making data driven decisions is well aware of the challenge of keeping their users’ attention. A typical Netflix member loses interest after 60 to 90 seconds of choosing…on one or two screens.”
Our research shows similarly striking numbers, with the drop in attention spans primarily due to the rapid adoption of mobile browsing and the endless distractions available to people on their phones.
For e-commerce teams, this means that they only have 1-2 page views to engage visitors – to grab their attention, be relevant, or they’re gone and moved on to the next best thing – a competitor site, a text chain. Being relevant is a must for engaging visitors, as it’s the only way to successfully compete and win their attention.
2. The average ecommerce retailer has 13,000 skus to manage.
This is far too many skus for even large, robust ecommerce teams to manage on their digital store-front. 13,000 products means so much product data on views, prices, purchases, inventory, sell-thru, returns, the list goes on and on. With so much to make senses of, all that product data often gets lost in the background. Swimming in spreadsheets, racing to hit targets, busy ecommerce teams often end up focusing their efforts on a small subset of products – with heroes and promotional products representing a large chunk of these.
However, if retailers take another look at the status quo, or dipped their toes back into the data, they’ll realize that all that product data is their customer data. Behind the rows in those endless spreadsheets are people, making decisions – choosing whether they feel inspired, whether to buy from your site or to abandon cart and look elsewhere.
3. 70% of revenue comes from longtail products
Hero products will always be core to retailers – but focusing efforts on hero products doesn’t guarantee success or lead to efficiency – focusing exclusively on heroes would still mean managing nearly 4,000 products.
4. Hero products are relevant to only 16% of visitors.
If 70% of revenue already comes from products that are not actively merchandised, retailers are missing out on huge opportunities to boost revenue within their product long-tail. But to effectively do this, ecommerce teams need better tools and technology that can help make sense of all the product and customer data and dynamically promote the right products to the right visitors.
Using our AI capabilities we have estimated the relevance of heroes across a large sample of our customers. We found out that, most of the time, hero products are far from relevant, despite being pushed the most. If customer preference was taken into account homepage relevance would jump from 16% up to 42%, a whopping 3.2x boost. Relevance matters – it matters to your customers in those fleeting one to two page views. And since it can mean the difference between a bounce or a sale, it certainly matters to retailers and their top line revenue figures.
5. Most product recommendations aren’t as personalized as they say they are
Most product recommendations engines use out of date technology to power their ‘smarts.’ These traditional recs systems build product to product correlations rather than understanding the customer to product relationships. The result is that all visitors see more or less, very similar product selections despite their unique and varying tastes and preferences.
Customer to product relationships develop as the customer takes a journey through your product offering. Product recommendations that are powered by sequential deep learning are what make it possible to scale on retail sites. This shift in technology puts the most relevant product selection in front of your customers every step of the way. Because every second, every interaction and every page view matters.
At Qubit we always start with the data, this enables us to have a laser-focused approach and gives us higher success rates for our customers. Our research shows that your customers need more relevant experiences. Otherwise you’ll lose them to the competition or the next thing competing for their attention. More can be made of your product catalog, especially the long tail, but you need up-to-date technology to effectively operationalize and see the impact. Most of the time product recommendations tools are outdated and not fit for purpose, sequential AI is the next leap. Download the ebook – Vuja De: The Personalized Merchandising Shake Up to find out how you can stay ahead of the game