The right products at the right time: Product personalisation is one of the things that most often comes to mind for many people when we talk about personalisation.
In its most simple form, we see this in product carousels that can be placed in different areas of a website or in an email. The carousel will contain a set of products that an algorithm has determined that a customer may have an interest in, based on data about the customer that has been collected. This may be as simple as basic gender and geographic information, what they have previously purchased. It may also include data on which products they look at, the content they consume, and even what they post about the brand on social media.
No matter how much data they use and how sophisticated their personalisation engine is, brands will only see a limited uplift using product carousels alone. Most carousels will contain up to 10 different products and the likelihood of one of these products being the one the customer wants to buy at that point in time is relatively low, even if curated by an algorithm. It is certainly worth doing but, if it is the only personalisation that is done, it's not going to change the world.
A step further in product personalisation is where search results, including catalogue browsing, is personalised.
This means that product listings and search result pages are personalised, ensuring that the products that the personalisation algorithm determines that the customer has an affinity to are pushed upwards in the pages or that a different set of rules are used when determining the order of products shown on these pages.
This technique is less specific than a product carousel as it is effectively adding weightings to certain types of products to ensure that they are pushed further towards the top of pages. However, a retailer utilising this technique can expect to see a larger uplift than when introducing product carousels as the customer will consistently see products that they have a closer affinity to more often.
Seeing the right Content
Many retailers are using rich content to give them an edge over their competitors. E-commerce is not just about a product catalogue, it is about an experience and rich content can often be an important part of that experience.
Content is another area where I see a lot of brands personalise. For example, it is very common for a brand to have a large hero banner on their home page; often pointing to a particular category or product range. Many brands will now personalise this by creating multiple copies of a banner and showing the most relevant one to each customer based on the data that has been collected about the customer.
Therefore it makes sense to show customers the content that is most relevant to them rather than just showing the same thing to everyone. This technique has been proven to increase KPIs such as conversion rates and is almost always worth doing but, on its own, it is not going to have a dramatic impact.
Let's think of a very simple scenario where a multi-category retailer sells clothing as well as homeware. While they will have customers who purchase from multiple categories, they will also have others that will primarily purchase from a single category. Most brands will decide which category to display first within the main navigation based on many factors, often focused on which product category generates the most overall revenue, most often, same for everyone who visits the website. However, if they know that I mainly purchase from the clothing category, there is a strong argument to show that category to me first in the navigation. This, alone, is not going to have a dramatic effect on KPIs but, combined with other changes to the experience, it can start to make a big difference.
Another, more sophisticated, example could be changing the order of the faceted navigation (filter options that sit on the product listing and search results pages) based on a user's behaviour. Maybe a specific user uses the price slider much more often than they filter by colour. If so, the retailer should consider pushing the price slider towards the top of the faceted navigation. Many brands will have a large number of facets that a customer can filter on but will always show the same facets in the same order to every customer. Why not show the facets that each individual uses most at the top? You could do the same for the sort options or even automatically show a certain number of products in each row on the product listing page based on how each individual has previously interacted with the website.
How about delivery preferences? If a user most often chooses click and collect, show that option first ahead of other delivery methods. The same goes for payment methods. If they always use PayPal, make that prominent and the first choice. When you start to think about it, there are many distinct parts of the experience that could be altered to perfectly match an individual.
How can it be done
Some of the best tools a brand can use to deliver a holistic personalised experience are ones that started life as A/B testing tools but have evolved into experience personalisation platforms. Tools like Dynamic Yield, Monetate and Optimizely work by intercepting the HTML before it is loaded by a user's browser, manipulating it based on certain rules and then showing the altered experience to the user. This technology allows these tools to change almost any aspect of the website, from the products, the content, the navigation and almost any part of the experience.
A brand can test multiple different experiences at once and measure which delivers the best results.
When they were first developed, these tools knew very little about the individual users and would randomly select which users were show which experience based on the percentages the brand would configure. However, over time, these tools have evolved into personalisation platforms where they gather data on each user's behaviour; which products and categories they view, what content they consume, what they search for, what they buy and many other aspects of the user’s behaviour. They then user artificial intelligence and machine learning to calculate a user's affinity to a certain category, product, or any other aspect the brand may want to consider. The system will then dynamically segment customers based on this behaviour and affinity and the brand is then able to define which segment will see which experience.
It is even possible to combine this with data from other touchpoints such as product returns, social interaction and even purchases from physical stores.
All of this data can be used to provide a 360-degree view of the customer and can be further used to provide personalised experiences such as ensuring online and email product recommendations take in-store into account.
Using this technology will allow a brand to deliver an online experience that is wholly personalised to each customer, putting the right products content and overall experience in front of everyone; having the maximum positive impact on KPIs.