The funny thing about monitoring is that there’s a ton of data to collect, but few people know where to start. A few months ago, Steve Souders, one of the co-chairs of Velocity, asked us if we’d teach a workshop to try and fix this.

It’s a reworking of many of the things we cover in the book, plus an attempt to explain the math and reporting in an accessible way. Here’s the slide deck.

Related posts:

1. [Synthetic and Real User Monitoring] Knowing When Things Go Wrong
2. Slides from performance and KPI webinar

We had a good discussion about performance and its impact on KPIs like analytics and conversion with Strangeloop this week. Here are the slides, available for download or viewing, on Slideshare.

Related posts:

1. Proof that speeding up websites improves online business
2. Guest Post: How I increased traffic 1,176% in 24 hours

Do faster web pages mean better business? Definitely. We’ve seen hard evidence from major web operators like Shopzilla, Google, and Microsoft. But what about other websites? How big an impact does performance optimization have on the business metrics of a typical media or e-commerce site?

Here’s some concrete data on how reducing latency changes the key metrics, such as bounce rate, pages per visit, conversion rate, and shopping cart amount. It’s a pretty detailed discussion, but it if you want to understand the ROI of improving web performance on your site, dig in. If you want to read this more easily, here’s a PDF.

Enough theorizing: Page latency affects your business

Years ago, the need for web performance was anecdotal. Researcher Mihaly Csikszentmihalyi had, years before, shown that human beings are more engaged, and more likely to enter “flow states”, when response to their actions is immediate (see this article for more details on flow and web optimization.) But while it seemed like a good idea to make a page load quickly — and companies like Zona Research made headlines with their 8-second rule — we didn’t have any empirical evidence to that effect.

There’s no longer any debate. There’s reliable, reproducible evidence that web page latency is directly tied to the bottom line. At Velocity, Microsoft, Google and Shopzilla made this abundantly clear in a series of awesome presentations: detailed, controlled testing proves that slower pages hurt the bottom line. In Google’s case, adding delay reduces the average number of searches a visitor does each day even after the delay is removed.

Microsoft, using data from Bing, showed that slow pages affect other KPIs:

In other words, if your website is slow you’ll get:

• Fewer search queries per user
• Less query refinement
• Less revenue per visitor
• Fewer clicks, and lower satisfaction
• A longer time for visitors to click something
• Fewer searches per day
• Lower search engine rankings

Ouch.

There’s a great writeup of all this at Artzstudio, and we covered Marissa Mayer’s talk on Bitcurrent. My only complaint was that the sessions all left me with questions about how this applies to smaller, more targeted sites.

So what about sites we mortals run?

Generally speaking, there are a few KPIs that really matter. These measure attention (how many people find out about you), engagement (how much they interact with your site), and conversion (whether they do what you wanted, and how much you benefit from it.) Here’s a simple view of some of those KPIs. Different KPIs matter as visitors turn into customers, buyers, or enrolled users.

In this diagram, inbound traffic either consists of returning or new visitors. Some of those visitors leave (“bounce”) immediately, while others continue on to other pages, spending time on the site, creating content, and viewing ads. On an e-commerce site, a subset of those visitors convert (by doing something we want them to) and there’s a value to those conversions.

The question, then, is how does a faster page load time affect metrics such as these?

Comparing performance optimization in analytics

At Interop Las Vegas, I had dinner with the guys from Strangeloop Networks (disclaimer: I have friends who work there). They make a web acceleration appliance that speeds up page load times. It occurred to us that they were in a perfect position to tie page performance to analytical results, because their appliance could actually modify pages on their way out, on a visitor-by-visitor basis. So if they only optimized some of the visits, and marked them as such, they could later compare the business performance of optimized and unoptimized segments.

The Strangeloop guys ran sprinted with this idea, testing a couple of different types of site, including media and transactional businesses. Here’s what they found, which they’ve agreed to share with us provided we keep certain details confidential.

Total visits

The system was set up to optimize half of the visits and leave the other half untouched. But when they analyzed the results, that’s not how things looked in Google Analytics, where significantly more optimized sessions were recorded. In all, roughly 14,000 visits were segmented within the analytics package; but the number of optimized sessions that were recorded was significantly higher.

This may be due to problems with the analytics scripts running on slower connections, or it may be a sign of increased page abandonment before analytics has a chance to load. Either way, it’s worthy of more study.

Percent of visits from new visitors

Optimization seems to have an effect on the number of new visitors to the site, too, though it’s not clear why this happens.

We’d expect that since new visitors’ browsers have more to load — there’s nothing in their cache yet, and so page load times are higher — there would be fewer aborted visits for new users, which in turn would increase the relative number of new visitors to the site. But that’s not what happened.

Bounce rate and optimization

The next metric we looked at was bounce rate, that is, how many visitors left quickly from the first page they saw. Faster pages delivered a lower bounce rate than slower ones.

This seems intuitive, since when a page loads quickly you’re less likely to leave out of frustration. That extra percent of visitors who stick around turns into more opportunities to sell something or otherwise engage a visitor.

Engagement: Pages per visit and visit duration

Optimization also affects the number of pages a visitor views. When the site is slow, people read fewer pages.

It’s not clear from the pages-per-visit metric alone whether this is because people spend a finite time on a site, and simply get bored after that time, If that were the case, we’d expect a faster-loading page to result in more pages per visit simply because more pages could be loaded before the visitor got bored. A second KPI, Average time on site, clarifies this. Optimized visitors to the site spent 7 minutes more than unoptimized visitors. So it’s not so much the pages per minute of time on the site, but the actual number of minutes, that increases.

This is great for visitor engagement; if you’re running a media site, it also means a chance to deliver more impressions and make more money per visitor.

Impact of optimization on e-commerce results

But what about e-commerce and retail? Strangeloop instrumented a second site in the same way. The beauty of tagging visits up front is that the actual business outcome of that optimization can finally be quantified. In this case, optimization resulted in a 16.07% increase in conversion rates and a 5.50% increase in average order value.

We can’t share the actual conversion rate and order value amounts, but what these numbers do is allow you to actually quantify the ROI of a performance improvement investment.

Want to learn more?

There’s lots of good data behind these results, which we’ll be looking at in more detail in a Webinar on October 8 at 2PM EDT (you can sign up online at bit.ly/perfwebinar). Strangeloop will also have some data on how much performance improvement visitors experienced by then, and you can ask Hooman Beheshti (their VP of Products) and I questions about the experiment if you want to know more.

Some caveats:

I’m always wary of presenting vendor-specific data, because we try to remain impartial. Strangeloop isn’t paying me to talk about this, and I decided to cover it because it’s useful to the web monitoring community and I asked the question in the first place. I’ve reviewed the information fairly closely and have good reason to trust Strangeloop (for one thing, their VP of products, Hooman Beheshti, is a sometime contributor to Bitcurrent and an expert on web performance who reviewed Complete Web Monitoring and gave us detailed feedback.)

Instrumentation happens as follows:

• Every visitor who requests a page gets a segmentation cookie regardless of who they are. Since Strangeloop’s technology takes advantage of certain features in more modern browsers, not every visitor who is accelerated will benefit from the same performance improvements.
• The numbers reported in the analytics package (Google Analytics) are a result of the segmentation cookies seen, which is tied to how many visitors’ browsers made the analytics request.
• The reporting Javascript within the page sometimes can’t find the cookie that was set (this happens about 5% of the time). This could be a consequence of security restrictions, browser limitations, and so on. However, the “cookie not found” errors occur relatively evenly across optimized and unoptimized visitor, so they don’t distort the numbers.
• Google Analytics’ IP filtering was used to block out internal users, which might distort numbers too.

Related posts:

1. Slides from performance and KPI webinar
2. What business are you in?

Disclaimer 1: All site-related data found in this post comes from compete.com.  The company was kind enough to give us a “pro account” to help us research the O’Reilly book that we wrote called Complete Web Monitoring (thanks, you rock!).  However, compete.com did not sponsor this post (nor did any company, for that matter).  And yes, we know – compete.com numbers are simply estimates.

Disclaimer 2: I (Sean) worked for Akoha as Community Gardener while we launched at TechCrunch50 2008; but I’m now doing metrics, web analytics, performance, and social computing consulting.  The views found below are mine, and do not reflect those of Akoha in any way.  For the record, Akoha is awesome!

About us: This post was written by Sean Power with Alistair Croll.

Dear TechCrunch50 Startups,

Congratulations. You made the list. You’re finally launching, and that pent-up frustration of not being able to tell people about it for a month is almost at an end. Now, you have to live with a weekend of cold, hard fear that your demo will explode. You’ve got an interesting week ahead, and I know you’re short on sleep, so let me get to the point quickly.

You’re probably excited about the TC50 bump. I first saw the term used by Josh Kopelman of First Round Capital on the RedEye VC blog. The bump refers to the pounding your website is about to experience from TC50 attendees, readers, bloggers and their friends.  It’s not to be underestimated.  Here’s a glimpse at how the bump looked like for all TC50 startups in 2008.  If you squint a little, you’ll see Akoha somewhere in there!:

This is an unprecedented influx of attention. It may be the single biggest traffic spike you’ll ever experience. Thousands of visitors will drive by your site, stay for a minute, and leave — never to return. After the bump, you’ll feel a tremendous rush of adrenaline, then deep, soul-sucking disillusionment as your traffic dwindles back to its former levels.

Don’t waste this opportunity. If you take the right steps, you can make the most of your fifteen minutes of fame.

Being Talked About Is Nothing. Being Remembered Is Everything,

You’re probably celebrating the traffic you’ll get to your site. But you shouldn’t be, unless you capture the minds and hearts of your visitors.  The best way to do that is to target them.

On a normal day, you know very little about your site traffic. But today, you know lots. And that means you can tailor the experience to your audience. Let’s say you knew for a fact that 95% of your visitors loved the color red.  You’d probably redesign your site to make it bright red, right?  Well, next week, over 85% of your visitors will all come to you through TechCrunch somehow. Think about it. For a single week, you know exactly what kind of visitor will be visiting you.

TechCrunch50 visitors will load your site, scan quickly, and may even register (if registration isn’t broken, and is simple enough). Then they’ll leave.  In that crucial 10 seconds of play time, your job is to make sure that as many people remember your site as possible, and that you gain the ability to reconnect with them later, and to engage with them.

During TC50 2008, Robert Scoble blogged about Demo and TC50 websites as “sucking big time”. Knowing that he was an important influencer, we put his face on the front page of Akoha, with a speech bubble saying “If your name is Robert Scoble, click here!”, which brought the user to a small blog post explaining where we were going with our site design.

It worked. Scoble reciprocated in kind:

Being memorable is everything at an event like TechCrunch 50.  Find ways to be remembered by those who visit you.

Gathering Data

Since you’ve just launched, now is the time to start figuring out how unsolicited users react to your site. You’ve got a great lab for a few short, sweet days. Try a couple of site designs, and see which one works best.  This is where simple abtests becomes useful. Make sure you target people specifically. And turn on all the monitoring you can (without making the site slower.) It’s an excellent time to make baselines.

Google Analytics Is Not Enough

It really isn’t. For one thing, it doesn’t show you results fast enough in a flash crowd situation. Consider a more real-time analytics tool like Getclicky or Woopra, so you can see traffic as it happens — not the day after. Then find someone to watch it. While you’re on stage, or schmoozing investors, have someone back at headquarters (or whatever you call your mom’s basement) looking at the traffic.

But even if you have up-to-the-minute analytics, it’s not enough. It’s one thing to know what your users did, but chances are that you don’t even have goals set up in Google Analytics, and even if you do, you’re only measuring if people reached the goals, not why.

Measure Peoples’ Behaviors

Take a look at your site right now.  In two years, it will look completely different, because you’ll learn. What if you could speed up that clock? You can, if you measure what people do on your site and iterate rapidly.

A Web Interation Analytics tool measures where people click (or don’t), helping you figure out where your design flaws are without even asking people.  Go install CrazyEgg or ClickTale right away.  You can’t afford to miss out on this data. If you’re really serious, put Coradiant or Tealeaf in front of your site.

Ask People What They Think

People are weird. They’ll do all kinds of things you don’t expect. It’s one thing to watch what they do — but often, the real key to unlocking your business potential is to know why they did it. The simple way to find this out is to ask them. Do yourself a favor and set this up now.  Install 4Q, Kampyle, GetSatisfaction or UserVoice right away on your site.  Have the guts to ask what people like and don’t like about your site so you can fix it faster than your competitors.

And don’t just rely on tools. Be on Twitter, and be reachable. When that girl/guy in your basement isn’t watching analytics, have him or her respond to people online. When Alistair & I present at conferences, we take turns teaching and reading Twitter — fielding questions from the audience, responding to folks who couldn’t be there, and seeing if the folks in the back of the room can hear us. Do the same thing for your site.

Make Sure Your Site Can Weather The Storm

If your site is slow, people will leave.

This isn’t just a platitude: Google and Microsoft have both released empirical evidence that latency correlates with departure, and reduces conversion rates. Worse, people who had a slow experience will use the site less even when it gets faster again.

Bounce rate (the number of people that see one page and leave) may be a function of site latency. So test it! Call Keynote, Gomez, AlertSite or WebMetrics and start testing your site’s speed and availability every 5-15 minutes. Still haven’t got funding? Consider AlertFox or pingdom.  Are you at TC50, trying to load your site and it’s not working?  Not sure if it’s your computer, TC50’s Internet or your site?  Use AOL Page Test to test it remotely and find out why it’s sluggish.

Don’t forget to test the key pages and functions you want people to use, such as enrollment.

Have a Simple Call To Action

You’ll have a lot of folks on your site. Make it painfully obvious what you want them to do, whether that’s signing up, giving you permission to contact them later, trying it out, or telling friends. Don’t give them several choices — give them one. If 95% of people are just there to find out what you do, tell them. You can show them press releases or a list of your board members later. And consider tightening up your site copy, too. Less words is good.  Still not sure?  Talk to Josh Porter.  He can help.

Listen and Learn

Ultimately, you want to hear what people are saying about you elsewhere — not just on your site — and respond to the criticism and compliments.  Call ScoutLabs, TechRigy or Radian6 right away.  In most cases, all you need is a credit card and off you go.  Set up searches for keywords that matter and find conversations that concern you.  Twitter search is your friend. Google Alerts are too. There’s no better way to make a first impression than by actually being there to make a first impression.

Running The Booth?

Read this excellent post by Jason Calacanis on the subject.  It will help you avoid disillusionment when you come back from the launch party.

Learn From The Jedi Masters

Once TechCrunch50 is over, you’ll have very little time to rest on your laurels.  Chances are that your startup will die in the next year.

(Don’t die!  We don’t want you to die!)

Learn with the masters instead.  Follow Eric Ries, Dave McClure and Avinash Kaushik religiously.

Come And See Us In November (Or Call Us Before)

It’s awesome that you’ve come this far. Hopefully something in this list will help you learn from the storm, and turn the bump into a ramp. Whatever the case, once the dust settles you’ll have a lot of data to dig through. We’d love to help.

If you’re a finalist at TechCrunch50, we’re impressed that you’ve read all the way up to here.  Thanks!  We know you have so many things to do this weekend, and we appreciate that you’ve given us a few minutes of your time.

We feel for startups, we love ‘em!  So we’d like to make you an offer.  Once the next couple of weeks are done, give us a call.  We’ll spend an hour with you on the phone or online, gratis, and see if we can help you sort out your data, no strings attached.  You can reach Alistair and myself on Twitter, or simply email me directly – sean at httpd dot org.

We’ll also be doing an eight hour bootcamp called “Communilytics” at Web2Expo in New York in November on the subject of community metrics; if you’re coming, let us know.

Good luck out there, and knock ‘em dead!

- Alistair & Sean

ps; For those of you curious to know what sort of traffic the TechCrunch 08 attendees received, you’ll find the result of our findings below.

The Bump By The Numbers

When analyzing the TC50 08 finalist sites, the numbers from compete.com can be misleading.  For most sites, the traffic volume is relatively low, and therefore more prone to inaccuracies due to small sample size.  Consider that a disclaimer.  Averages were calculated between the months of October 08 and August 09 in order to avoid skewing the data as a result of the TechCrunch Bump.

It would be useless to lump all the sites in one bucket.  Grouping transaction and collaboration sites side by side is like admiring the similarities of hippopotamuses and cream puffs.

To get better insight, we must segment into the four types of sites found on the Internet.

Executive Summary

• Media sites showed relatively large traffic volumes, with unique visitor count often going up and to the right.
• Collaboration sites showed consistent traffic patterns from the month of October to august, but had no month to month average growth.
• The TechCrunch bump for transaction sites represented a small portion (8%) of the total unique visitor count they would receive for the next 11 months.
• The TechCrunch bump for SaaS portal on average represented roughly the total amount of traffic they would receive for the next 11 months.

Media Sites

These sites offer content that attracts and retains an audience. They make money from that content through sponsorship, advertising, or affiliate referrals. Search engines, AdWords-backed sites, newspapers, and well-known bloggers are media properties.

Between the months of OCT 08 (one month after TechCrunch50 ‘08) and August 09 (one month before TechCrunch50 ‘09), media site finalists:

• received, on average, 1,340,000 unique visitors (860,000 unique visitors with outliers removed (MIN and MAX)).
• showed peaks with a 41% increase in unique visitors, a positive influx of 148,000 unique visitors for that month.
• showed valleys with a %29 decrease in unique visitors, a loss of 171,700 unique visitors for that month. (ouch!)
• with the highest unique visitor count (MAX) showed an average monthly growth of 21.33% from 84,600 unique visitors totaling 3,530,000 unique visitors.
  
• with the lowest unique visitor count (MIN) showed an average monthly growth of 2.57% from 16,800 unique visitors totaling 99,600 unique visitors.

On average, the TechCrunch bump brought 57,900 unique visitors (55,300 unique visitors with outliers removed) for the month of September 08 and represented 8.55% of the total traffic they would receive in the following 11 months (8.17% with outliers removed).

Transaction Sites

A site that wants visitors to complete a transaction—normally a purchase—is transactional. There’s an “ideal path” through the site that its designers intended, resulting in a goal or outcome of some kind. The goal isn’t always a purchase; it can also be enrollment (signing up for email) or lead generation (asking salespeople to contact them), and that goal can be achieved either online or off.

Between the months of OCT 08 (one month after TechCrunch50 ‘08) and August 09 (one month before TechCrunch50 ‘09), transaction site finalists:

• received, on average, 272,000 unique visitors (124,000 unique visitors with outliers removed (MIN and MAX)).
• showed peaks with a 700% increase in unique visitors, a positive influx of 65,700 unique visitors for that month.
• showed valleys with a %49 decrease in unique visitors, a loss of 23,700 unique visitors for that month. (ouch!)
• with the highest unique visitor count (MAX) showed an average monthly growth of 15.49% from 113,800 unique visitors totaling 2,916,000 unique visitors.
  
• with the lowest unique visitor count (MIN) showed 0 growth and traffic.  (oof).

On average, the TechCrunch bump brought 35,300 unique visitors (26,600 unique visitors with outliers removed) for the month of September 08 and represented 8.83% of the total traffic they would receive in the following 11 months (6.64% with outliers removed).

Collaboration Sites

On these sites, visitors generate the content themselves. Wikis, news aggregators, user groups, classified ad listings, and other web properties in which the value of the site is largely derived from things created by others are all collaborative.

Between the months of OCT 08 (one month after TechCrunch50 ‘08) and August 09 (one month before TechCrunch50 ‘09), collaboration site finalists:

• received, on average, 138,000 unique visitors (86,400 unique visitors with outliers removed (MIN and MAX)).
• showed peaks with a 316% increase in unique visitors, a positive influx of 29,500 unique visitors for that month.
• showed valleys with a %63 decrease in unique visitors, a loss of 24,600 unique visitors for that month. (ouch!)
• with the highest unique visitor count (MAX) showed an average monthly growth of -1.02% from 108,000 unique visitors totaling 997,000 unique visitors.
  
• with the lowest unique visitor count (MIN) showed an average monthly growth of -50.70% from 1,400 unique visitors totaling 5,000 unique visitors.

On average, the TechCrunch bump brought 37,600 unique visitors (26,500 unique visitors with outliers removed) for the month of September 08 and represented 47.23% of the total traffic they would receive in the following 11 months (29.58% with outliers removed).

Software-as-a-Service Sites

These sites are hosted versions of software someone might buy. SaaS subscribers expect reliability and may pay a monthly per-seat fee for employees to use the service. Revenues come from subscriptions, and a single subscriber may have many user accounts. On some SaaS sites, users are logged in for hours every day.

Between the months of OCT 08 (one month after TechCrunch50 ‘08) and August 09 (one month before TechCrunch50 ‘09), software-as-a-service site finalists:

• received, on average, 62,000 unique visitors (47,000 unique visitors with outliers removed (MIN and MAX)).
• showed peaks with a 337% increase in unique visitors, a positive influx of 18,200 unique visitors for that month.
• showed valleys with a %65 decrease in unique visitors, a loss of 15,400 unique visitors for that month. (ouch!)
• with the highest unique visitor count (MAX) showed an average monthly growth of 1.52% from 26,000 unique visitors totaling 250,000 unique visitors.
  
• with the lowest unique visitor count (MIN) showed an average monthly growth of 64.46% from 1,400 unique visitors totaling 5,500 unique visitors.

On average, the TechCrunch bump brought 23,800 unique visitors (21,800 unique visitors with outliers removed) for the month of September 08 and represented 100.23% (!!!) of the total traffic they would receive in the following 11 months (91.81% with outliers removed).

Notes:

Big thanks to wa7iut for helping me with the stats!

Alfabetic.net was not included because it was impossible to determine what kind of site alphabetic.net is (site is down, nothing on Google cache or archive.com and the ustream demo doesn’t make it clear).

We used a combination of fotopedia.com and fotonauts.com data due to a site rebrand.

footnote.com was not included in the analysis.  By the time that they were presenting at TC50, they had 1.3million unique visitors a month, with 1.4 million unique visitors in December 2007.  Statistically (and from a TC50 perspective), footnote.com was too much of an anomaly.  We do, however, think that their site is totally cool.

Related posts:

1. DemoCamp Guelph

We’re doing a presentation that’s excerpted from the book at DemoCamp Guelph tonight. Should be an interesting conversation; we have an “exercise” planned. Sean can’t be here (he was at Podcamp and has to get real work done after a weekend of editing the 400+ figures in the text!) but will be joining on Twitter. If you have photos from the event, or questions for Sean, we’ll be using the #CWM hashtag (for Complete Web Monitoring, the title of the book.)

One of the projects we’ve been working on is trying to create a single, comprehensive overview of the Complete Web Monitoring process. Here’s where we’re at (and an early glimpse at a poster we’re working on.)

First of all, a complete monitoring strategy includes the many questions a web analyst needs to answer:

• Web analytics (“what did they do?”)
• Web Interaction Analytics (“how did they do it?”)
• Voice of the Customer (“why did they do it?”)
• Both synthetic and real user performance monitoring (“could they do it?”)
• Community monitoring (“what are they saying?”, “who’s talking?”, and “where are they saying it?”

Any strategy also has to look at several different stages in monitoring:

• Arrival (“I visited the site”)
• Usage (“I played with it”)
• Engagement (“I’m a part of it”)
• Revenue (“I paid for it”)
• Referrals (“I spread the word”)

If these look somewhat like Dave McClure’s Pirate Metrics, it’s because he’s awesome and we borrow heavily from his thinking on startup metrics. Anyway, this PDF is a work in progress of trying to align the big questions analysts need to answer with the various stages of visitor engagement. Once we sex it up a bit, we’ll make some posters.

I’ll put the DemoCamp slides up here shortly.

Related posts:

1. Slides from DemoCamp Guelph 08
2. An Open Letter To All TechCrunch50 2009 Startups: The TC “Bump”, What It Really Means & How To Navigate It

Uh oh.  Is the site is down?

Yahoo! site inaccessible

Site downtime is rare these days, but it still happens, and when it does, thousands of people can be affected.  But how do you know that an entire web property is down, and that it’s not just down for you?  How can you figure out who’se affected by the outage?

I tried to load Yahoo! a few moments ago, but I was unable to do so.  To figure out if it was just my machine that had an issue, or if others were affected as well, I did the following:

First, I launched tests from webpagetest.org and Gomez’s free synthetic monitoring service called Gomez Info-Center (since they’re free tools, I used both to make sure that I was getting a second opinion).  I use these services to understand if I’m the only one experiencing problems with websites.  If a site is slow, or if I’m getting “Address Not Found”, it’s the first place that I’ll go if there’s nothing obviously wrong with my own Internet connection.

Here’s what webpagetest.org told me:

webpagetests.org results for Yahoo.com

Weird stuff!  At least I know I’m not the only one.  This was further confirmed by a few twits that I received confirming that people were experiencing mixed results when accessing Yahoo!.

Twitter response

So is it *just* Yahoo!, or might there be other sites affected as well?  I went over to Keynote’s Internet Health Report to have a look.

Keynote's Internet Health Report

I didn’t see anything yellow or red, indicating that all backbone providers seem fine.  So chances are that the problem is local to Yahoo!, or Yahoo!’s data center.  If I had to guess, I suspect that they had some momentary DNS issues, because sure enough .. after a few minutes, Yahoo! was back up and kickin’:

Yahoo! is alive!

Sure enough, as I was writing this blog post, an alert from the folks at AlertSite kindly informed me that something happened in Washington DC.

AlertSite notification - way to go, ServerBeach!

This isn’t really a post about Yahoo! being down (chances are that the Yahoo! data center closest to me was in Washington, the same data center that AlertSite was announcing availability issues with).

Incidently, I actually use Yahoo!’s main landing pages as a benchmark for success.  Their pages load quickly, and their availability (uptime) is generally excellent.

This post is about something much more important.  How can you know if your blog/site is slow, or even worse, simply not loading for others all around the world.

We’re currently in the process of getting the synthetic monitoring and the real-user monitoring chapters reviewed; they’re almost done!  They talk about two things:

1. Making sure your site is accessible from around the world
2. Making sure that your site is accessible for all of your users

Synthetic Monitoring

This technology checks your site for you at intervals (every minute, five minutes, fifteen minutes, hour, whatever), and gather all kinds accessibility (is the site up?) and performance (how fast did it load?) stats, from many different types of machines located all around the world.  You can tell synthetic monitoring technologies to alert you as soon as something happens that you don’t want (like a page not found or taking more than X seconds to load).

Thanks to the distributed nature of many synthetic monitoring platforms, you can also figure out if certain parts of the world are having more issues accessing your site than others (which is what I suspect happened in Yahoo!’s case – the west coast seemed fine, but those of us in the east seemed affected).  In fact, if Yahoo!’s downtime wasn’t expected, I’m positive that some form of automated alert informed them that they were experiencing issues.

Real User Monitoring

But sometimes, it’s not enough to test from a whole bunch of desktops around the world, because they don’t necessarily represent real traffic – the actual traffic that you have on your site.  That’s where RUM comes in.  RUM tools monitor every single transaction on your site, and tell you how fast (or slow) transactions were for your users.  RUM data is extremely insightful.  For example, you can figure out exactly how many users in NYC have page load times over 9 seconds.  Maybe you have a widget on it being served from Australia that is dramatically reducing performance for users in North America.

We show you what you need to know to implement these kinds of solution, and go into many of the pros and cons of these technologies; whether you’re a blogger or an SVP of Operations for a Fortune 500 company.

In the meantime, if you haven’t seen this video, load it next time your favorite website is down. http://www.thewebsiteisdown.com/

No related posts.