# Data Science Visualizations to Improve eCommerce Sales | LinkedIn
URL:: https://www.linkedin.com/pulse/data-science-visualizations-improve-ecommerce-sales-steven-maurer/
Author:: linkedin.com
## Highlights
> Let's start with a performance histogram and what I call "performance sensitivity". I created the chart above using excel and invented numbers, but it's very close to typical data you see at major name brands. (Note: if this were real, it would come from what customers' browsers reported through the W3 navigation timing standard into a RUM or Real User Measurement JavaScript collection snippet that you would include on your site.) The green stripes are counts of users who came to the website. Some users get a fast experience. Others, a slow one. The height of the stripe is the count of user visits with the average page load speed they had marked (in seconds) at the bottom. As you can see, the peak of the green curve shows that around 49,000 users got average page load speeds averaging 3.7 to 4.1 seconds each. That's the performance that this (made up) website most commonly delivers.
> Addressing performance does pay off. Sales gains of anywhere from 5% to 30% for improving site performance is not unheard of, depending mostly on what issues your site currently has, and how important performance is to your customers. As a rule of thumb, if customers bother to call you up to complain about how slow your website is, it's at the upper end of that range. Most of the rest won't. They just leave.
---
Title: Data Science Visualizations to Improve eCommerce Sales | LinkedIn
Author: linkedin.com
Tags: readwise, articles
date: 2024-01-30
---
# Data Science Visualizations to Improve eCommerce Sales | LinkedIn
URL:: https://www.linkedin.com/pulse/data-science-visualizations-improve-ecommerce-sales-steven-maurer/
Author:: linkedin.com
## AI-Generated Summary
None
## Highlights
> Let's start with a performance histogram and what I call "performance sensitivity". I created the chart above using excel and invented numbers, but it's very close to typical data you see at major name brands. (Note: if this were real, it would come from what customers' browsers reported through the W3 navigation timing standard into a RUM or Real User Measurement JavaScript collection snippet that you would include on your site.) The green stripes are counts of users who came to the website. Some users get a fast experience. Others, a slow one. The height of the stripe is the count of user visits with the average page load speed they had marked (in seconds) at the bottom. As you can see, the peak of the green curve shows that around 49,000 users got average page load speeds averaging 3.7 to 4.1 seconds each. That's the performance that this (made up) website most commonly delivers.
> Addressing performance does pay off. Sales gains of anywhere from 5% to 30% for improving site performance is not unheard of, depending mostly on what issues your site currently has, and how important performance is to your customers. As a rule of thumb, if customers bother to call you up to complain about how slow your website is, it's at the upper end of that range. Most of the rest won't. They just leave.