Kopari’s revenue per visitor has increased 60% in 15 months by personalizing the site based on the specific shopper, down to the product description she reads or the products featured in a photo.

Cosmetics retailer Kopari is obsessed with testing everything, says Megan Whitman, the retailer’s chief digital officer.

Since KopariBeauty.com first launched in 2015, the coconut products retailer tested different social media ads to see which messages would drive consumers to the site and then purchase. Once Kopari built up a good amount of traffic on the site, it would A-B test different website layouts and descriptions to see which would most encourage shoppers to purchase, Whitman says.

Kopari text messagesIn January 2017, Kopari decided to take its A-B tests to a new level and started working with personalization vendor Dynamic Yield Ltd. Previously, Kopari used vendor Optimizely for its A-B tests, but decided to switch to Dynamic Yield because of the availability of tools and functionality without added fees, Whitman says. As a small business, Whitman felt she would get more support from Dynamic Yield, she says.

When Kopari first started using Dynamic Yield, its focus was on the best homepage layouts and product page layouts. Now that the retailer has determined those, Kopari has turned its focus to personalizing those pages even more based on how the shopper came to the website, where she is in the purchasing funnel and what device the shopper is on, Whitman says.

For example, if the shopper has never visited KopariBeauty.com, she may see more images and messages that display high-level branding about Kopari, such as a message about how its products are coconut-based and all natural, whereas a returning shopper will likely see a hero image featuring a product she has previously browsed or bought.


Currently, Kopari is heavily marketing its deodorant product, so a new shopper may see an image and messaging on the homepage about that. However, returning shoppers have heard about this product in email marketing campaigns for a few weeks now, so the retailer may choose to feature a different product that the shopper may never have heard of instead, Whitman says.

Dynamic Yield uses cookies to determine if a shopper has visited the site before or not, even if the shopper is not logged in or a previous customer, according to a Dynamic Yield spokeswoman.

Another way Kopari personalizes the site is based on the ads a shopper sees. For example, one Kopari social media ad highlights the ingredients in a product, whereas a different ad focuses on the results a shopper receives from using a certain cosmetic product. Depending on which ad the shopper clicks on, Kopari will change the product’s description on the product detail page to focus on what the ad talked about.

“If they clicked on the ad, we’re assuming they saw something they liked, and we’re reiterating that message,” Whitman says.


For the device type, Kopari noticed that certain layouts worked well for desktop users but not for mobile users. For example, a shopper has a product in her cart but has not yet checked out. When that shopper returns to KopariBeauty.com on a desktop, her cart will slide out on the screen, so the shopper can easily pick up where she left off. On mobile, however, that popup reminder about her cart covered the whole screen, and shoppers found it to be more of distraction than a gentle reminder. Therefore, Kopari does not have this feature enabled for mobile shoppers, Whitman says.

These personalized tweaks are paying off: Kopari’s revenue per visitor has increased 60% in the 15 months since using Dynamic Yield, Whitman says. Revenue per visitor is a measure of sales on the site divided by unique visitors.

Kopari uses this as its measure of success because it’s all encompassing, Whitman says. For example, for a while Kopari was testing its free shipping threshold. With a $60 free shipping threshold, average order values may increase because people are buying more to reach $60, but conversion rate may be lower because not all shoppers will do that. The inverse was true for when Kopari tested a lower free shipping threshold—the site had more conversion with smaller baskets.


Because of the variability in these metrics, Kopari decided that revenue per visitor was the best way to determine personalization’s effectiveness, she says.

However, when looking just at conversion rate, that number is looking bright. Kopari’s conversion rate has increased 30% year-to-date compared with January-September 2017, Whitman says.

“That’s not something that happens overnight,” she says. “It’s from constantly iterating and constantly testing and drilling in on personalization month over month.”

Kopari pays Dynamic Yield a quarterly fee based on its traffic. Whitman wouldn’t reveal costs, but she says the return on investment is “large,” and in Q1 2018, the revenue per visitor Kopari generated from using Dynamic Yield is about five times what it cost.


Next up, the retailer is testing messages to get shoppers to subscribe to a product. For example, a shopper can purchase the product once, or purchase it on a recurring basis, and Kopari will ship the next product after a few weeks or months.

Describing what the subscription is and how much it costs, however, takes a lot of text in small amount of space, Whitman says. Kopari is testing a few options but doesn’t have a clear winner yet.

Kopari also wants to further personalize the site for its loyalty program members, for example, by making it easier to find out how many points she has and what she can use them toward, Whitman says.


“For our highly engaged customers who are already subscribers or part of our loyalty program, we make the site very personalized and easy to use so they continue to come back, and shop with us and not necessarily at a third party,” Whitman says.

Kopari also sells its products at several retail stores and websites, including Sephora USA Inc. (owned by LVMH, No. 131 in  Internet Retailer 2018 Top 500), Ulta (No. 91), Anthropologie (owned by Urban Outfitters Inc., No. 39) and Nordstrom Inc. (No. 16).