With a wide variety of customers, Bloomingdale’s marketing team had to find a way to personalize their email marketing campaigns to customers’ needs and behaviors.
Read below how the brand was able to integrate behavioral marketing into its email program for a 50% increase in revenue driven by triggered campaigns.
An upscale contemporary department store with locations across the United States, Bloomingdale's has a wide variety of customers.
Sean Duffy, Operating Vice President of Customer Loyalty and Contact Strategy, Bloomingdale's, is part of the brand’s marketing department and heads up the customer loyalty team, serving those customers. He oversees the loyalty program — Loyallist — and oversees content strategy, which covers email marketing, direct mail and mobile marketing.
Because Bloomingdale’s carries an array of merchandise at different price points, the customer base is varied, but does tend to skew female, he said.
“[The Bloomingdale’s woman] tends to have a slightly more elevated household income and definitely has more of that contemporary style,” he said. “We try to give [her] the … inspiration-base on the new trends that we're seeing out there so she can make sure to update her wardrobe accordingly.”
Most of the email marketing Bloomingdale’s had previously done with customers was closer to batch-and-blast campaigns.
“[The emails] were speaking to trends or marketing offers or initiatives that we were controlling centrally and then based on the different messages we had,” Duffy said. “We would decide which segments of customers to send each to, but those campaigns were not necessarily directly related to the customer's immediate behavior.”
While the team would look at shopping and browsing history to determine who should receive certain messaging, they didn’t have any triggered email sends based on behavior. They also sent out transaction-related emails with order and shipping confirmations.
The clickthrough on those emails was usually fairly strong, Duffy said, and even generated a substantial amount of demand.
“Which you wouldn't ordinarily expect because the customers just bought something, but a lot of customers were engaging with those emails and going back to the site to buy something else,” he said.
Sending those transactional emails based off of customer purchases suggested that it would be beneficial to trigger emails based on additional customer behaviors.
“Marketing emails that are triggered based on products the customer has engaged with but hasn't yet purchased to encourage them to really pull it over the finish line and make that sale was very appealing to us,” he said.
This campaign “really helped us get more personalized in the marketing messages that we were speaking to our customers about — just getting closer to that one-to-one personalization, really talking to her about what's relevant to her,” Duffy said.
Starting in early 2015, Bloomingdale’s committed to this campaign and established an overall goal to drive online sales.
“But through that, we also were very interested in the typical KPIs that we look at for our emails, which are open rate, clickthrough rate,” he said. “From there, once they get to the site, how many of those customers actually convert and buy something?”
The team began developing campaigns such as price reduction alerts, sale previews and post-purchase follow-ups.
Step #1. Walk before you run
This campaign began as “a little bit of like a crawl-walk-run, where at first we wanted to start with some of the most obvious trigger campaigns that you see out there in the ecommerce space a lot,” Duffy said.
For the team, that meant an abandoned checkout and abandoned bag campaign, with the email focused on remarketing to customers to remind them about products they added to their bag, or the checkout process they started but did not complete.
“We really worked with [our vendor] to figure that right way to go about that. Before working with them, we might not have known anything or realized anything about what they call drips,” he said. “We would have thought, ‘Okay, you abandoned the bag in the checkout. Let's send you one email about it.’”
Duffy and his team worked with the vendor to develop a drip campaign that had the first drip going out within a couple hours of the abandonment and a second drip going out 24 hours later.
“We started with that as what we saw as the biggest opportunity in terms of the trigger campaign. And as we saw the success in the campaign, we brainstormed internally and with the [vendor] to come up with other types of triggers that we could leverage,” he said.
That lead them to develop browse abandonment campaigns, through which customers get emails about products they’ve browsed more than once, or look at products within the same category.
“Let’s say handbags as an example. We'll send them an email featuring the handbags they looked at and some other product recommendations in that category as well,” he said.
An additional feature of the abandoned bag and checkout series was put in as well. If the price of an item that a customer had added to their bag was reduced, they would get an email notifying them of the change. The customer would also get an email notifying them if Bloomingdale’s had a promotional sale that affected the abandoned item.
Step #2. Develop campaigns based on customer data
Beginning in 2016, “once we had a lot of those very direct behavioral campaigns in place and were seeing great success, we worked … to come up with some more model-based campaigns,” Duffy said.
Based on customer data, the team began trying to anticipate customers who were highly engaged with the website but for some reason hadn’t purchased anything yet. Also, they began identifying customers who were perhaps steadily disengaging with Bloomingdale’s.
By sending timely emails using data, they believed they could prompt customers to convert more readily than with the previous general marketing campaigns.
Using reactivation as an example, Duffy said, in the past, customers who hadn’t engaged in the past six months were put into the same campaign bucket.
However, by analyzing customer’s behavior more individually, they’re able to “notice when the customer's behavior seems to be shifting and, let's say, starting to downward migrate a little bit. We can try to catch that customer earlier in the process, maybe before they've already disengaged with us,” he said.
Step #3. Consider the customer experience
As this effort has gone on, Duffy said, there are areas where the team has adjusted how they go about approaching customers with the triggered emails.
“In one direction, there would be times where we would tweak the logic about how many of each of these emails a customer could get within a certain time period,” Duffy said. “We started off conservatively where we would say if they get any … triggered email, make sure they can't get another one for a week or two weeks.”
As they began to layer in a few more of the campaigns and saw how strongly they were performing, he added, “we loosened those restrictions a little and continued to see strong results.”
So, while the team may limit how many abandoned cart campaigns a customer can get, that same customer could still get a browsing campaign. Whereas previously, while still testing, they would have been held out of other campaigns for a set time period.
They also tested the abandoned browse campaign to decide if customers should be shown a single item they’ve abandoned or an array of products.
“You may have looked at one women's handbag, let's say, three times, but you also looked at two others a few times,” he said. “Do we send you an email just about the handbag that you looked at the most, or do we send it with all three handbags you've looked at to get the full story in there?”
The team decided to focus on customers who were looking at multiple products across the same category to ensure that people getting the abandoned browse email had a strong interest.
These efforts have reaffirmed, Duffy said, the belief they had that the more personalized and the “closer to one-on-one marketing you can get, the better your return can be.”
“I think the takeaway from this … is just directly using the customer's actual behavior on the website and using that to inform your marketing message to her has been enormously productive and has been highly successful for us,” he added.
The team was aware at the beginning of this campaign that there were opportunities for them in an abandoned cart and checkout campaign, but “it’s opened our eyes to what the next levels we can take it to are — to do more advanced campaigns based on the customer’s direct behavior,” he said.
Since a unique aspect of department stores is that they carry all different categories of merchandise, Duffy said the team is now considering ways that they can dedicate certain offers or content to customers who are interested in particular brands.
“We are adding more campaigns, but the growth and demand that we're getting from [these] campaigns hasn't really slowed down since they launched with them,” Duffy said.
These types of triggered emails make up an increasing percentage of the total email demand, he added.
“It's grown from being pretty low to … double digits in terms of the percent of our email demand that comes from triggered campaigns like this,” he said.
The results the team has seen are:
“Over the past two years when we've had comparative [sends], that growth rate has stayed pretty consistent, so we haven't seen it start to taper off yet, which is great,” he said.
SmarterHQ – Bloomingdale’s behavioral marketing vendor
Download the free Quick Guide to Email Marketing
7 Steps to Discovering Your Essential Value Proposition with Simple A/B Tests
Drive better results when you discover what it is about your business that customers love.Get the Guide >
Increase Mobile Conversion Rates
These five free micro classes (each under 12 minutes) apply 25 years of research to help you maximize the impact of your messages in a mobile environment.Get the Course >
Test Discovery Tool
Show business leaders all the results of your testing efforts with this free tool.Get the Tool >
A Model of Your Customer's Mind
These 21 tools and concepts have helped capture more than $500 million in test wins.Download the File >
Research-based Lead Generation Swipe File
22 valid marketing experiments to give you ideas for your next A/B test.Download the File >
Receive the latest case studies and data on email, lead gen, and social media along with MarketingSherpa updates and promotions.