Sophelle Blog

By Jeff Neville

Best practices for implementing optimization

Personalization engines dominate retail marketing, but they come with a hefty price tag. Some CFOs find themselves asking a simple question: is this software worth it?

With the likelihood of a recession getting higher and costs becoming more scrutinized, growth-minded retail leaders know they need to invest wisely to compete in today’s market. CFOs sign off on personalization software because companies selling the systems promise big results. But if personalization is underleveraged, a retailer may not see the ROI they’re looking for. 

As a retail consulting agency supporting over 70 clients with ecommerce personalization services, Sophelle has developed a framework for getting the most out of a personalization platform. Here’s what you need to know when it comes to measuring, utilizing, and supporting personalization software for maximum ROI. 

KPIs to Assess eCommerce Personalization

The first step in determining ROI on a personalization investment is knowing how to measure it. Personalization platforms produce a significant amount of data, but we recommend retail leaders closely monitor and measure a handful of KPIs to determine how well the system is performing.

Most ecommerce retailers track and review these four metrics on a month-over-month and year-over-year basis for each site page:

• Average order value 
• Conversion rate
• Bounce rate
• Abandon cart rate

If personalization software is working effectively, each of these KPIs should change significantly (increases in AOV and conversion rates; decreases in bounce and abandon cart rates). A corresponding increase in revenue should eventually follow. The extent of the change depends on a number of factors, ranging from retail segment to target audience to current site performance. 

Understanding the influence of each experience test on these individual KPIs is crucial to measuring the impact of a personalization platform. Proper A/B testing practices should be followed, and retailers should resist making any other experience changes outside of the personalization engine during testing. Doing so ensures that any performance improvements can be directly attributed to the platform and helps more clearly document its ROI.

5 Best Practices for eCommerce Personalization

To move the needle on core KPIs, personalization software needs to be operated in accordance with industry best practices. Sophelle has created a detailed guide to ecommerce personalization best practices, a few of which are summarized here: 

1. Consider all touchpoints in the customer journey, starting with the point of entry

Personalization elements should vary depending on where a customer is in the purchasing process and consider how they navigate to and through your online space.

Understanding how customers get to your website makes it possible to personalize and accelerate their journey. Knowing the device they’re using, their location, the term they searched, or the ad they clicked on can help retailers determine the customer’s intention.  

2. Understand who your customers are and how they behave

Understanding and leveraging customer demographics (including age, gender, ethnicity, nationality, and income level) often lead to increased conversion rates and the ability to identify new opportunities. 

Additionally, retailers should analyze how customers shop, including how they interact with a retailer’s website, as that information provides excellent insight into who they are and what interests them.

For example, Sophelle recently helped a leading vendor of residential and commercial construction supplies that wanted to increase ecommerce revenue through personalization. After analyzing their algorithms and adjusting their strategy to leverage customer-specific, real-time recommendations, the retailer achieved a 265% lift in revenue. 

3. Don’t over-engineer personalization experiences—let the software lead

The benefit of machine learning is that it can optimize itself, getting smarter with every customer interaction. Let the software work its magic for you, and don’t over-engineer the process. 

The simple truth is that the amount of data that AI can take in, synthesize, and analyze far surpasses human capabilities. Allow the personalization tool to work to its full potential.

We have two examples to illustrate these points. First, a cosmetics retailer had a complex system of over 25 algorithms making customer-specific product recommendations across web and mobile channels. Sophelle created advanced data models and implemented A/B testing to identify and realign algorithms that maximized engagement and revenue. As a result, the retailer saw an 11.5% lift in revenue on their highly trafficked product detail pages.

Second, a major apparel retailer wanted to capitalize on an increase in ecommerce that occurred during the height of the pandemic, but they were relying heavily on recommendations entered manually by merchandisers. Sophelle analyzed and compared the performance of all product recommendation algorithms, revealing through testing that AI algorithms outperformed merchandiser intuition by a landslide.

4. Provide a seamless customer experience

Today’s consumers are constantly switching back and forth between devices. Incorporating cross-device connection improves the customer experience and leads to higher conversion rates.

Social login enhances a seamless experience, allowing customers to register and sign in to websites with their social media credentials. Customers appreciate this option because it simplifies signing into a new website, and retailers can learn a customer’s age and birthdate, interests, hashtags, and even their favorite sports teams once customers are connected.

5. Segment users based on device

We know device usage typically breaks down by age group. The performance of ecommerce personalization testing must be assessed separately for desktop and mobile users, as they may very well be customers from different generations with very different needs.

For the complete list of best practices, download Sophelle’s Focus Report on eCommerce Personalization. 

Do You Have the Right Skills on Your Team for eCommerce Personalization?

The final element required to obtain significant ROI on a personalization investment is people. To implement these best practices and optimize personalization software, retailers need a team that has the right skills, capacity, and mindset to do the job.

• Core competencies. Someone who excels at personalization understands analytics, develops insights, and implements data-driven processes. Traditional retail marketers are often good at engaging customers and driving interactions, but the scientific thinking skills required to test a personalization platform may be outside of their core competencies. 
• Bandwidth. Personalization is an ongoing effort, not a “set it and forget it” project. While it may not be a full-time position, too many personalization efforts suffer from a lack of attention: operating and monitoring the system is assigned to a team member as one of their many tasks. When other items become more pressing, the system is abandoned or forgotten. It’s vital to have personalization overseen by a team with adequate bandwidth and time to test, analyze, learn, and iterate consistently.
• Test and learn mindset. Personalization evolves and improves through trial and error. This “test and learn” mindset runs counter to the approach of many marketers, who often face immense pressure to get a campaign “right” the first time. The team running personalization software needs an iterative, engineering-type approach, not necessarily a marketer’s creative mind. The ability to be comfortable with experimentation—and failure—is critical to success. 

The importance of these qualifications begs the question: is it best to manage personalization processes and software in-house? While some companies are large enough to have dedicated teams focused solely on personalization, many retailers benefit from finding a partner who can act as their personalization experts. Personalization specialists have the core skills, capacity, and mindset required to get the most out of a personalization platform. 

Retailers also benefit from the collective learnings of outsourced personalization expertise. These companies have access to real-time testing data across multiple retailers and can provide insights about trends and themes in the retail personalization space gleaned from their full portfolio to better serve all their clients.

A Simple Way to Evaluate Your ROI

Determining the exact ROI of personalization software can be difficult, particularly if best practices aren’t followed or the right team isn’t in place to operate the platform. However, here’s a quick back-of-the-napkin calculation for evaluating the current return on investment. 

Evaluate the four core KPIs and determine the total benefit since implementing personalization software.

Add up total costs, including:

• Technology costs of the platform
• Implementation costs
• Ongoing fully loaded costs of the time people spend developing new experiences to test, implementing the tests, and managing the optimization program and the solution vendor

Then, determine the payback period. It is up to the individual retailer to decide whether that payback period is acceptable.

Keep in mind that many personalization investments can be further optimized with the right best practices and the right expertise. If you’re interested in learning how Sophelle can optimize your current ecommerce personalization efforts, contact us today.

If you haven’t implemented personalization in your business yet, contact Sophelle about a risk-free personalization trial.