9th October 2015 -

Generalisation vs Personalization

Having recently spoken with many retail experts, I found one of the key discussion points to be around what differentiates one personalization software from another.

There are many companies today offering varying levels of ‘personalization’ from basic on-site pop-ups to highly intelligent behavioral profiling and merchandising but, as more and more companies are now adopting personalization the question is ‘are they investing in the right solutions?’

Here’s the thing, without behavioral profiling true personalization simply cannot be realized. This means the way to achieve Amazon’s level of personalization is by segmenting customers into groups with other extremely similar customers, not just falling into the trap of generalizing by presenting trends from every one of your visitors.

Let’s dig a bit deeper…

Firstly what do I mean by generalization?

Generalization gives the appearance of personalization but in actual fact for me is the complete opposite of true personalization. A site that offers a more general approach would typically make its recommendations based on general customer data showing recommendations such as “best seller” or “people who bought this also bought this” and these recommendations are all based around product attributions, along with linking products together based on overall behavior rather than the individual.

Personalization presents much more relevant results as it is based on a specific behavioral profile. The system knows the visitor on your website often purchases cameras and has a high average order value, therefore the recommenders around your website will show high-value camera related items.

What should you be aiming towards when choosing a personalization solution?

To answer that I need you to think about a time when you left a shop having received a great experience. You probably engaged with the sales assistant, you probably then discussed your requirements with the sales assistant perhaps probing some more. Finally, you were probably recommended exactly the right item to suit your needs along with a couple of extras that you had never considered purchasing originally but perfectly complemented your main purchase.

Now; take yourself back to a time when you received a poor in-store experience. You were probably approached by a sales assistant who had no intention of listening to your exact requirements but instead tried to recommend a few items that had just arrived in store and the most popular item sold that day. Rather than seeing and purchasing the product you entered the store for, it is likely you became frustrated and/or overwhelmed and left the store without making a purchase. Just because the majority of the population purchased an iPhone, doesn’t mean that everyone wants one, right?

So true personalization is the delivery of a great individual experience on your ecommerce site. The type of great experience that will make you want to move right through the customer journey and on to place an order. To translate this great experience to your ecommerce site you need to start placing all of your customers into behavioral profiles based on how they interact with your brand, what products they like, share, buy, view and how other customers who are similar to them act.

Often customers need to be placed into multiple behavioral profiles, this simply enables an ever more tailored experience. Once you gather even more data on the customer then you can place them into more defined profiles and so on right until we can conquer that customer and target them with complete relevance.

Once a customer is placed into a specific behavioral profile you can then personalize the look and feel of certain pages by placing personalized pop-ups or banners showing specific promotions to that individual. Admittedly there is only so much the naked eye can spot so PureClarity will analyze the mountains of data, spot trends for you and notify you when a new profile has been found. This frees up your time allowing you to focus your marketing efforts on attracting more, relevant traffic and new customers.

When looking at placing page recommenders you have to consider the customer journey and focus on areas where customers may typically leave or ‘drop-off’ the journey. To move away from generalization you need to place recommendations that are relevant specifically to that person and these should be based on a combination of two things:

1. The past activity of that individual.
2. The past activity and trends of other people who have similar interests of that individual.

You, therefore, move away from that generalization model of recommending “top trending items” and move into very personalized recommendations by displaying “people like you who viewed this item also went on to view”. I would also make sure these recommendations sit in the right places and typically place them in all landing pages, product list pages and most definitely the basket area to give you the best chance of increasing your conversion rate and delivering that great customer experience.