If you’re walking through a store at the mall with your best friend, and she pulls a shirt off the rack, puts it in front of you and says, “this will look great on you!”, she’s probably done a better sales job than any sales associate on that showroom floor. Without a doubt, you’re more likely to buy that shirt now than if you’d been to the store alone, or even if you’d been heading to that store intent on buying a new shirt.
The fact is, when someone we know provides an endorsement of a product or service, 80% of our normal psychological barriers to the purchase are torn down. That recommendation automatically tells us the item is valuable, attractive, and personally beneficial to us.
Having your best friend with you at an online store
Online shopping is a little different, of course, because we tend to do it alone. Unlike walking through the mall, online shopping lends itself to a more thorough decision-making process as customers compare prices and shipping details to isolate the very best option.
The introduction of social media several years ago inserted a level of camaraderie to the online shopping world and retailers were quick to take advantage of the opportunity. Now, it’s common for people to first see products on Pinterest or Facebook, where a friend’s “like” starts to break down those psychological barriers for us.
But social media will likely never live up to the value of a real person standing next to us at the shirt rack, primarily because it lacks immediacy. So, your friend’s opinion isn’t going to necessarily affect your purchasing decision.
What about personalized product recommendations online?
Online tools that help stores create that social feel for personalized product recommendations, generated not by people, but by machine learning algorithms. Basically, the recommendation engine uses data collected from many different interactions the visitor has already had with the site. For instance, the keywords they search for, the specific links they click, the links they ignore, the products they already have in their shopping cart or wish list – all of these go into the “secret sauce” of the recommendation algorithm. Imagine this scenario involving an online clothing store:
You click on a Pinterest pin that shows a gorgeous skirt at GorgeousSkirts.com. You click through to the site and wind up on the product page for that skirt. Now that you can see it a little closer and you find out what material it’s made from, you decide it’s not right for you. But the style is still really nice, so you notice another similarly-styled skirt recommended at the bottom of the page, and you click on it.
Now, you’ve only been on GorgeousSkirts.com for a few moments, and it only really knows one thing about you: you clicked on a pin with a picture of a particular style/color/size skirt on it. So, the recommendation engine they use analyzed the data it had and presented you with other products that were similar.
By clicking on that recommended link and looking at a new product, you’re giving the recommender more information. While the other skirt was light green, this one is deep blue and it’s just a bit shorter. As you’re absorbing the details of this new skirt, the personalized product recommendations include skirts that are a bit shorter and in darker colors. Two of them catch your eye.
As you continue clicking through to different products, perhaps search for an item or two, and either choose or ignore recommendations, the engine gets smarter and smarter. Finally, it hones in on exactly what you’re looking for: that beautiful royal blue knee-length skirt that’s going to be the perfect match for the shoes you bought last week.
So, you buy it.
That’s how product recommendation engines like the Strands Retail help online retailers make sales, and customers find exactly what they want. If you’re an online retailer and you’re not taking advantage of the next best thing to your customer’s best friend, you need to consider it.