Mobile web isn’t the only thing taking a hit. Radware’s Spring 2013 State of the Union – a look at page speed and web performance – found that internet performance is getting worse. Here are some key findings from Radware:

• Load times have increased by 22% in just one year. The median load time for a first-time visitor to a home page in the Alexa Retail 2000 was 7.25 seconds, compared to the median of 5.94 seconds recorded in December 2011.
• Top retailers underperformed the rest of the pack. The median top 100 site had a load time of 8.23 seconds, 14% slower than the overall median load time of 7.25 seconds. Top sites are also slowing down at a faster rate: 28% compared to 22% for the top 2,000 sites.
• Firefox outperformed Chrome and Internet Explorer. With a median load time of 6.64 seconds, Firefox was 8.4% faster than Internet Explorer 9, which had a median load time of 7.25 seconds. More significantly, we noted that median load times have slowed down for all three browsers. This suggests that, despite the huge performance leaps made by browser vendors, development cannot keep pace with the demands placed by ever-growing and increasingly complex web pages

Based on historical trends, we expected to see some degradation in page load time; this has been the trend since 2010, when this set of sites was initially measured. We did not, however, expect to see such a dramatic performance decline.

Which major news organization has the fastest loading website? What about the slowest? PBS’ Idea Lab did some research (using GTmetrix) and here’s what they came up with:

What’s slowing these websites down? Phillip Smith, who wrote the article, suspects it’s all the add-ons found on websites these days. From Idea Lab:

I had a hunch it might be widgets, buttons and badges. The Web is now awash in third-party services — Facebook, Google+, Twitter, just to name a few — that all provide their own handy little code snippet for you to embed on your site. Each of these little code snippets is time-saving for the developer implementing it, but potentially time-consuming for the end user. Could widgets be to blame? I asked [Alex Krohn, founder of ... hosting company Gossamer Threads, the company that created the GTmetrix performance testing tool] that question too.

“Each one of these adds a ‘hit’ to the site, in terms of the resources required and the total time to load everything. If you do it in a way that’s smart and doesn’t slow things down for the user, then it can be a good thing. However, if you do it wrong, it can impact the speed of the site,” Krohn said. He went on to explain, “It’s important to have a good understanding of the numbers and what they mean. So, for example, a ‘15-second page load time’ doesn’t necessarily mean the user is staring at a blank screen that whole time, and using a lot of resources may be OK if the user gets what they want out of the site within the first couple of seconds.”

Phillip takes quite the in-depth look at the issue. Check out his findings at Idea Lab >>>

You knew it was coming. Despite spending piles of money to air an ad during the prime time of prime times (the Super Bowl) some of the companies just weren’t ready for the resulting onslaught of web visitors.

Here’s a look at which companies could have used a little more load testing (or just under estimated how many visitors their ad would draw in), from Yottaa:

Coke
Coke had the worst performance of the night by far. Coke took an interactive approach to their commercial, directing users to vote for the ending at www.cokechase.com. However, www.cokechase.com had a page load time average of 62 seconds – which was longer than their entire commercial! Since users expect websites to load in 2 seconds or less, they were not very happy. Hundreds of tweets from angry users came flooding in.

Moral of the story, if you’re advertising during one of the most watched events in history, plan for more traffic than you could ever imagine. …

To make matters worse, Coke’s main U.S. homepage only had 92% uptime for the night! This marks the second straight Super Bowl where Coke’s website(s) has crashed.

SodaStream
SodaStream created a lot of buzz with their ad showcasing their cool soda maker. Millions rushed to their website to check out more of the product. The surge in traffic instantly caused a spike in performance by 5 seconds! Soon enough, tweets came rolling in that the website was unavailable.

Calvin Klein
Users flocked to the website to see more of the, ahem – “product”, but many were left disappointed as they could not access the website. …

The website’s performance increased by only 1 second over the ensuing hours, but a 1-second delay can decrease conversions by 7%! When it comes to the magnitude and scale of an event like the Super Bowl, every millisecond counts.

Axe
The website Axe sent people to, www.axeapollo.com, only had 88% uptime for the night and their homepage … was only up for 93% of the time.

Visit the Yottaa post to see a collection of traffic and load time charts, plus some examples of tweets that disappointed visitors inevitably shared when sites were slow to load.

Moral of this story: when you’re advertising during one of the most watched events in history, make sure your website can handle the insane amounts of traffic it will likely bring.

Forget losing customers, your business needs a mobile website in case of an emergency. As superstorm Sandy proved last week, a major event can knock out power for an extended period of time (some people are still without power more than a week later). That means no computers (once laptops lose their charge), no landline phones, no televisions, no refrigerators, microwaves or electric stoves. But people will still need information, and they’re going to turn to their mobile devices to get that info. Take a look at these stats from mobile website platform Usablnet, as reported by Venture Beat:

Mobile traffic to websites of utility companies like Con Edison increased by 16,000 percent both during and after hurricane Sandy.

Other sites saw major traffic boosts as well. Visits to transportation pages spiked by 495 percent, and telecoms saw their own traffic climb by 126 percent.

While these numbers are unsurprising on their own, they do show just how important the mobile web has become for finding information about natural disasters both before and as they are happening. The trick, then, is to ensure that people can access that information quickly and easily.

“The ability to access the mobile web becomes even more essential during times of crisis,”  Usablenet marketing chief Carin Van Vuuren said via email.

Read the full article at Venture Beat >>>

Providing vital information isn’t limited to utility companies, news and weather. If you can communication via social media or your mobile website that your storefront or restaurant is still open, you can literally be a lifesaver. During the past two weeks, people turned to the mobile web to share news of open coffee shops, buildings with make shift “charging stations” and gas stations with power and gas.

If your area is hit by a disaster you’ll want to make sure your mobile site is up and ready to provide the community with updates and news. And judging by the traffic numbers generated during Sandy, you’ll want to make sure you put your mobile website through some good load testing!

The launch of HTML5 game development portals seems to be very common lately. On Monday, I posted about the launch of Gamorlive, an HTML5 web portal for developing multiplayer games. It seems that a similar platform, Goko, launched last week. Unfortunately for Goko, the company overlooked load testing which caused so many problems that the web portal had to be pulled back into beta testing.

According to an article by Tricia Duryee in AllThingsD:

“In an apology on its site, the company explained: ‘As we moved out of beta and into the live world, we learned that we had problems with our backend and weren’t able to scale fast enough to support the load. As a result, we experienced numerous problems and created a really bad experience for the fans trying to access our games or even get on our website.’”

Goko wasn’t ready for real world usage, and as a result the application had a failed launch. Duryee in AllThingsD also adds that now it will be even harder for the company to convince developers the platform is workable once it launches again. With load testing you can ensure that your web app is ready for peak, real-world usage the first time it launches.

For more information on load testing click here.

The Mars rover Curiosity had a highly anticipated, highly nerve-wrecking (but successful) landing over the weekend. But for all of NASA’s concern that the landing go off without a hitch, they still had time to consider – and test – the landing’s live video stream. NASA knew a lot of people would be tuning into the live landing feed, and they wanted to make sure it could stand up to the demand.

For the actual testing, NASA teamed up with load specialists (and uTest partner) SOASTA. GigaOm writer Derrick Harris contacted the organizations to find out how they went about testing. Here’s a few highlights of what SOASTA did:

• SOASTA generated load from six Amazon EC2 regions across the world, generating more than 25 Gbps of traffic and pounding the application for nearly 40 minutes.
• After 20 minutes, they terminated 10 instances (see Arrow 1 on the chart) to see if their stack and Amazon’s cloud could handle the failure. This temporarily reduced the amount of traffic the system could handle, but Amazon’s Elastic Load Balancer service had the failed instances back up and handling 25 Gbps in about 5 minutes.

Read more…

Automation is tricky business. Knowing when to employ it (and when not to) is a decision with major consequences. Here with some good advice on test automation is James Bach, who had this to say in a 2009 interview with uTest’s Matt Johnston. Check it out:

Here’s how I think of it:

- Is the product highly controllable and observable? A command line tool that provides its output solely to the console window is inexpensive to automate, compared to an iPod touchscreen app. I want to get under the GUI.

- How expensive is the tool I’m using? I urge you not to use expensive tools, even if they work. Never let your manager buy them. Because expensive tools become something you MUST use, even if they don’t work. A free tool may be freely abandoned. This gives you flexibility.

- How well can I automate the oracle? Will the bugs be able to elude my automation because it can’t tell if a complex graphic is rendered correctly?

- What is the learning and testing value I’m giving up by using automated checks? I find that doing a test multiple times also causes me to learn and see new things in the product. Furthermore, when I re-run tests, I often run them in a different way, and that allows me to find new bugs.

- Can the automated check be parameterized and randomized, so that I get lots of similar checks for very little additional investment? I like automation more for data intensive testing, because I get new tests just by changing the database.

Read more…

The 2012 London Olympics are underway and they’ve already run int0 a few technical problems – most notably, the Twitter overload in London. But it looks like the Olympic committee might have been set up for failure from the beginning. According to a July 9 study the official Olympics site wasn’t prepared for the sheer volume of visitors to all-things-Olympic almost a month in advance.

A test gauging site performance put the official London Olympics website – london2012.com – at the bottom of the pack. Compuware, who conducted the test, measured the performance of london2012.com, londonolympics2012.com, visitlondon.com/london2012, timeout.com/London/Olympics-2012, tfl.gov.uk and lastminute.com against median performance metrics from the Alexa 100 index. Here’s how they did it and what they were looking for, from PCWorld:

Compuware completed its analysis on 9 July using SpeedoftheWeb.org, where the websites were tested according to 15 key performance indicators relating to user experience, browser, content, network and server. The performance of the sites was benchmarked against the performance of the Alexa 100, a global web index.

You would think thatlondon2012.com would be on their A-game since they are likely to be the most popular site during the Olympics. But alas, they fell short.

Read more…

The Olympics have only just begun, but we’ve already seen several technological issues surface. Perhaps the biggest mishap occurred yesterday when viewers experienced failed coverage of the cycling event.

What exactly happened? Interestingly, the Olympic committee organizers are blaming crowds of Internet-hungry attendees for failed television coverage. Due to the overload of fans’ text messages, tweets, and status updates during the race – the wireless network crashed. To top it off, the Olympic committee is asking attendees not to use their phones during the events.

As covered by Tom Cheredar of VentureBeat:

“‘From my understanding, One network was oversubscribed, and OBS (Olympic Broadcasting Services) are trying to spread the load to other providers,’ IOC Communications Director Mark Adams told The Guardian. ‘We don’t want to stop people engaging in this by social media but perhaps they might consider only sending urgent updates.’

Regardless of how it happened, I think it’s pretty much impossible to tell a crowd of excited sports fans to stop communicating with friends not at the event.

The network failure, however, isn’t a complete shock. U.K communications regulatory body Ofcom predicted that mobile data usage would double during the games, and previously said it was rolling out a plan to ensure sufficient spectrum was available deal with any congestion. But obviously, they didn’t do nearly enough.”

The bottom line is you can’t ask attendees not to access the Internet during the events. Not only is it impossible, but also cruel to our tech-savvy population. Instead, any bottlenecks needs to be found and taken care of. Obviously, it is hard to plan for these real world crashes. They occur randomly and unexpectedly. This is why a plan simply isn’t enough. The best practice in preparing for a global event like the Olympics, is to load test in-the-wild to see how all wireless networks will hold up under pressure.

There are two weeks of Olympic Games left, and hopefully the Olympic IT crew can quickly load test and resolve any problems.

It’s old news that a slow loading or buggy sites will drive users away, but users are apparently intolerant of sites being down for maintenance as well. That’s the finding of a recent survey conducted by 1&1 Internet , an UK analyst firm. According to a report by PCWorld,  1&1 Internet surveyed 1,500 web users and found that, overall, “consumers expect sites to be available 24/7.” And people are just getting more demanding – 44% of survey participants are more intolerant of web errors today compared to five years ago, and 40% said they wouldn’t return to a company because of a faulty website.

So what are the top five biggest complaints about websites?

1. Slow running websites – 71%
2. Sites down for maintenance  - 49%
3. Broken links – 44%
4. Online orders that freeze - 44%
5. Broken website pages - 42%

And if consumers can’t use your site, they WILL use a competitors.

Frustrations with a business website meant over two-thirds of those surveyed had left the site and gone to a competitor’s. …

Read more…