How Much Traffic Can Your Website Handle?

How Much Traffic Can Your Website Handle?

Most websites load reasonably fast when visited by their average number of users. However, performance rapidly deteriorates when a site is overwhelmed by peak traffic (the times when the site’s traffic is the highest) and during traffic spikes.

In a quest to learn about the art and science of peak traffic estimation, I began to study some publicly available data to see if I could try to discover a connection between peak traffic and the average traffic of a website, as well as the type of traffic it can receive.

Summary and Data

Data for peak and average site traffic analysis came from Quantcast.

Here is a typical traffic chart from Quantcast:

For the purpose of this study, we will only look at traffic data from the US. We also obtained the average and maximum daily traffic over a 6-month period. This data was first retrieved on February 12, 2011. Only MRC accredited data was used.

The following summarizes the data obtained.

High-Ranking Websites (About 400,000 Visitors Monthly)

Rank URL Monthly visitors Average daily visitors Maximum daily visitors *Factor
4,988 409,400 13,647 61,700 5
4,912 415,400 13,847 39,900 3
4,908 415,600 13,854 29,500 3
5,009 407,500 13,584 104,000 8
4,937 413,600 13,787 53,000 4
4,978 410,200 13,674 50,600 4
4,964 411,700 13,724 80,800 6
5,010 407,400 13,580 18,700 2
5,072 403,000 13,434 25,300 2

*Factor is the maximum daily visitors divided by the average daily visitors.

Notice that the maximum daily visitors are 2-8 times higher than average daily visitors. This is from monthly traffic in the 400,000-visitors range.

Middle-Ranking Websites (About 190,000 Visitors Monthly)

Rank URL Monthly visitors Average daily visitors Maximum daily visitors *Factor
9,938 195,300 6,510 56,400 9
9,943 195,300 6,510 28,900 5
9,948 195,200 6,507 10,400 2
9,952 195,100 6,504 12,600 2
9,967 194,800 6,494 26,300 5
9,969 194,800 6,494 21,700 4
9,973 194,700 6,490 22,900 4
9,978 194,600 6,487 19,100 3
9,916 195,900 6,530 18,300 3
10,155 191,200 6,374 62,900 10

Note that at this traffic range, maximum daily visitors are now 2-10 times higher than average daily visitors. Monthly traffic that each site received was about 190,000.

Lower-Ranking Websites

Rank URL Monthly visitors Average daily visitors Maximum daily visitors *Factor
19,895 89,900 2,997 5,100 2
20,146 88,700 2,957 110,000 38
17,722 103,200 3,440 88,300 26
19,996 89,500 2,984 6,100 3
20,025 89,300 2,977 7,400 3
19,920 89,800 2,994 16,200 6
20,181 88,500 2,950 6,700 3
19,904 89,900 2,997 8,600 3
19,763 90,600 3,020 13,200 5
20,173 88,500 2,950 6,500 3

You may notice that maximum daily visitors are greater than average daily visitors by a factor between 2-38 times. The variance has thus increased dramatically. A simple explanation would be that higher-ranking sites with high average monthly visitors are less likely to be affected by sudden surges in traffic.

Large sites like CNN, TechCrunch and Mashable have high, stable visitorship, and the impact of a popular news story such as protests in Egypt or a new Apple product would be much lower as compared to a site with low average visitorship.

On the other hand, a site like can be flooded with traffic if a particular piece of news about the site owner becomes viral.

Let’s take a closer look.

Three Types of Traffic Patterns

There are three main types of traffic patterns.

High Stable Increasing Traffic has high stable traffic of 44.1 million monthly US visitors. It has a peak daily traffic of 2.2 times of its daily average traffic.

Extrapolation of the rising trend puts their estimated peak daily traffic to be about 4 million by June 2011.

Seasonal Peak Traffic

Some sites experience spikes in traffic due to seasonal promotion or normal business cycles. Let’s take a look at data from Holidays Central.

Notice that there is seasonal peak traffic. In this case, we can estimate based on past data, but we need to also factor in growth in traffic.

Low Stable Traffic

Let’s take a look at the chart for

On December 14, 2010, Michael Moore decided to contribute $20,000 to Julian Assange’s bail, resulting in a lot of press, which coincides with the drastic spike in the site owner’s website traffic. This news was reported on major news channels, causing traffic to spike 38 times higher than normal.

Defend Against Sudden Traffic Spikes

It is always hard to predict peak traffic for sites like because there is always a possibility that some of its content becomes viral.

According to Scott Galloway, Clinical Associate Professor of Marketing at NYU, there are 3 elements of viral content:

  1. Authenticity
  2. Humor
  3. Social Debate

In Michael Moore’s case, we can see these elements coming into play.

What is interesting, though, is that most of the time, going viral normally catches people by surprise. They are not prepared for sudden fame. Neither are their websites.

Imagine your website being hit with 100,000 visitors in an hour — you should be overjoyed, right? But many site owners actually have a bad experience because the site is slow or unavailable.

If your website falls into the category of having low, stable traffic with the chance of going viral, you should not hesitate to estimate that your requirements would be more than 30 times your average traffic.

Controlled Traffic Spikes

Besides being viral, normal marketing and promotion can also cause a spike in traffic. As a site owner, this traffic spike should be more predictable to you and is more likely to happen than accidentally achieving viral site traffic levels.

For example, the site completebody recently did a Groupon promotion. This resulted in a traffic spike from 150 regular daily visitors to 7,000 in one day. That is more than 46 times more than the normal traffic.


We can follow these guidelines for estimating peak traffic:

  1. If you do not have prior peak traffic data, if your site has low visitorship, and has content that could go viral, account for peak traffic that can be up to 30 times your average daily traffic.
  2. If your site has high and stable visitorship, peak traffic can be up to 5 times your average daily traffic.
  3. If you have prior peak traffic data and timing of peak traffic is predictable (seasonal traffic and controlled traffic spikes), use past data and add a percent growth in traffic to arrive at the final number.

Developers and website owners should determine their site’s capacity by using load-testing tools such as Load Impact (full disclosure: I work for Load Impact) to simulate traffic to their website. That way, they will be able to understand how the website performs under high traffic and tune the website before peak traffic arrives.

Related Content

About the Author

Julia May is a freelance writer based in Ukraine. Learn more about Julia over at her personal site and follow her on Twitter @indigomay.

This was published on Mar 4, 2011


Brian Carey Mar 04 2011

Great read, thanks!

Robert Mar 04 2011

A very interesting analysis and thanks for sharing. We also see large peaks in traffic when we run Email Marketing campaigns for our clients which is always expected. It is a good idea to invest in cloud hosting, despite being more expensive it allows you to easily monitor your bandwidth and simply upgrade your processor power and memory allocation instantly. It’s also a good idea to have website monitoring so if you suddenly become famous overnight, you’re the first to know and can ensure your website it working. Hope that helps someone.

Intrahost Mar 04 2011

It’s interesting to see how popular some of these sites are. One of the concerns for websites that aren’t used to a huge spike of traffic is whether there hosting provider can handle this amount of traffic.

Many of these websites will be running many dedicated servers of even using the “Cloud”.

David Mar 04 2011

Informative, thanks

design boy Mar 04 2011

Thank you for the useful advice, great read.

Rhoda Bernstein Mar 04 2011

A good summary of sizing challenges. Thanks.

Also, folks might want to check out node.js and Joyent. I’ve been doing some contract work for them (marketing) and if you look at their almost counter-intuitive approach to autoscaling, it helps deal with unanticipated spikes. I am only pseudo techy, so I could be completely off base. But, I am interested to know what those of you who are truly scaling geeks (yes, I am not worthy) think about Joyent’s approach and node.js.

Alwyn Mar 04 2011

One of my sites, recently got stumbled, and as a result of the stumble had 2 other articles “stumbled” pushing a ordinary site up from a small visitor count to just over 90 000 visitors for Feb. The climb has steadily increased in the past three months. with new highs being reached every month.

I thought that it will be a nice way to see when the site would crumble, but to my surprise everything went well apart from me having to jump in and get extra bandwidth a time or two.

This all happened on a shared hosting plan with not even caching or anything.

I think that a proper host is prepared for more than we might imagine.

Ryan Sharp Mar 04 2011

The most ironic thing about is the fact they are using Apache. Nobody should be using that outdated, synchronous ball of cruft in this day and age.

Harry Mar 04 2011

Definitely a great read, though I must say I can only dream of having ‘too much’ traffic to handle right now!

Brankica Mar 04 2011

Wow, nice info and detailed research. Loved it and glad I can see it before it happens. I had the biggest spike a few days ago and it was about 5 times my usual traffic. But fortunately it didn’t crash my blog :)

Thanks for the warning :)

Giannis Mar 04 2011

Have you ever thought to analyze this data through SPSS for a more scientific study? For example you can analyze Quantcast data through Principal Component Analysis

Usman Mar 05 2011

The bandwidth of a website will be exceeded if more then expected visitors get inside a website

A good website has a hosting company that can handle huge traffic on its servers

Surrey Mar 05 2011

Not too much of an issue for my site at present, but great read none the less :)

I think the key issue is not whether the website is able to load, but rather how fast does it load? That is a more qualitative feature since most people get irritated by slow loading websites. Maybe your bandwidth can handle the load, but because of your server caching settings, inefficient code, database errors, redirects.. the website still loads slowly under high traffic.

David Mar 06 2011

Nice article, we have spikes in traffic when running google ads campaign, recently thinking of moving to a beter hosting company to handle the traffic.

Chase Sagum Mar 20 2011

It’s about time somebody wrote about this. What do you think of these “grid services” web hosting providers are offering? I was under the impression that these grid services were a solution to sudden peaks in traffic.

thedevelopertuts Apr 27 2011

The loadimpact test helped me a lot. Thanks for talking about it, I would have never known it.

And the article is great!

This comment section is closed. Please contact us if you have important new information about this post.