I have a terrible habit. Every morning, I get on the scales to check to see if I’ve lost weight or fat. In fact, the habit is so bad that sometimes I’ll get on two or three times a morning and even through the day. Measurement is fine and good but sometimes daily measurement of small or difficult to measure changes can be frustrating and inaccurate (unless it says I lost weight – then it’s accurate).

The other day I noticed that by simply allowing 30 minutes to lapse without doing anything, my body fat percentage dropped by four percent and I lost 100g (apparently). Well, I thought, forgo food and activity and by lunch I should be at a decent percentage body fat and a pound lighter! Obviously my instrument of measurement was not reliable.

I had an experience with a web analytics tool which seemed to be under reporting for a site I was working on. I was making changes, seeing the effect in other ways (sign-ups) but analytics refused to budge. Analysis of the code showed some variations but apparently it was all OK. There was something wrong though in my opinion and an audit of our raw log file stats illustrated in tangible terms what I had been struggling with for weeks – the measurement was indeed flawed.

Relying on web analytics has become something of an addiction among some businesses. With constant feeds of web traffic, the minute-by-minute tracking of where traffic is going to on their site and pages ranked by popularity real-time, web analytics are as essential as their other addictions – coffee, toast and antacid tablets. The problem with this is that most web analytics are wrong and are misreporting statistics. Add to the mix a misunderstanding of some of the analysis of the data and it’s a recipe for stress.

There was a comparison of the various statistics packages which showed the variation among the packages available in 2007 and how differently they reported on the exact same website. In the report Stone Temple assert that “as Jim Sterne is fond of saying, if your yardstick measures 39 inches instead of 36 inches, it’s still great to have a measurement tool”. I would argue that some statistic packages are not continuously measuring 36 inches as 39 inches due to cookie reliance, poor implementation and other factors and thus are not a proper measurement tool in those conditions.

How can web business operate in an environment where reliance on web analytics has to be tempered with something else? Using more than one measurement tool can help. With the former Urchin product, now Google analytics, now available for free it is easier than before to run two statistics packages. I would argue that someone in your organization should make web analytics their project. If possible, train up someone to dedicate at least an hour a day to web analytics. The more this web analytics person understands, the more appropriate use your company can make of sometimes flawed analytics.

Web Analytics are not the be all and end all of measurement online. I’d argue that a holistic approach needs to be taken and that online is the same as offline – it’s simply not conducted face to face. Footfalls are not the only measurement of a shop’s success and nor is unique users. Sales are not the only measure of a company’s success and nor are sales online. Websites are more than just online brochures, they are a way for people to engage with your brand.

I’ll still get on the scales every morning, but Sunday is the day I use now to benchmark my weekly weight loss. Diet, exercise and health levels all cause variance in my scales ability to correctly measure me. Your website’s health, code implementation and use of first or third party cookies (the code kind, websites don’t eat cookies) will determine the accuracy of your web analytics. Don’t rely on analytics alone, find an internal champion and make sure your website code is correct for a happy, healthy web analytics experience

I read an interesting article the other day which discussed the importance of targeting and bounce rates and it got me to thinking about the clients that I have had, past and present and the buckets of money I have seen thrown away. So, I decided to put a list together of 10 ways you or your client may be wasting money they could be paying you:

1. Stop sending paid traffic to your home!
   1. Your homepage is rarely the page your potential customers are looking for. If I am searching for “buy sony flatscreen 1080p” PLEASE send me to a page with sony flat screen tv’s on it!
   2. Check your homepage’s bounce rate. It is 60% than you should only be sending people searching for “sitename.com” to that page otherwise, you are throwing at least 60% of your money away. Why do I say at least? Because if you are sending all your paid traffic to your homepage than you probably haven’t optimized your paid search ads and are just looking for a high click through rate.
2. Stop duplicating terms in your paid search accounts
   1. It never ceases to amaze me when clients are using the same keyword in multiple ad groups with the same geo-targeting and timing. Don’t do this
3. Have goals for your website
   1. Set goals for your website. “I want to make money.” Isn’t a sufficient goal. Come up with trackable metrics to get you there so you can optimize your website to eventually make you some more money.
   2. Publishing sites - # of page views, repeat visitors, high CPM.
   3. Ecommerce – lifetime value
   4. Lead-gen sites are a monster in and of themselves. If you generate leads so that they become sales then focusing on the lead only is going to inhibit you from making money. Remember, unless you are selling leads, they don’ t make you any money until they become a sale.
4. Don’t cut the long-tail
   1. If you aren’t tracking performance over-time, you are probably missing all of your long-tail terms. Don’t cut terms that are attracting people in the educational phase of the sales-cycle. If you market to them differently you can increase your conversion rate over time. Don’t cut them just because these phrases take a little longer to convert - MARKET TO THEM DIFFERENTLY. These are long sales cycle leads
5. Use negative keywords in your paid search campaigns
   1. This may sound basic, but if you aren’t using negative keywords than you are paying for traffic that will NEVER make you a dollar, pound or euro. If you have paid-search clients or use paid search and using mostly broad match than you really need to spend some time investigating exactly what terms are being clicked on.
6. Learn what terms lead to conversion and optimize your site accordingly
   1. Welcome to the beauty of the internet and web analytics. You can actually tell how people get to your site, what they do there and if they are making you money. It’s brilliant. So why are you optimizing your site based on what you think will work well?
7. Stop making creative decisions based on what your boss thinks is a good idea.
   1. See point 6. We have the ability to test and optimize, which means we can run multiple ad creatives (banner and search) and see what drives the greatest revenue rather than based on the CxO’s affinity to the color blue. Back your decisions with data
8. Test, test, test and SPLIT TEST!
   1. See point 7. Products like Google website optimizer and Omniture’s Test and Target (formerly Offermatica) can optimize your landing pages based on conversions. There is no excuse for not testing your landing pages and assuming that your designers first try is always correct. In all likelihood it is not optimal and assuming it is will make an ass out you but not me because I’m split testing
9. Track how changes to your website affect your bottom line
   1. If designers, programmers, your kids or whomever are making website changes that affect your business– especially your homepage – you should be recording that this is being done and how this is affecting organic placement and conversions. If there is no tracking there is no accountability.
10. Know what your users are clicking on
    1. Heatmaps and click analysis tools, like Crazy Egg are great at letting you know what your users are clicking on and responding to. Using this information to optimize page layout should help you to make the most of the traffic that you have.

If you’ve ever extolled the virtues of the total accuracy of online metrics to anyone, you’ve been guilty of lying and you’re going to burn in hell. I’ll be there too though, so try and find some good indie music and maybe some cheese dip and try and drag those along.

In web metrics, discrepancies abound. If you’ve ever run two sets of metrics on the exact same data, you’ve most likely received two totally different sets of numbers. There are reasons for that, of course, but they may not make sense to someone who isn’t overly technical (or bright, like many of my past clients.) So when we’re telling someone that yes, he or she should funnel more cash into online marketing because it’s really easy to get accurate stats, what we’re really saying is that yes, he or she should funnel more cash into online marketing because the stats are probably more accurate than offline measurements, but they’re still not REALLY really accurate.

Like many of our SEO arguments that we make to each other and to our clients, the accurate measurement argument is a relative one. You obviously cannot accurately measure the number of viewings of a highway billboard, for example, even if you knew the exact number of people who’d driven past it in a certain period of time. If any of those drivers/riders were like me, they’d have been paying strict attention to the immediate dangers of highway life (bad song on the radio, band of wandering midgets who’ve gotten loose from the circus, those dusty plant things that roll down the streets in Westerns, etc.) You may have circulation numbers of a periodical but you have no clue as to whether the subcriber even opened it. You can’t tell when someone’s turned down the volume on the radio when a commercial is aired.

Online measurements, at least, are a bit more easy to conduct, since a user has to do something in order to trigger his or her action to be recorded. However, there’s quite a lot that goes on with the programs used to measure this raw data. Depending on your session length, for instance, you could count a visitor once or twice. When I used to, ahem, cloak my sites, I realized that the analytics program I used was counting each hit twice, counting the cloaked page once and the actual page served to the user once. Another analytics program I used on the exact same sites did not do this, though.

So, if we can agree that web metrics are MORE accurate than most other forms of marketing measurements, what difference does it really make if they’re still potentially not actually all that accurate? If we’re used to estimates from keyword software and we use those as a guideline for how to best utilize our content to capture the highest ranking spots and the most converting users, why can’t we be happy with inaccuracy in terms of how many people visited our site this month?

The New York Times has a great article about the dangers of online web metric inaccuracy, one which lays out the major problem with this: advertiser money. The following bits are the most telling…

“Many advertisers pay Web publishers each time their ad gets an impression, meaning that it is viewed by a reader, but each company uses its own methodology to count impressions.” So think about it: if you’re being told that your ad has been viewed 10 million times, you’re going to be charged for that. What if it had truly only been viewed 8 million times? You’re paying for 2 million views that never happened.

For major advertisers being lured by sites with the premise that their viewership is massive, this kind of discrepancy can also add up to tons of lost money, whether it’s being lost by inflated stats or by the promise of a high readership that turns out to be untrue. Speaking about site owners attempting to gain major advertisers, Gian M. Fulgoni, the chairman of ComScore, states that “It’s in their interest to make their audience look as big as possible.”

“To make matters more complicated, consumers who delete cookies — small bits of computer code that track their online wanderings — are also overcounted by publishers’ servers, by most accounts. Some news sites have tried to improve their systems by asking their visitors to register, but many people refuse.” I delete my cookies all the time, especially when I get bored. I also refuse to register on most of the sites that ask for it just to be contrary. I’m obviously adding to the problem here but I see no way around that.

Should we have to settle for constant inaccuracy? Probably so, if we’re dealing with any form of marketing. I’m not really all that concerned if I get two sets of visitor numbers for the same site, but what DOES seriously concern me is that my conversions may not be accurate. Currently I run conversion tracking on a small set of paid ads for a client, and he can verify that those numbers are correct because the conversion is on a contact us form that he receives. Currently, that’s about the only bit of accuracy that I’m enjoying though. It would be nice if that changed, but I’m not too optimistic about it right now.

Bounce rate is a critical web statistic that can give you tons of valuable information about landing pages on your website. As it’s a simple percentage (based on how many users hit a page without going elsewhere on your site in a session divided by the total number of users who hit that page) it’s a measurement that doesn’t tax your brain too much (hopefully.) If you’re going to take the time to drive traffic to your site, the least you can do is make sure users don’t leave immediately. However, it’s not a clearcut statistic like you may think it is (and like I have kind of implied that it is, too.)

Avinash Kaushik states that “My own personal observation is that it is really hard to get a bounce rate under 20%, anything over 35% is cause for concern, 50% (above) is worrying. I stress that this is my personal analysis based on my experience, but hopefully it gives you a feel for what you are shooting for.” Reading these numbers filled me with fear and loathing, I must say. After looking at the bounce rate for a particular site that I shall refer to as When Animals Attack Magicians, I felt like I should be as busy as a cat covering it up. I also did a bit of exclamatory cursing, something that Southern ladies should not do.

What this metric tells you is that your page/site has the potential to entice, bore to tears, or infuriate a user. In the case of When Animals Attack Magicians, it seems that I’m mostly infuriating people. I feel like Johnny Rotten with only a slightly better attitude. But oh wait! When Animals Attack Magicians is a blog so how do we deal with that? As everyone knows, it’s quite easy to read a ton of posts on a blog without going to another page on the site…

Here’s where things get a bit tricky. If I’m on a blog and I read 15 posts, then leave without viewing anything other than that great big long page of posts, I am counted as a bounce even though I’ve read a good bit of the site. If I look at individual pages and check THEIR bounce rate, this may make more sense in terms of giving me an accurate depiction of whether or not people are offended as deeply as I am when I see George Bush on television talking about how necessary the war in Iraq is. However, if these bounce rates are high, could that simply mean that the person has already read everything else on the blog and has followed a link into the site only to leave for a reason other than simply finding fault with the site? Yes of course it can, but how are we to tell that from the this one metric?

Let’s take a non-blog site then, which we’ll call America’s Funniest Dancing Cats, and think about its bounce rate, which is possibly easier to accurately measure. A high bounce rate here could definitely mean that the landing page is as charmfree as Paul Weller’s latest offerings. Unless you’re running an iframe or you have all your content on one page, a high bounce rate here should trigger a closer look at WHY these people are leaving without delving deeper into the wonder that is your site about dancing cats. Maybe, on your home page, you had a non-funny dancing cat photo and people simply thought you were nuts, for example.(Like there IS such a thing as a non-funny photo of a dancing cat!)

In addition to the type of site that you’re working with, you need to pay attention to the TYPE of pages that have high bounce rates. With some pages you should probably expect a higher bounce rate. Let’s say that someone was searching for your contact information and went directly to the contact page. Unless you’ve cleverly replaced your actual contact information with a blurb about ministering to the poor, the user is getting what he or she needs and will possibly find no need to go anywhere else on your site. Thus, a high bounce rate here would not be cause for concern. It’s common for ecommerce sites to have high bounce rates on the final review page of purchasing, as well, as people realize that really, they do not need yet another ceramic lighthouse.

In closing, let me just say that while I have possibly confused you about a statistic that IS actually quite valuable, my point is simply that this is a metric that cannot be viewed on its own with no thought about what it actually represents. You may be doing really well with a blog that has a 70% bounce rate, so don’t be alarmed when you see such a high percentage until you actively look around and see what is going on. This statistic does not exist in a Dyson you know…so use it and bow to its importance, but only in conjunction with a bit of human analysis. Numbers are our friends, but you can’t remove the need for interpretation.

**UPDATE**

Avinash was kind enough to comment on this post and lead us to two more posts that deal with issues raised here…so read them. NOW!
Tips for Measuring Success of Your Blog
Standard Metrics Revisited

In the South we have a saying that “Even a blind hog finds an acorn now and then.” For all you Yankees and people who’ve never watched “The Dukes of Hazzard” or my personal favorite, “Mama’s Family”, this means that we’re all occasionally lucky. However, to consistently do well in SEO, you need more than simple luck. You might need to actually think a bit too, or get a smart friend.

It’s fairly common to do a bit of testing in PPC campaigns to see which keywords have the highest conversion or which ad title attracts the most clicks. It’s also quite easy to see this information in the ad interfaces, which may explain why those of us who are kind of lazy will actually make the effort to test here. Even if we’re not intentionally testing anything out, we can still see all the numbers and formulate a plan of how to make things better. This isn’t so easy with organic marketing.

I’m sure that you run web analytics on your sites. Considering all the good free packages out there, you’d be insane not to utilize all of the data that you can get even if you’re normally a cheap basta…person. So the question remains…what are you doing with all those numbers? Are you making adjustments to your campaign in the same way that you would if these were PPC statistics? If not, here’s why you should wake up and realize what a goldmine of information this data really is for your organic marketing. In particular, you’re being given two fantastic pieces of information that you’d be criminally negligent not to use:

Traffic sources: how are your visitors getting to you? It’s obvious enough when you’re talking about a paid ad in Google (because unless something is really wrong or you’re really not that bright, a user didn’t click on a paid ad in Google by searching in Yahoo), but the overall traffic sources for your site can tell you quite a bit about the areas that need a boost in visibility.

If you’re getting most of your traffic from referring sites, you might want to boost your search engine rankings and get your name out there for more direct traffic. If your site only does well because of inbound links to it, you’re too dependent on other people for your traffic. If almost all of your traffic comes from people directly typing in your URL, then congratulations on whatever you’ve done to get the word out (did Lisa make you wear an SEO Chicks t-shirt?) but you need to beef up your search engine rankings and your links most likely.

You get the point here: you’re never going to wind up with a totally balanced traffic source pie chart (ummmm pie…) but you really should use this information to figure out the areas that need the most work.

Search engines: which ones seem to like your site? It’s strange how you can rank in the top 10 for all your major keywords in Yahoo, for example, and still get all your search engine traffic from Google. Just because you do rank well in one engine does not mean that you’ll see this reflected in how users actually get to your site. And honestly, ranking well isn’t worth much if your target market isn’t using that search engine.

Here’s where you have to actually do a bit of work though…don’t just see which engines send you the most traffic. Go and check on your rankings and see if it matches up. If you find that you’re usually in the 100th spot for a keyword but it’s still getting tons of traffic for you, a lightbulb should go on in your head. If you get lots of traffic from ranking that low down, you should seriously invest some time into boosting your visibility for that keyword. This is like giving the Wolfman a locked room full of cute young coeds. Chances are, he’s gonna take advantage so follow his example. In this case, at least. Don’t go around scaring the bejaysus out of people though or let yourself get TOO hairy.

Traffic sources and search engines are such basic aspects of your overall data, but if you’re not someone who routinely uses web analytics to test out new things, you might want to start out slowly, so as not to cause any blood pressure spikes or Tourettes-style cursing. It’s not a difficult thing to do really, more like something new to annoy you. And here’s the kicker: you have to be prepared to actually USE the information you glean from all your testing. Use it to better your marketing efforts, or just to have a good laugh.

It can also be a bit obsessive for those of us who get ultra-competitive and crazy and refuse to ever play any board game with someone who’s at least 75% as intelligent as we are just in case we might not win. Luckily that is so completely crazy that I don’t know anyone like that. Once you’ve completed round one of “really basic analytics data that might help me” you will feel confident enough to move on to actually getting into serious analytical mode and then the sky’s the limit.

In closing, please note that I did NOT once mention Adam Ant in this post! The shock therapy seems to be working well. I’m switching back to Mark Linkous anyway.

Alexa.com is regarded by many people as an authority. However, if you’ve been reading a lot of SEO blogs recently you’ll know that many of us in the industry are fed up with this site and its utter lack of accuracy. Save everyone some time and read this Slashdot article for an excellent piece on just how bad Alexa really is. I couldn’t have put this any better.

So therein lies the issue of what else to use for internet measurements? Alexa may be the name that everyone knows, but it certainly doesn’t have a monopoly on this type of information. There ARE alternatives. OK there are TWO free alternatives that I found that seem even remotely decent (meaning they don’t show N/A for important areas and they give me numbers that I know are somewhat in the range of accurate), but here they are:

Quantcast
Quantcast bills itself as “the world’s first open internet ratings service” and it’s FREE (like anyone wants to pay for what may not be accurate data.) Quantcast gives you estimates (like Alexa does) but also allows more accurate reporting if you’re willing to paste a bit of javascript into each page on your site. Since you may already do this with Google Analytics, it might not irritate you but for those of you who are grumpy non-green monkeys, this might not be something that will make you put on some Scroobius Pip and moonwalk. The main issue with Quantcast is that it’s currently only applicable for US-based audiences. That’s not going to help my little nutbrown hare Mahoud over in London. The wording for the results IS quite nice though and gives you tidbits such as this:

“This site reaches approximately 24,640 U.S. monthly uniques. The site attracts a slightly male slanted, more affluent, skewing older following. The site’s visitors’ interests are specialty retail (Cost Plus, Pier 1, West Marine, Crate & Barrel), sports (PGA Tour) and television (DIY - The Do It Yourself Cable Channel, Fine Living Network, Home and Garden Television (HGTV), Food Network).”

Pretty cool information, but is it ACCURATE information? This was for a site that doesn’t use the javascript they give you and the numbers seem reasonable enough, but this site also gets 85% of its traffic from the U.S. so how does that affect the statistics? Whether or not Quantcast is the most accurate reporting tool on the planet, it does give you lovely little graphs so that those of you who need to communicate with bosses who only understand pretty pictures will have an easier time. However, considering you will only get truly accurate numbers by adding code to your site, I don’t expect to see people lining up to use this one.

Compete’s Site SnapShot
Another FREE one…Using my above site example, Compete gave me approximately 21,568 monthly uniques. This is also based on U.S. traffic. This may be fine for some sites but to Alexa’s credit, their stats are international and that’s what I need on many of the sites that I work with. There aren’t as many nice graphs as on Quantcast but we’ve all been wowed by visuals when they’re needed to hide numbers haven’t we?

What really gets me about Compete is this: “for small sites and/or sites with significant international traffic, Compete will report site traffic and rankings that may vary significantly from log file reports.” Enough said. My sites fall into both of these cases so I’m not going to bother with this one right now.

Overall, in researching this I would say that my mood went from the usual mostly OK to downright nasty. Why is this information so difficult to get? Why are the main alternatives only U.S. targeted? Why aren’t these programs coded so that great big skews are less likely to happen? I do not personally rely on any of this information in the work that I do currently, so this isn’t anything that should impact how I do my job. However, I do realize that many people (including my clients, bosses of my clients, other people who simply don’t have a clue) rely on this type of information in their decision-making, which DOES affect me. Crude comparisons are fine…that’s how we do keyword research and figure out how well we’re doing day to day. However, there comes a time when crude isn’t going to cut it. Obviously this doesn’t apply to my sense of humor, since crude is the only way I know to be, but when it comes to numbers, I really could use a little more refinement. If you’re able to use trends and work from there though, you’re a lucky lucky ducky.

There are also programs that you can pay for, if you’re so inclined. I’ve even consulted my sweet little EvilGreenMonkey about this, since he knows all. Hitwise is one that he mentioned and it’s also the paid program that kept popping up during my research. It’s network is international (not just U.S. targeted like Quantcast and Compete) and uses your logs to give you information on your site and your competitors’ sites. I’ve heard figures of around $2k U.S. per month is the norm, so if you’re thinking about this one, expect to pay that. Apparently many large companies use this, and you can see why…you get great data. Go through the product tour and you’ll be drooling like you’ve just had a giant veggie pie placed in front of you.

If you have any other recommendations, I’d love to hear them!

A short month after I was bestowed the auspicious status of SEO-Chick, I finally got let out of my SEOmoz cage and am thus just now authoring my first post. (Sorry Lisa, it’s been a busy month at the MozQuarters. Please don’t abuse me in yet another one of your drunken stupors.)

Anyway, one of the tasks I was busy with was putting together extensive site reviews for a couple of clients. One of the sections of the site review was an analytics overview, detailing what actions should be tracked, how to use your analytics, and specific action tracking recommendations. I’m sure a lot of you are rolling your eyes and thinking “You’ve got to be kidding me. This is all common sense stuff.” To that I say clearly you’ve never had clients, because you’d be surprised at how even the most basic stuff flies over their heads.

Not that analytics is basic, of course. All of those numbers, pie charts, and links can be pretty damn overwhelming, especially if you’re a non-math, non-business, non-anything analytics-related whatsoever major like me. Like anything worth getting the hang of, however, analytics can be conquered if you simply try, try, try. And by try, of course, I don’t mean log in once a month, poke around a bit and ooh and ahh over the purdy charts, and leave.

No sir (or ma’am), you have to get anal-retentive about your analytics. I’m talking obsessed. Like logging in every morning and checking your stats from the day before. Checking in every half hour to bask in the traffic bump from the Digg (or Sphinn?) Effect. Not noticing your non-SEO friends’ eyes glaze over as you spout on and on about referring links and search terms and reduced bounce rates. Once you get into the habit of checking regularly, it won’t matter if you studied home decorating in school–you’ll know what the hell you’re doing.

That is, of course, if you know what to look for. Any schmuck can log in every day and click around frantically like a coked-out bunny rabbit trying to hop across the freeway. What the hell are you supposed to focus on? Well, this e-commerceguide.com post is a good start. The article recommends keeping track of the following:

• referring search terms from search engines
• referring URLs
• your site’s most popular pages
• visits to the home page
• site overlay reports
• site bounce rate

To that list I’d also add anything that converts on your site–any call to action, such as an email link, any “Request a Demo/Brochure/More Information” link, add to cart links, checkout links, etc. Basically, what’s the point of having calls to action on your pages if you’re not keeping track of which ones work better than others? Also, be sure to pay attention to your analytics after you’ve made any changes on your site so that you can track whether or not those changes worked well.

The most important thing to keep in mind is that you should essentially pay attention to any data or stat that can help you improve whatever your site’s inherent goal is, whether that be to sell products, increase sign ups, encourage user participation, etc. Also, don’t wet your pants over a sudden spike in traffic in the assumption that OH MY GOD it’s a trend! You need to watch your analytics over a larger chunk of time in order to analyze long-term results.

There ya have it. Get off yer ass and start obsessin’!