Triumph International, an intimates company whose products are sold in 80 countries, recently completed a four-month pilot program with India-based tech firm ZS to implement its Personalize.AI software to recruit, retain and recapture lapsed customers.
The result was a 40 percent conversion lift in the Asia-Pacific markets of Singapore, Malaysia, Taiwan, Hong Kong, China, Vietnam, Thailand, Australia and India.
On May 31 the companies agreed to continue the partnership for at least another two years.
Related Stories
“What we’re able to do…for the first time at scale is provide considerable levels of personalization,” Neil Lewis, Triumph International’s vice president of marketing, APAC, told Sourcing Journal. “The AI is really helping us to communicate back to the consumer in a really relevant way.”
In a nutshell, the program takes the personal data of each consumer, such as their purchase and search history, and the AI predicts what product he or she would most likely want to purchase next. Those images and offers are presented in email and other marketing materials as part of a “curated journey” with the brand.
ZS partner Arun Shastri said the AI function includes fundamental components such as basic ones like content generation or summarizing text, as well as more complicated ones such as interacting with the customer in dialogue.
“If you have a good member base, right then AI immediately can help you leapfrog some of your legacy traditional problems, because personalization is now an expectation from the customer,” Gopi Vikranth, partner and ZS and lead for its Personalize.AI program, told Sourcing Journal. “Maybe 10 years back it was a bit more optional, but now you expect it as a consumer that I want something relevant from the corporation.”
The tricky part is getting customers to collect that information from, and according to Vikranth, Triumph’s strong loyalty program put it in a better position to succeed with Personalize.AI.
“They also have a lot of physical presence in terms of the stores across all these different countries,” Vikranth added. “Also the business understanding and associates that are capturing all that [customer] intelligence.”
Lewis said Triumph’s initial data dive before launching the pilot focused on finding out what increases customer loyalty. The number of purchases required varied in each market, usually correlating to the amount of competition there, and measuring the loyalty of those who had redeemed points vs. those who hadn’t.
“So there was a trigger point around getting customers to redeem points and we were able to take those learnings to understand what are the elements of the program that drive level of engagement,” Lewis said. “Then we played those back in a campaign targeting consumers who hadn’t shopped with us for over 24 months. Suddenly, they were receiving personalized communications with their name, their number of points, or products they actually liked, colors, sizes, fits. Previously they had just been sent communication that was just not relevant for them because we were sending one same communication to the entire base.”
Responsibly handling consumer information is important for both the consumer and the company trying to get the most out of Personalize.AI, Shastri said.
“You don’t want to spam your customer. You don’t want to get data on the customer, especially if they [wish] to be forgotten; they do have a right to be forgotten,” he said. “Which data you’re going to reuse, how good your models are, etc. are an important aspect of trustworthiness and that’s something we take very seriously within Personalize.AI.”
That’s especially when selling intimate apparel.
“Fit is really key. So when we know you buy a certain fit of a bra, it’s unlikely you’re going to change that fit unless something significant happens,” Lewis said. “What bra can we sell her that is the same fit but dressed up in a different way? That really worked for us.”
Lewis said the biggest surprise of the four-month pilot run was how quickly the technology was able to scale.
“Our CRM system allowed us to do one-to-one dynamic content with a different version of the communication,” Lewis said. “It’s not just sending one communication and forgetting about it; they’re really planning a campaign to have a four-to-six week period where we can introduce something and follow up with [the customer] a week later if she hasn’t done something else.”
The simultaneous combination of e-mail and SMS blasts at the same time, is another feature Lewis wants to grow.
“We want to build that into new channels like web personalization and maybe push notifications on apps and things like that,” Lewis said. “So we can really think about all of the touch points the consumer has.”
.
