October 22, 2015
Case Study

B2C Marketing: Site and search optimization leads to 15% increase in average order value

SUMMARY: It’s difficult for online stores to match the level of one-on-one attention of their brick-and-mortar counterparts. That was the challenge Sourav Sharma, Head of Ecommerce and Online Marketing, Café Britt, encountered when it came to the specialty coffee and chocolate B2C company.

However, by adding a visual element to its internal search engine and including a sophisticated recommendation engine to the site, Café Britt was able to overcome this obstacle. Learn how this effort helped to increase customers’ average order value by 15%.
by Kayla Cobb, Reporter


Café Britt is an international brand that sells gourmet coffee, fine chocolate and specialty gift items. Currently the brand features coffee from Costa Rica, Peru and Columbia. At the time of publication, Café Britt had 92 gift stores in eight countries as well as an online store.

"We also have another division, which is travel retail, which is the biggest, where we have close to 140 stores across 13 different countries. Those are primarily at airports and hotels," Sourav Sharma, Head of Ecommerce and Online Marketing, Café Britt, said.

First and foremost, the brand's typical customer values a high-quality coffee. The brand sees both male and female customers who are typically 35 years or older. However, this age range is more flexible when it comes to Café Britt's online store.

"Of course, you have a profile for an ideal customer that you definitely need who definitely is buying with you. But Web allows you that kind of flexibility to stretch and go after the Millennials and also an older group of people as well," Sharma said.


Because gourmet coffee is a specialty product, it was important for the buyer's journey to be more personalized than for the average product. While Café Britt was able to achieve that level of one-on-one interaction at its brick-and-mortar locations, that experience was missing from the Web.

Specifically, the team at Café Britt wanted to make finding relevant products easier for first-time visitors on the site, Sharma said. However, they also wanted to better the experience for the company's regular customers.

"And our regular customers, they should feel we are taking care of them and providing them the best personal experience," he said. "They want to see things first that they really want to see."

With this focus on bettering the ecommerce customer's experience in mind, the team at Café Britt worked with its vendor to implement a site personalization plan.


In the summer of 2014, Café Britt started working with its vendor on this site personalization campaign.

Since this campaign affected the ecommerce site directly, Sharma said that the team wanted to achieve an increase in revenue, but the biggest goal for the team was increasing the site's engagement and conversion rates. The team also wanted to increase the amount of regular and returning Café Britt customers.

"That can only be possible if we are giving [customers] relevant experiences and making them feel they are finding things easily, we are taking care of them, we are thinking about them and providing them what they need," Sharma said.

"That was the goal we had in mind. I'm glad we are succeeding in that," he added.

Learn how this ecommerce personalization effort led to an 18% increase in conversion rates and a 15% increase in order size for Café Britt.

Step #1. Improve site search

After teaming up with its vendor, the first aspect the Café Britt team focused on was improving the site's search quality. This was accomplished by adding a visual component to the site's search engine.

According to Sharma, "when you talk about search, we live in the world of Google where it needs to be precise, it needs to provide relevant information and it needs to adapt."

Before this implementation of visual elements, if a customer searched for a particular product on the site, they would be taken to a new page where they would see a text list of products.

"There was no visual appeal to that, and being ecommerce, we knew that we were [making] that mistake," he said.

The team changed the search engine so, as the customer typed, a helpful image of the potential products they were searching for would appear.

View the Creative Sample

Click here to see the full version of this creative sample

This system was designed to be easy to use on the backend so that Café Britt's changing line of products can be quickly added to this search function. Sharma largely credits the ease-of-use of this backend to Café Britt's vendor.

The team quickly received positive feedback on this new feature.

"Thankfully, we have a pool of great, loyal customers, and we started getting some good feedback from them as well. 'Hey, listen, you added this new feature and it looks beautiful. It helps us find more information easily.' So we knew we were in the right direction with this," he said.

Step #2. Create segmented customer profiles

After the site's search was optimized, the Café Britt team turned their attention to a site recommendation engine. This measure was taken to give the ecommerce experience a more one-on-one feeling.

This complicated system was implemented to make recommendations based on customer behaviors as soon as they land on the site. The system starts by using the customer's geolocation and what browser they're coming from to construct a starting profile for that customer. This profile also takes into account other sessions that customer has had on the site for returning users. For new users, it utilizes products other users who match the user's geolocation and browser preference have ordered.

"If a person comes from this location, most of the time with this kind of browser or this kind of OS, operating system, he tends to go towards this. So that's our first level at the start," Sharma said.

The recommendation system was designed to adjust constantly based on how the user behaves.

"The recommendation engine adapts, so we have a widget that tells you — actually on the homepage — that tells you when you land what is recommended for you," he said.

View the Creative Sample

Click here to see the full version of this creative sample

"In coffees, you have different kinds of coffees. You can have light roast and a dark coffee, an espresso, an organic. So the moment we see [the customer] navigating our website, it is constantly teaching us what he might like," he said.

The system also assists customers who may not know exactly what they want.

"Say for example, if he adds a French press into the cart … we might recommend there's another accessory that might complement your French press," Sharma said. "That's how this whole engine works, and that's how we do the profiling, so it's very dynamic. It's not static at all, and that's the best part of the solution."

Step #3. Incorporate cross-selling into recommendation engine

Once the recommendation engine was put into place, it wasn't only utilized to make product-specific recommendations. Cross-selling was incorporated into recommendations as well.

To incorporate cross-selling, users were intentionally segmented at multiple levels, allowing users to fall into multiple segments.

"The segmentation can be done at multiple levels. One is on the category level — you can do that. Then within the category, the product types, and one user can be in multiple segments," Sharma said. "So that's the reason when we are recommending him or her a product again based on, if it's a returning customer, it's based on his past behavior."

According to Sharma, this new system has helped in educating both new and returning customers on all the products Café Britt has to offer. For example, a customer who exclusively buys coffee from Café Britt can learn for the first time that the brand also produces chocolates, which can then impact their buying habits.

The success of the cross-selling recommendations is tied in part of the site's ratings and reviews section for its products.

View the Creative Sample

Click here to see the full version of this creative sample

"Several customers have added [the] comment, 'I didn't know you had some nice gifts,' because those customers were old customers who have been buying for the last 20 to 25 years, and sometimes they just call and say, 'Hey, can you just repeat that order?' because we have a nice subscription program too," Sharma said.

That was the reason, he added, the company was able to divert customers and also sell products in related categories, which hadn't happened previously.


From this site customization effort, Sharma and his team were able to achieve:
  • Almost an 18% boost to the site's conversion rate

  • An increase in average order value by 15%

However, the biggest accomplishment for the team was being able to give Café Britt's regular customers a better and happier buying experience. Customer feedback is very important to the Café Britt team.

"A good example — one my customers sent an email that said, 'Listen, now I can find my product that I order on your homepage. Earlier I had to go the gourmet coffee category, and then go within the region like Costa Rica and click on the product.' So there were three clicks involved," Sharma said.

"Now based on his behavior, when he goes to this homepage, he sees that product right away. So, that is one of the things that has been amazing for us. We see that people are engaging more," he said.

Based on the success Café Britt has seen on the desktop front, the company is considering expanding its personalized ecommerce experience to mobile.

When asked what advice he would give to marketers who are considering launching a personalization campaign, Sharma recommended starting small.

"Don't try to implement everything at the get-go. Learn from it and keep improving little by little, and then go all-in," he said. "That has been probably the biggest reason why we were successful. Because if we had implemented throughout, then we wouldn't have learned so many things in the process."

He also focused on the importance of listening to your customers.

"When we did this [campaign], to be honest, we weren't sure what to expect or how much to expect. But this has continued to impress us in so many ways and we are happy with the results," he said. "Now, we probably will go to the next level and go more deep."

Creative Samples

  1. Search image

  2. Recommendation engine

  3. Product ratings and reviews


Café Britt

Reflektion (Café Britt's vendor)

Related Resources

MarketingSherpa Summit 2016 — At the Bellagio in Las Vegas, February 22-24

B2C Marketing: Senior living provider drives a 12% lift in online leads with questionnaire

Landing Page Optimization: Testing location-based pages increases conversion 88%

Search Engine Optimization: Website overhaul increases revenue tracked to organic traffic 25%

Ecommerce Chart: The most effective types of personalized product recommendations

Content Marketing: Dynamic content recommendations increase time on site 70%

Improve Your Marketing

Join our thousands of weekly case study readers.

Enter your email below to receive MarketingSherpa news, updates, and promotions:

Note: Already a subscriber? Want to add a subscription?
Click Here to Manage Subscriptions