Six case studies of machine-learning powered email marketing – Econsultancy

Machine learning has changed the game for email marketers.

Once hyped as the ‘next big thing’, it is now being put into practice by a wide range of businesses to improve the effectiveness of email.

Here are just six case studies that demonstrate its success.


Language plays a huge part in why consumers respond to marketing (and why they don’t). To figure out why its email engagement rates were declining, Dell partnered with Persado in 2016. Persado is a technology company that machine-generates email copy (categorised by emotion) and analyses which key phrases or words succeed where others don’t.

Through Persado’s AI-driven tool, Dell was able to optimize language for each of its segments, ultimately enhancing the effectiveness of the channel.

For example, the software helped create email copy that was more than just descriptive, it injected characteristics like exclusivity to prompt engagement.

In the example below, the top Persado subject line (invoking exclusivity and inspiring gratitude and anxiety) reads:

‘Update: These massive deals – Up to 50% off, because you’re on our list’

As a result of using Persado’s platform, Dell saw a 50% average increase in email CTR, and a 46% average increase in conversions from email. It also generated a 77% average increase in add-to-carts, and a 22% average lift in page visits. Dell has since used Persado to improve the marketing copy of its Facebook ads and display banners.


Adobe’s Sensei platform uses artificial intelligence to power more than just language. It also optimises send time for each recipient, and estimates the likelihood of interaction based on customer profile.

Sky is one brand that has used Adobe Sensei to improve customer interactions across digital channels, including email. As Rob McLaughlin, Head of Digital Decisioning and Analytics at Sky UK, states: “We have 22.5 million very diverse customers. Even trying to divide people by favourite television genre can result in pretty broad segments.”

By using an AI and machine learning framework, Sky was able to make sense of its huge volumes of customer information to discover the recommendations, services, and experiences that resonated the most with each individual customer.

READ ALSO  How machine learning is transforming retail both online and offline – Econsultancy

For example, the brand optimised emails for customers that have Sky Sports under their plan but have not downloaded the Sky Sports app, ultimately drawing them in with personalisation.

As McLaughlin reiterates: “It gives us the chance to surprise and delight customers with recommendations that are not only relevant, but wanted.”

Virgin Holidays

We’ve featured Virgin Holidays’ AI-powered email marketing before, but it’s certainly one that stands out in terms of results. Previously, the travel brand faced a number of challenges with internal structures, particularly relating to the sign-off processes involved in email marketing and copywriting. This resulted in a lack of creativity and the absence of a testing culture.

To combat this, the brand turned to Phrasee, which is a platform that uses artificial intelligence to automate and optimise subject lines. Through its technology, Phrasee generates marketing language that is relevant and engaging to users, in turn freeing up marketers who would otherwise spend lengthy amounts of time manually A/B testing.

Virgin Holidays turned to Phrasee during one particular marketing campaign, whereby ‘directness’ and ‘curiosity’ had previously been the most successful types of subject lines. However, as the brand entered its peak sales phase, Phrasee predicted that ‘directness’ and ‘urgency’ were going to be more effective, and immediately started to auto generate subject lines in that area.

Virgin Holidays saw a 22% increase in clicks from Phrasee subject lines, as well as a 2% increase in open rates. AI-powered email campaigns also generated 66% increased awareness and a 33% increase in web traffic.


Another area where artificial intelligence can be hugely beneficial is in email automation, with the technology allowing brands to send reactive and relevant messaging at the right time. Trendyol, a leading commerce company in Turkey, faces big competition from global brands like Adidas and ASOS – particularly for sportswear. To help enhance its email campaigns, it partnered with vendor Liveclicker, which specialises in real-time personalisation.

Trendyol’s email campaign used AI to distinguish which messages would be most relevant to which customers. It also created an offer for a football jersey using the platform’s ‘Liveimage’ tool, superimposing the recipient’s name on the back to ramp up personalisation.

READ ALSO  Marketing's CX efforts are being thwarted by technology [stats]

These emails generated a 30% lift in click-through rates for Trendyol, a 62% lift in response rates, and an impressive 130% lift in conversion rates.

Harley Davidson

Albert is a machine-learning powered tool that can be applied to various marketing channels including social media and email. The software predicts which consumers are most likely to convert and adjusts creative and copy autonomously.

Harley Davidson is just one brand that made use of Albert. The motorbike manufacturer analysed existing customer data to determine the characteristics and behaviours of previous customers who had acted positively. These actions include purchasing, adding items to carts, or spending more than the average amount of time browsing the site. From this, Albert created micro segments of customers that resembled them, and scaled up test campaigns accordingly.

Results show that Harley Davidson increased sales by 40% by using Albert. The brand also had a 2,930% increase in leads, with 50% of those from high converting ‘lookalikes’ identified by the AI.


Before turning to AI engine Emarsys, lingerie brand Cosabella relied on manual efforts to create its email campaigns. However, it was struggling to produce relevant and personalised email at scale, with employees unable to understand or make use of the multitude of data available.

The AI interpreted two years’ worth of customer data into easy-to-understand charts. From this, Cosabella was able to draw on a recommendation engine to create personalised and targeted email at scale.

According to Emarsys, the vendor helped Cosabella increase revenue from email 60% year on year. It’s email database doubled, and click-through rates increased 100%. Alongside this, the time taken to create email campaigns decreased by more than half, freeing up its marketing team to focus on more creative and over-arching strategic tasks.


An Econsultancy subscription gives access to:

Source link

WP Twitter Auto Publish Powered By :