Growing a business internationally is never easy. But when you are a family-run operation competing against global players in multi-lingual niche, things can get challenging. So what would success look like in that situation? For FreestyleXtreme, it was an 8% hike in sales from its web site following the implementation of AI-enabled marketing automation.
“We traditionally operated in very small niches, sub-segments of action sports,” explained FreestyleXtreme’s managing director, Shaun Loughlin, in an interview with DataIQ. “We’re going toe-to-toe with businesses ten times our size and marketing budget, but often we’re winning.”
For a business that can trace its origins to a van and a garden shed, that is an impressive acheivement. (It’s a story that’s worth a read here.) Even more so when you consider that it now has half a million customers across 60 countries communicating in 14 different languages.
It was this growth and internal resource constraints that led to a step-change in approach. “We were struggling to run a simple email marketing strategy,” explained Loughlin. “We had 20 local web sites with 14 different languages and customers interested in different sports in each market, so we wanted to be able to send 20 different emails. That would be hard with a large marketing team - we have just two people.”
The solution came in the form of an AI-enabled cloud marketing platform from Emarsys, including its Automation, Smart Insight, Web Extend, Predict (web and email) and Smart Content components. The impact was immediate just from an efficiency perspective. “Instead of taking two or three days to build an email campaign, it now takes a couple of hourse,” he said.
FreestyleXtreme provides a link from its customer database into the platform which then combines online behaviour with product interest and purchasing history to generate recommendations both for the email campaign and online. “What started as a basic newsletter, targeted by language, is now a matrix of language, product, relevance and lifestyle interests,” said Loughlin.
Having AI driving its digital marketing has delivered a significant impact. “The result is the important thing - we have added 8% to our web site revenues. That is enormous for a company our size,” he said. Overall, the new platform has been responsible for a 4.2% rise in directly attributable new revenue. “It is a black box that seems to produce magic,” said Loughlin.
Raj Balasundaram, VP solutions and strategy at Emarsys, greeted that description with a laugh during an interview with DataIQ from the US. “It isn’t a black box, but our customers don’t need to know how it works. For them, it is a platform. It is about combining strategy, understanding the customer and having the right data.”
“The way we look at artificial intelligence is not as a single entity - it is about solving particular problems with a combination of things, including machine learning (ML) and data science,” he said. Balasundaram pointed out that deploying ML to identify patterns in data is not an end in itself, it has to be focused on solving a business problem, such as whether visitors to a web site are converting or not. “That is what marketers are trying to solve on a daily basis,” he said. “Only now, instead of asking the marketer to do it, you can ask the machine. Marketers having been doing that using a sample extract and looking at a small number of groups - they can’t solve that problem when they are dealing with millions of people. That is what MIL and AI can do.”
For companies like FreestyleXtreme who are open to that shift in how they work, the benefits are significant. Others can be wary, fuelled by fears over what machine-led marketing might mean for them. As Balasundaram said: “When I meet people, often their first question is, ‘will AI take away my job?’ The answer is, not yet.” In fact, he views it in the historical context of IT which at every stage has been considered a direct threat to employment, but has generally created new, albeit different, forms of employment.
Balasundaram expects to see revenue gains across the board once marketers embrace these new tools, not least by avoiding margin loss and churn. “When you recommend a product to somebody, you need to make sure you use the right offer. Traditionally, retailers have offered big discounts to everybody when they go into Black Friday or other sales. In reality, not everybody wants a discount,” he said.
Pure-play ecommerce brands already struggle with low margins because of price pressure and the ease with which customers can compare them to their rivals. “Our customers are asking how to increase their margins - one lever they can pull is to give the most appropriate offer to every individual,” said Balasundaram.
For Loughlin, the results may seem like magic, but they did not come without some pain. “We now have a dashboard using Emarsys Smart Insight which puts our customers into brackets by activity level, retention rate, spend level, product category and brand preference. That has given us real insight into our data,” he explained.
“A major shock for us when we first got that dashboard was the percentage of inactive customers we had. We were aware to a degree, but having it presented in black and white means there is nowhere to hide. So now we have put measures in place, like a structured retention campaign,” he said.
Loughlin is also keen to ensure the company find the line between using customer data to personalise emails and recommendations while not taking that too far. As he noted: “There is a definite balance to strike between better marketing communications because of what we know about them and not using too much data and creeping them out. We have not got it quite right yet, but we are learning.”