Web Analytics Wednesday is this Wednesday, the 27th of August, at Bar Wellington. Here’s the link to let Patrick know you’re coming! http://www.webanalyticsdemystified.com/wednesday/ Alright, so I shrill for WAW. It’s true. But it’s one of the big events in Toronto for us Web Analysts. Our community isn’t terribly large, and, moreover, it’s one of the few occasions that I get to heckle Mike Sukmanowski. And who doesn’t like to heckle Mike? It’s also a great opportunity for eMarketing folken to come out, along with WA-related developers, IA’s, and data modelers. In fact, a huge invite for the data modelers to come on out and meet with the web analysts. The call to action is: come out dammit! Detail in the[…]
Complexity is hard. Happily/Sadly, web analytics projects are fraught with complexity. My initial reaction is often to get a sort of ‘deer in headlights’ look, and then, I cope with complexity. Here are a few methods. Backcasting Backcasting asks, initially, “what’s the end state?”, “What’s the goal?”, or, most fundamentally, “why”. If I understand the end goal, it becomes much easier to start thinking about the prerequisites. And the step back. And step back again. Often, you run into a situation where you discover that you should have started work 4 years ago, and that, indeed, is part of the fun. Chunking / Atomization Breaking a problem out into its component parts helps. Immensely. Drawing out those boxes, and then[…]
The reaction to the Garbage Can has been fun. My good friend pointed out that I never really explained the entire theory of the Garbage Can, and he’s right. I didn’t. I’ve only briefly described the topline thought process behind it. I talked briefly, in the original post about the Garbage Can, about how process can really alleviate some of the symptoms. If key participants agree on the problem we’re trying to solve, and agrees on the criterion on the solutions, then an actual effectual outcome is more likely. Of course, even the best process in the world can save you if there just isn’t enough energy in the can to make it pop. “But how do you go about[…]
A big thank you to Andrea Hadley of eMetrics/NetSetGo for hosting the WAARM, the Web Analytics Association Regional Meeting, and for having me out. I want there to be a strong incentive for our Chicago readers to attend WAARM Chicago. 🙂 So, I’ll share a few high level outcomes. My fellow panelists were John Hossack of VKI Studios and Jason Carmel of ZAAZ Seattle. The big takeaways are to write a job description that focuses on finding very smart people – the skills development can be loaded in systematically. There’s a belief that the recession is causing an increase in demand for measurement and web analytics, and it appears to be broadbased, and, if you’re going to freelance and go[…]
I’m afraid this is yet another one of those philosophical blog entries. J Consider a web analytics vendor selection process. What a mess that can be! You have some people who genuinely believe that something is the best, and then you have some people whose salary premium merely depends on their ability to (fakely) pull reports out of a hyper-specialized piece of software, you have people who like a particular vendor because of who they know, and then you have everybody else. Worst case scenario: kick-backs or related ‘perks’ – which, just for the record, is one of the absolutely worse bases for making a decision. It’s hard to be objective when you’re making a decision though – but let’s[…]
Much of December was spent struggling with the List Processing language. I believed I could use genetic algorithms, similar to what I had done for advanced pattern recognition, and apply that technique to an emerging field called ‘website morphing’, which is coming up in a forthcoming paper. A ‘morphing’ website is one that actively changes itself in response to a user input. I’m not talking about changing which pages are displayed based on a series of clicked links as is the case with Web 1.0, but rather, how a user is interacting with the site. It’s a direct use of web analytics to enhance the user experience in real time (if we take the notion of morphing to its extreme).[…]
Marketing Sherpa released an interesting chart of the week that I figured is worth sharing because it raises an important problem – how should you signal to a company that you don’t want what they’re sending you? So often, one signs up for an email with an expectation that you’re going to get what you’re signing up for, and I’ll admit – when I invite Air Canada into my mailbox for airfair specials, I expect to only get airfair specials. Only. I don’t want to be cross-promoted with their latest Aeroplan offering, and on and on. Or about any sort of “courir pour la poutine” fun run and bar-b-q that they’re having in support of the Heart and Stroke Association.[…]
Inflexion points in a career are really quite interesting things. There’s a certain something about pushing yourself outside the base comfort zone – like the first time you do a SEM campaign, or code in jQuery, or do your first applied linear regression to awful, awful data. (But you do it). And then there are points where you push yourself well beyond that comfort zone. There’s a bit of a cycle to it, that I’ll compare to trying to swim. First, you get out of the hottub and you walk around the pool and assess the situation. You see a shallow end, and a deep end. You know it would be far more safer to jump into the shallow end[…]
Hamel got things started with “Conversion is King” blog post: So I was talking with my co-worker Patrick Glinski about my notion of conversions being the only metric that matters and he brought up an interesting problem with it. If you have a business, and your only concern is sales (conversions) then over time you will lose because you are ignoring the other parts of your business. To continue with my Rolex example, if Rolex only focused on the sale, the conversion, and not on the aspiration, eventually everyone who could buy one, would have then Rolex wouldn’t have any new potential customers. To which, Sukmanowski replied: Take one of mine and Novo’s favourite metrics for example – recency. The[…]
Remember this equation? Y = mx + b The dependent variable, Y, responds to a change in the slope (m) times an independent variable (x), plus the intercept point or constant (b). Typically, we strip out the letters and replace them with words to make it nice and readable. (Number of Conversions) = (conversionRate)(PageViews) + constant What’s the constant? The best way I can describe it simply is “left over unexplained stuff”. Ideally, you’d have an equation that intercepts the Y-axis at 0 (the origin), but, that’s fairly rare, because there are other factors involved, and it’s really rare, at least in marketing science, to get an equation that completely explains everything, perfectly. (Though, it’s not impossible, as the above[…]