How Web Performance Affects Business Results (18 Case Studies)

Oct 22, 2020 | 6 min

Improving website performance is like regular exercise:

Everyone says it’s important, but few people actually do it.

Here’s why:

Optimizing performance/site speed is hard. And to be perfectly honest - it’s not all that exciting.

That’s why it’s essential to know why we should go through the trouble of doing it.

Put simply, web performance matters to business goals, like brand loyalty, sales and revenue.

But don’t take my word for it.

Below you can find 18 case studies that show speed’s direct effect on business results.

These studies were conducted by some of the most successful online businesses, including Amazon, Google and Netflix.

Let’s dive in.

1. Amazon: Site Speed Impacts Sales

Year: 2009

Industry: eCommerce

Affected Metrics: Conversions

Context: Back in 2009, few people cared about web performance. But then the biggest eCommerce website in the world said that site speed directly affects their sales. While the original case study is hard to find, Amazon has been used as an example of why speed matters ever since.

Key Finding: Every 100ms of latency cost Amazon 1% of their sales.

Amazon site speed case study

2. Google: Delays Under a Second Affect User Engagement

Year: 2009

Industry: Internet

Affected Metrics: Searches per user

Context: The next giant that started talking about speed is Google. In 2009 they also had an “intuition” that speed matters a lot. And data showed that they were right. Since then, no one has done more to advance web performance standards than Google.

Key Finding: The cost of bad performance increases and persists over time.

Google site speed case study

3. YouTube: Web Performance and Slow Connections

Year: 2009

Industry: Video sharing

Affected Metrics: Traffic and user engagement

Context: As a Google company, it’s logical that YouTube also started investing in web performance around 2009. Their team dug deep into the speed problem and saw that entire continents couldn’t use their service. This realization eventually led to the creation of imgix.

Key Finding: People experience different latencies, depending on their device and network connections.

YouTube site speed case study

4. Firefox: 3-part Case Study On Speed and Conversions

Year: 2010

Industry: Internet

Affected Metrics: Conversions

Context: In 2010, the browser war was firmly underway. That’s when Firefox tried to replicate Google’s web performance findings. Firefox’s results surprised even their own team, as site speed affected conversions much more than anticipated.

Key Finding: Reducing the average page load time by 2.2 seconds increased download conversions by 15.4%.

 Mozilla site speed case study

5. Walmart: Technical Deep Dive Into Web Performance

Year: 2012

Industry: eCommerce

Affected Metrics: Conversions, bounce rate and revenue

Context: Three years after Amazon announced its focus on speed, Walmart published this detailed presentation of their web performance. It’s clear that by 2012, Walmart already had a sophisticated web performance team. Earlier that year, around 2,2% of the Global Internet usage had gone through their website on Thanksgiving alone.

Key Finding: Speeding up pages by 100ms affects incremental revenue by up to 1%.

 Walmart site speed case study

6. Staples: Bounce Rates and Conversions

Year: 2014

Industry: eCommerce

Affected Metrics: Conversions and bounce rate

Context: Staples knows how important user experience is. Naturally, site speed became their focus after they saw how much it affected giants like Amazon, Yahoo and Walmart. Their case study starts with a very simple but important message: “Performance is a Business Problem”.

Key Finding: Refactoring CSS and JavaScript code reduced homepage load time by 1s, improving conversions by roughly 10%.

 Staples site speed case study

7. Optimizely: Slowing Down Websites (On Purpose)

Year: 2015

Industry: Media

Affected Metrics: Pageviews and returning users

Context: 2015 was a rough time for mobile web performance. That’s when Google (AMP) and Facebook (Instant Articles) both tried to address speed problems on mobile news sites. As an A/B testing company, Optimizely decided to delay The Telegraph on purpose to test just how much pageviews decreased.

Key Findings: An 8-second load time delay results in 17.52% fewer pageviews

Optimizely site speed case study

8. Financial Times: Speed’s Effect on Media Consumption

Year: 2016

Industry: Media

Affected Metrics: Article views and revenue

Context: As a media company, the Financial Times (FT) was also affected by the shift to mobile devices. And they knew they had to improve their website’s performance to stay relevant. That’s why they set out to see how much speed affected their core business metrics.

Key Finding: 1-second delay in page speed leads to a 4.6% drop in article views.

FT site speed case study

9. Mobify: Faster Websites = Deeper Engagement = Higher Conversion

Year: 2016

Industry: eCommerce

Affected Metrics: Conversions and revenue

Context: Mobify works with many online retailers, so it's no surprise that they have lots of web performance data. Their 2016 report focuses on the impact that speed has on homepages and checkouts. It also shows the direct effect of web performance in dollars.

Key Finding: For every 100ms of improved checkout load speed, Mobify’s customers saw 1.55% more conversions on average.

Mobify site speed case study

10. Zitmaxx Wonen: Building a Faster Mobile Experience

Year: 2017

Industry: eCommerce

Affected Metric: Conversions and revenue

Context: Zitmaxx Wonen is an online store that found itself in a very typical situation back in 2017. Mobile traffic was skyrocketing, but their website wasn’t optimized for mobile. This case study details the project they took on to measure, analyze and improve their performance.

Key Finding: Getting a perfect 100 on Google’s PageSpeed Insights and reducing the website’s load time under 4 seconds led to mobile conversions increasing by 50.2%.

 Zitmaxx Wonen site speed case study

11. Google: Analysis of 900,000 Mobile Ad Landing Pages

Year: 2017

Industry: Internet

Affected Metrics: Conversions and bounce rate

Context: By 2017, Google had more than enough data about site speed’s impact. They also knew that people demanded faster mobile websites. In this analysis, they computed how much bounce rate probably increases as page load time worsens.

Key Finding: As load times go from 1s to 3s, the probability of bounce increases by 32%.

Google 2017 site speed case study

12. Netflix: Prefetching and Optimizing HTML, CSS and JavaScript

Year: 2018

Industry: Tech & Entertainment

Affected Metrics: Time to Interactive and user clicks

Context: At this point, Netflix has become synonymous with tech excellence. Their infrastructure is beyond impressive and they constantly share their approach to microservices and DevOps. That’s why it’s no surprise that they also spend tons of time and effort on web performance.

Key Finding: Prefetching HTML, CSS and JavaScript reduces Time to Interactive by 50%.

Netflix site speed case study

13. BBC: How to Bring Sports to Over 100 Million Devices

Year: 2018

Industry: Media

Affected Metrics: Bounce rate

Context: BBC is one of (if not the) biggest traditional news publishers online. According to Similarweb, BBC's main site alone generates over 1.5 billion pageviews per month. Optimizing a website for that type of traffic is a problem that I hope we also have at some point.

Key Finding: BBC’s website loses 10% of its visitors for every additional second of load time.

BBC site speed case study

14. Google: How 10 Brands Improved Their Web Performance

Year: 2018

Industry: Internet, technology, lodging and eCommerce

Affected Metrics: Conversions, user engagement, Speed Index, Time to Interactive and others

Context: This is a big success story compilation by Google. It’s a practical look at ten businesses (including Walmart, Airbnb and eBay) with their reasons and approaches to improving site speed.

Key Finding: It is hard to pick just one finding since this is a collection of case studies. But an important theme that runs through the entire talk is this: Everyone in your organization must understand that web performance is a vital factor for success.

Google 2018 site speed case study

15. Furnspace: The Power of Image Optimization

Year: 2018

Industry: eCommerce

Affected Metrics: Conversions, bounce rate and mobile traffic

Context: I’ve talked about why image optimization is often the best thing you can do for site speed. This case study perfectly illustrates my point. Using ImageEngine, Furnspace saw impressive site speed and revenue results only by optimizing their images.

Key Finding: Image optimization decreased load time by 65%, leading to a 2x increase in conversions.

Furnspace site speed case study

16. Google: The Aftermath of Making Page Speed a Ranking Factor

Year: 2019

Industry: Internet

Affected Metrics: Abandonment rate

Context: Back in 2018, Google announced that page speed would become a ranking factor for mobile searches. The SEO world went crazy. Everyone scrambled to fix their broken mobile websites, which they should’ve done years ago. This article summarizes the results from this experiment a year later.

Key Finding: With speed improvements across the globe, Google saw a 20% reduction in abandonment rate for navigations initiated from search.

Google 2018 site speed case study

17. Google: Why The Core Web Vitals Are Necessary

Year: 2020

Industry: Internet

Affected Metrics: Abandonment rate

Context: The addition of the Core Web Vitals was Google’s biggest contribution to web performance in 2020. In this article, they explain how these three metrics affected vital performance indicators during their research.

Key Finding: Websites that met the thresholds for all three Core Web Vitals saw up to 24% less abandonment rate.

Google Core Web Vitals site speed case study

18. Web.dev: Improving LCP and Reducing Bounce Rate

Year: 2020

Industry: Media

Affected Metrics: User engagement, revenue, bounce rate and Largest Contentful Paint (LCP)

Context: This is one of the first comprehensive case studies about LCP’s effect on user engagement. As a news station, NDTV needs its largest above the fold elements to load quickly (like Yahoo and Google News). That’s why improving their LCP and CLS had such a massive impact on bounce rates.

Key Finding: 55% improvement in LCP time led to bounce rate decreasing by 50%.

 NDTV site speed case study

Final Thoughts

To sum up, I believe this quote from Mobify’s report (#9) encapsulates this article perfectly:

Faster Websites = Deeper Engagement = Higher Conversion

It’s. That. Simple.

This has been proven over and over again by some of the most successful businesses ever.

In short, if you want to compete online, you need to fix your website’s performance.

Google’s PageSpeed Insights is a great place to start diagnosing web performance. It gives you an overall optimization score, tracks vital metrics and provides suggestions for improving your site speed.

NitroPack PSI Score

Check out our PageSpeed Insights guide to see how you can interpret and improve your optimization score.