Drive business growth with machine learning personalization - The Entrepreneurial Way with A.I.


Tuesday, December 1, 2020

Drive business growth with machine learning personalization


Powerful personalization, now for every business.

The rapid development of technology and the growing amount of time people spend online has increased the number of consumer touchpoints and interactions with brands they engage with. The concept of personalization (although not new), offers retail organizations the ability to improve brand loyalty, grow sales and revenue, and increase efficiencies by using data to create a more sophisticated and unique customer experiences. According to a 2019 McKinsey study, organizations that have implemented personalized recommendations and triggered communications have realized 5%-15% increases in revenue and 10%-30% increases in marketing spend efficiency.

As the ability to deliver more sophisticated shopping experiences has evolved over time, consumers today expect real-time, curated experiences across digital channels as they consider, purchase, and use products and services.

However, many retail organizations struggle to deliver personalized experiences due to scale limitations of static, rule based systems, complexities and costs implementing a machine learning (ML) based solution, and friction with platform integration.

Amazon Personalize enables developers in retail organizations to build applications with the same ML technology used by for real-time personalized recommendations – with no ML expertise required. Amazon Personalize makes it easy to develop applications for a wide array of personalization use cases, including specific product or content recommendations, relevant product rankings, and customized marketing communications. Let's take a look at how MECCA improved their customer experience to increase engagement with Amazon Personalize.

Delivering the same personalized experience online as in-store

MECCA, a beauty and cosmetics retailer based in Australia and New Zealand, prides itself on its highly personalized in-store services. Over time, shopping at MECCA becomes akin to having a personal stylist, as the store collects information about customers’ skin tone, style, and preferences to create an ongoing beauty profile that allows sales associates to recommend products based on prior visits and engagements.

MECCA sought to translate its in-store service to its digital outlets, enabling customers to get the same level of personalization and care online as they would in the store. A fast and effective proof of concept with Amazon Personalize, led by the MECCA technology and CRM teams, in collaboration with their partner Servian, Mecca has been able to pursue this goal without developing its own recommendation engine for email marketing communications.

"Since integrating (Amazon) Personalize, we are seeing our customers respond positively to the new recommendations with a 65% increase in email click-through rates and a corresponding increase in email revenue relating to the products recommended by (Amazon) Personalize."

Sam Bain

eCommerce and CRM Director at MECCA

Using the success of the email program as a proof of concept, MECCA is now extending use of Amazon Personalize to its website—demonstrating the benefits of starting small with Amazon Personalize and then expanding use if, when, and where it makes the most sense for your organization.

Taking the next step

Retail organizations that most effectively create personalized experiences to increase engagement and revenue are likely to thrive. With minimal investment required and fast ramp-up time, Amazon Personalize can make personalized recommendations and incredible customer experiences a reality—not in six months, not in a year—but in just a few days. Get started.

Amazon Personalize custom recommendation and ranking inference runs on Amazon EC2 C5 instances featuring the latest Intel® Xeon® Scalable processors and AVX 512. Amazon EC2 C5 instances deliver cost-effective high performance at a low price per compute ratio for running advanced compute-intensive workloads like machine/deep learning inference.


, Khareem Sudlow