Mymotech

Comscore recently published a press release about 1st and 3rd party cookie deletion rates that they measured in the month of December.

Here are the important details.

• Duration: December 2006
• Sample Size: 400,000
• Sample Site: Portal with content and an Ad serving 3rd party cookie
• Average number of 1st party cookies per computer measured: 2.5 (100 visitors becomes 250 visitors)

So, the gist of this press release is to say that 1st party cookies are an imperfect method for tracking unique visitors over time, and that using them will inflate visitor counts by as much as 150%.

OK, then what should a web analytics person do?

1. Don’t panic - Everybody already knows that 1st party cookies are an imperfect form of tracking unique visitors over time. Cookie deletion happens, people surf at work then at home, and a whole multitude of other scenarios that pretty much make 1st party cookies a crap shoot.
2. Look and see what metrics you are using that depend on unique visitors. Here are two samples:
• Visitor Conversion - This could be a biggie, but guess what; you are not overstating conversion you are understating it. Grab last months Visitor total and divide by 2.5 now use that number to calculate top level conversion. (Orders/Comscore Visitors) Boo Yaa right? For all you web analysts that get paid by conversion boosts, it is time to figure out where the pool goes. And practically speaking this doesn’t really change all that many things except to normalize and baseline at a different level. Obviously I am keeping it simple. This is a blog post after all.
• Campaign Response - Ok this one could tighten a little bit, especially when the sales cycle for your product is longer than a day or two. So what can you do? Start by understanding the difference between regular old visitor conversion and a campaign visitor conversion. Then get about learning how to understand your entire conversion process from offline to online and vice versa. Do some surveys to understand visitors who purchased better, and quit demanding a 4 to 1 ROI on Adwords. In actuality these numbers should drive you to your customers for answers and that is a good good thing.
3. Next start to redefine your metrics realistically. You have to fight against the current of web analytics vendor marketing speak, but you can choose and use metrics that do not leave you over-extended and washed up when Comscore points out the existing flaw in your underlying data. This is actually one of the harder steps, because it takes a special analyst not to jump on the next greatest metric band wagon, and to realize that every metric that gets used in the rest of the business world is not necessarily ready to be used in web analytics.
4. Don’t start writing the business case to move to a complete session based cookie tracking system yet. Comscore only surveyed one site and that with their panel of users who are provided spyware tools that may or may not recommend the removal of 1st party cookies. See Eric Peterson’s Blog for more on that.

So, these are interesting times. My advice: beware anyone who gets up in arms on either side of this issue about cookie deletion. Most likely they are trying to sell something. Get to know your visitors and customers. Cold stats are good for breakfast but you need more than that to sustain you.

The old put your name on it and you will take better care of it thing is not working. But in the interests of letting everyone know I am still alive, here is a quick list on what I have been up to lately.

1. I read a lot. You can have copies of my OPML file of web analytics and SEO blogs if you want.
2. I do a show with my brother. The Clickherder roundup. We will be posting episode 3 tomorrow, so surf now.
3. I shovel snow a lot. (Because Easter is when all the snow comes apparently)
4. I work. Things are going well, and I am involved in some exciting ecommerce projects (of course, I can’t talk about that).
5. I am building a website. I have teamed up with a friend to develop one of them niche type of sites, with an aim to test my mettle in the SEO ring. I am starting out in a relatively easy niche, and once I get a #1 on Google I will be sure to let you know of it.

Well, this is a blog about web analytics, so I guess a web analytics type post is in order.

Since Google Analytics came out last year, there has always been a question:

Isn’t Google Analytics just as good as other web analytics tools that cost more?

And while there are many different answers to that question, I ran across something today which inspired a closer analysis.

When looking at keyword reports in Google Analytics you can view the search phrases

Or You can view the Search Engines

But you can’t view both at the same time. In other words, You can’t see a report that displays all the search engines and their underlying search phrases simultaneously and vice versa. In addition the export function only grabs what is on the screen, so there is no correlation of those two data points in any exported report.

Briefly, I want to describe why it might be useful to see search engines by search phrases or vice versa. Let’s say that I am trying to optimize my site for the phrases, “wowza”, “yowza”, and “gyoza”. Wouldn’t it be useful to be able to track the number of referrals I got from Google and Yahoo over time without having to expand and collapse multiple boxes within the report? But it doesn’t appear that that is the case with Google Analytics.

The lack of this feature makes Google Analytics inferior to the likes of WebTrends, Omniture, Coremetrics, and WebSideStory. Although there is still no beating the price.
BTW, I would be happy to be wrong, so if you know some way to accomplish this in Google Analytics then please let me know, and I will obviously post an update.

I have been letting this blog languish over the past few months for all the typical reasons. One result of this was that almost all of my pages had turned into supplemental results at Google. Or so I thought.

Shoemoney recently had a post about how he experienced these SEO woes, and how he has corrected them. Since I have now assumed my true identity on this blog, I think I owe it to myself to take a look at the current SEO state.

Sadly, it is not good. What to do? Well, I just so happen to be the proud owner of Aaron Wall’s SEOBook. So guess what, I am going to start applying some of those principles directly to this blog, and turn this mother out.

First goal, ensure spider crawlability. I have a sitemap down at Google webmaster tools, but it was out of date and old, so a quick update of that was in order. Next, get a robots.txt going. Then make sure the URLs are nice and friendly. A quick update to the htaccess file and shaboom.

Bring on the Google traffic.

In all my years of so-called blogging, I have never revealed my name on this site. I guess I didn’t want to take the risk of having something to live down (potentially). But as I have thought about it, and I have become more interested and involved in the field of web analytics I have decided that my voice makes more sense on this blog if my name is attached to it. Also, because I am vain and would like to rank No. 1 in Google for my name.

So, without further ado, my name is Michael Helbling and I have added an about page on the sidebar for all the world to see. Boo yaaa.

BTW, this doesn’t mean I will post more.

I wanted to quickly jot down some quick notes about Gatineau, Microsoft’s new web analytics tool, that is in pre-alpa.

I think we all know that Microsoft is doing this to follow in Google’s footsteps to provide a tool that will encourage advertisers to spend more money on their respective ad delivery systems. Unfortunately for Microsoft, since Google thought of it first they bought a better company than what was available to Microsoft. Urchin was purchased by Google and DeepMetrix was purchased by Microsoft. In my opinion Urchin was a stronger performer in the analytics space than DeepMetrix. My experience with both tools pre-acquisition is not terribly strong so this is more my perception than a well researched fact.
Also, one has to question whether Microsoft is doing a little too much “me too” when it comes to business innovation in the search space. This is something they used to great affect when launching IE, but I wonder if this method of taking over the world has become slightly antiquated. We will see, I am sure.

Microsoft has brought on Ian Thomas, one of the founders of Web Abacus, to drive the development of the Microsoft version of this tool. In my mind they will need to produce a better look and feel for the tool to make it worth adopting.

On the flip side Microsoft has a long history of entering the market slowly and crappily but eventually dominating the space. They will potentially be able to integrate some very rich demographic data into the analytics reports. I think that this is unlikely in release one. And even though they have lost some market share in the search space, they will slowly get better and better and their AdCenter product will slowly start gaining traction, especially since Google is making their AdWords product more and more complex and harder to enter for a small business.

The great news in all of this is that we will see even more attention paid to web analysis and metrics based decisioning in online business. That will be a big victory for everyone.

Obviously, I am hoping to get my hands on an early version of this tool to put it through its paces.

Once upon a time I needed to register a domain. And in my innocence I decided that Godaddy was for wusses and I needed something a little more geeky to register my domains on. Enter Registerfly. They seemed pretty cool and stuff and they also had some cool discounts on domain registrations. Awesome.

Over the next year I registered some 25 domains with Registerfly, because that is how I roll. (most of them $.60 .info domains meant for a small amount of blackhat SEO burnin’ and churnin’. Strictly for educational purposes mind you.)

Of course, in the ensuing time period I’ve had to recommend a registrar to many friends, and I have always recommended Registerfly even though the internet is littered with bad reviews of their service. link link link

(Continued)

It could probably be said that it is a new year of blogging for half the year, if previous years are any indication. But I will stick with it, and keep this blog going on the off chance that I will eventually have something to say.

One of the things I have notices other bloggers doing is putting a pretty picture up with each post, and so maybe I will try to do a little more of that this year.

Also, I think I will probably need to start a series or two because one of the things that has kept me from blogging is that putting together a post takes so long and I have so little time.

And finally, here is one web analytics new years resolution: get some exposure to usability testing and surveys this year. I think that these two tools to increase web site performance are uber powerful when combined with good analytics. Web analytics will deliver the what and user interaction will help to shape the why of web site behavior. Awesome!

One of the things that frequently gets clouded in all the hype about web analytics tools and packages is what does a business need anyway? At the E-metrics summit in Santa Barbara Xavier Cassinova spoke about his experiences both as a web analytics vendor and now as the president of a new startup business. He talked about a disconnect between what web analytics vendors were trying to sell and what business owners wanted to buy.

So the question is, what do businesses need online? The answer is money.
Businesses need more money. And they need information on how to get more money.

How can web analytics do that? I am glad you asked. The primary way is to measure conversion. What you have to do is start peeling the conversion banana to get a better understanding of factors that lead to or take away from conversion.
Another thing that matters a lot is retention. If you can give customers that convert a reason to convert again you are pretty much guaranteeing success on the web.

Ok this post didn’t come out as organized as I intended. I may take another shot at it. In the meantime, lets focus our analytics on what really matters. Making Money.

In my work I do a lot of implementing of specific tools, mostly WebTrends. I install, configure, train and then walk away and leave the organization to figure out where to go next.
Problems abound with this scenario. Usually there are a couple of things missing in almost all of these organizations where I do this: someone to understand the data and someone to promote the data.
Understanding the data takes analysis. At Emetrics, back in April, one of the most engaging presenters, Avinash Kaushik from Intuit, detailed very clearly what is needed in this area. You can read all about it on his new blog Occam’s Razor. It comes down to a smart person with business acumen. Someone like me. :)