Editor’s Note
This article explores the dual pillars of a “good customer experience” in e-commerce—ease of information access and a stress-free interface—as defined by an expert with extensive experience in site optimization using heatmap analytics. It highlights the inherent challenge in meeting diverse user expectations, setting the stage for a deeper discussion on practical implementation.
What constitutes a “good customer experience” that should be provided on an e-commerce site? Mr. Itabashi from Faber Company, who has supported site improvements for approximately 100 companies using Mieruka Heatmap, states that it must satisfy the following two criteria:
– Easy access to desired information
– Stress-free UI
However, what users seek varies widely. Even when speaking of “desired information,” it’s unclear whether it’s product specifications or pricing, and it also differs depending on whether the user has a specific item in mind or not. Furthermore, points of stress vary from person to person and by the type of product being handled, making it difficult to define “stress-free.”
Forming hypotheses in such a vague state results in low accuracy and inefficiency. Mr. Nishikawa from F.D.C. Products, which operates 4℃, mentioned that while they tried various approaches at 4℃, they struggled to achieve the desired results.
At 4℃, Mr. Nishikawa’s team, responsible for the e-commerce site, handles not only site management but the entire scope of e-commerce operations, from order receipt and inventory management to picking, packing, shipping, and even post-purchase customer support (fulfillment). Consequently, there was a tendency to directly reflect customer feedback received through customer support onto the site. While this approach isn’t entirely wrong, it risks relying too much on intuition and subjectivity, lacking objectivity, or increasing disclaimers on product pages, which can dampen purchase intent.
Therefore, Mr. Nishikawa aimed to quantitatively visualize user needs and create an environment where individuals from different internal roles—such as brand managers and e-commerce staff, or creative staff and e-commerce staff—could discuss improvement strategies from the same perspective.
Mieruka Heatmap primarily offers the following three functions:
– Attention Heatmap: Identifies areas of thorough reading.
– Scroll Heatmap: Identifies drop-off points.
– Click Heatmap: Identifies click locations.
How did 4℃ improve its site using Mieruka Heatmap? Two specific examples were introduced.
Site Improvement Strategy 1: Revamping the Brand Top Page
While the first view remained a slideshow conveying the brand message, the area below it was significantly changed. Previously, it displayed new arrivals in order of release date under “New Arrival.” After the revamp, the New Arrival area was removed, and a “Category” section was newly established, allowing users to select categories like necklaces, rings, and pinky rings to navigate to respective listing pages.
The reason was that heatmap analysis revealed that “users on the 4℃ site wanted Category information.” On the previous top page, the Category listing page was quietly hidden within the top-left hamburger menu. Despite this, approximately 26% of all page clicks were concentrated on Category.
Given 4℃’s high gift demand, Mr. Itabashi hypothesized, “Users likely have a somewhat specific item in mind and want to choose from within that category.” Mr. Nishikawa responded, “Especially on e-commerce sites, I had a vague notion that many people search for products based on categories, like ‘let’s gift a necklace.’ The heatmap data presented clear facts, turning my assumption into conviction and enabling us to proceed with the revamp.”
As a result of adding the Category section based on heatmap data, the transition rate to each category’s listing page improved by approximately 150%, and the product purchase rate increased by about 114%.
Site Improvement Strategy 2: Pursuing Stress-Free UI Based on A/B Testing
From the beginning of 2024, Mr. Nishikawa’s team focused on “improving cart abandonment” as a major theme. While struggling to find good solutions, Mr. Itabashi used the “A/B Testing function” of Mieruka Heatmap to test various measures, also referencing competitor sites. The main ones were:
1. Fixing the conversion point (the ‘Proceed to Order’ button) as a floating banner: Previously, users had to scroll to the bottom of the page to purchase; this was improved to allow purchase from anywhere.
2. Simplifying the first view: Temporarily removed disclaimers that had increased from a customer support perspective.
3. Removing the footer menu to eliminate exit paths to other pages: Many users left the site via the footer menu after adding items to the cart, so removing the footer prevented this drop-off.
After about two months of testing numerous approaches through A/B testing, they successfully improved the conversion rate.
Mr. Itabashi, who advocates that “using A/B testing to speed up the D (Do: Implementation) and C (Check: Evaluation) of PDCA is important,” emphasizes that “if the P (Plan: Hypothesis Formulation) lacks evidence, it takes time to persuade superiors (stakeholders), hindering PDCA acceleration.”
No matter how promising a measure seems, a hypothesis remains just a hypothesis—it’s not infallible. Nevertheless, verification incurs some cost, be it expense, time, or manpower.
Since user behavior and needs are diverse, simply copying 4℃’s improvement strategies may not lead to good results. The way to find a winning pattern is to formulate hypotheses optimal for your own company and, each time, persuade stakeholders while pushing through improvements.
