Discover Customer Trends with Ecommerce AI

The ecommerce environment has been full of changes in 2023. Brands experienced strong revenue growth in the first half of the year, followed by slowdown in the second half. However, at Ivy Insights, we also uncovered exciting opportunities. For example, affluent mature consumers drove profit growth for brands, more so than the past few years.

Ecommerce brands that stay ahead of customer trends are the biggest winners. Empowered by comprehensive customer insights, they tailor sales and marketing strategies to different customer cohorts, leading to superior return in today’s economy.

Fortunately, recent advancements in GPT make such insights accessible to brands of all sizes. In this article, I will discuss how you can utilize AI and ML to leverage your company’s full revenue potential.

Start with business goals

In this case study, let’s use a dataset from lifestyle brand Ivy Essentials. The brand is focused on maximizing revenue and reducing customer acquisition costs across its products. To achieve this goal, let’s break down the insights we need:

• Where customers come from and what they look like

• What they like to buy and how they like to buy them

• What prompts them to convert

With this framework, we develop hypotheses about data variables relevant to Ivy Essentials’ business case, and measure 30+ variables accordingly.

Perform customer segmentation

Then, we leverage machine learning to analyze customer characteristics. The chart below showcases one of the results from customer segmentation. Each line represents a customer and their characteristics measured by key factors distilled from the 30+ variables above. Additionally, line colors represent different customer cohorts.

Manually extracting insights from complex segmentation results would be difficult. Now, AI can handle the task effortlessly. Asking AI to interpret customer profiles from the 30+ data variables, we learn that income, age, and product preferences are the key factors that differentiate customer cohorts from one another, as shown below.

After further reviewing factor distributions, we decide to segment customers into six cohorts. This way, we preserve a large number of unique characteristics across customer cohorts, and ensure insights from customer segmentation are actionable for Ivy Essentials’ team.

Unpack customer preferences and conversion patterns

A critical insight we seek is identifying who purchases each product type, especially those premium products. Since the target audience of each product category overlaps to some extent, it’s challenging for us to find out such information using traditional methods. Luckily, customer segmentation highlights distinct product preferences across cohorts.

The chart above shows that Established Suburban Families are the main contributors to Premium products’ revenue. Moreover, they are less fond of Hero and Plus products compared to other cohorts. Marketing data also reveals that they are particularly drawn to creatives that emphasize product quality and durability.

Apply customer insights

Next, we will need to come up with new sales and marketing strategies based on our insights regarding customer preferences and conversion patterns. Such tasks can be daunting for human brains. But the good news is that AI excels in providing recommendations based on expansive business cases.

To fully leverage AI’s capabilities, we feed it business descriptions and real-time data as context. We also ask it to consider the current economic trends.

AI is able to incorporate each customer cohort’s demographics and purchasing patterns in its recommendations for promotional strategies, messaging, and creatives. By setting up location, age, and gender targeting according to segmentation results, we can quickly implement the new strategies in ad campaigns.

Discover optimal strategies through experiments

We get the best business outcome from customer cohort-based strategies when target audience, promotions, messaging, and creatives are fully aligned. To achieve that, we often need to run multiple multivariate experiments, where we test several factors at the same time. In these experiments, we aim to increase return on ad spend by cohort over time.

From our experience working with ecommerce brands, we discover that every sizable brand sells to multiple types of target audiences. By implementing cohort-based strategies, brands achieve an improvement of over 10% in ad campaign profitability within the first month. Profitability continues to increase as brands gain more in-depth insights about their customers and further customize their strategies using AI and ML.

Ivy Insights provides AI-powered customer insights platform for ecommerce brands. If you are interested in leveling up your business using AI, you can scheduling a time with us here.