What’s the deal with personalization?
Personalization. So many e-commerce trend reports focus on personalization as the buzz-word and new product that will redefine e-commerce - and they’re not wrong. These days everything is personalized, from the marketing you receive in your email to the products shown to you when you visit a site . It’s become the thing that consumers don’t demand, but that they expect. As a company, if you don’t have it, you’re already behind.
But the thing about personalization is people’s expectations are very high, and you have one chance to get it right. Netflix knows exactly what you want to watch next, but TV is less personal than clothing and less permanent. It’s also less likely to make a judgement on someone’s sense of style. When it comes to style, you better get it right when you tell me “I may like this”.
The challenge of personalization in e-commerce is to feel as if you’re shopping with a personal stylist who knows you and your style intimately. They know what’s in your closet, what silhouettes you like, what you dislike, what will make you look good. The personalization shouldn’t require extra work outside of a “getting to know you” session. Shopping should feel easy, where the overwhelming feeling disappears, and you’re presented with things that uniquely suit you. Personalization should turn online shopping into a joy.
Some of the things companies can use to tailor their personalization tactics:
Segmentation Studies
There are a lot of companies out there selling their “segmentation” technology, where they have tech finding patterns in users in order to generate segments. But I’m not talking about the segmentations that take consumer purchase behavior and split them up based on spending habits and whether they clicked on an email. I’m talking about the ones that incorporate psychographics into them. Because you need to know the “why” behind the actions. WHY are they choosing to spend, why do they shop the way they do? What triggers a purchase? You may have two identical groups of people based on their habits, but when you take their thoughts and feelings into the equation, you actually have two completely distinct groups.
Style Groups
Now this isn’t something that exists currently (that I’m aware of), but it’s something I developed for Ochre. Fashion, style and how people feel about brands and silhouettes is incredibly personal. It’s something that is constantly changing and that an algorithm couldn’t learn all by itself without bucketloads of personal data. Our Ochre algorithm takes into consideration the personal associations people have with brands.For example, a t-shirt by brand X is not the same as a t-shirt by brand Y regardless of how identical they are. Yet what the average algorithm sees is a white t-shirt within the right price point and shape. But if brand X isn’t within the style segment of the consumer or even in a related segment, they won’t ever consider it and it’s a failed recommendation.
Trend Research
Most companies spend thousands of dollars each year employing independent trend forecasters to help them decide what to produce/stock/purchase for the next season. Yet somehow these forecasts don’t always trickle down to the recommendation and personalization section. A tip: Keep your recommendations current. And if certain things are trending, make sure to highlight those things and keep them in focus. A company that does this well is The Real Real. Perhaps it’s because their product assortment is older and they’re forced to constantly reinvent, but their team of editors is expert at putting the focus on pieces that are relevant. But that being said, their automated “Picks for You” could use a bit of fine tuning, to put it lightly. I guess you can’t win ‘em all?
Off Buttons
Don’t you hate when something is recommended to you and you wish you could turn it off? Or tell them why you hate it so it (and things like it) stop following you around? Imagine if I could tell a computer that I thoroughly disliked the color orange and never have to see orange things ever again. Or that I already bought the style of black boots chasing me around, but please could they recommend some pants instead? Oh the joy and the possibilities!