I just got two emails about Web Analytics Demystified in the last two days. The first was from a fan, saying:

“Just wanted to let you know the “Web Analytics Demystified” rocks. I have been toting this book around in my backpack for the last 6 months. Up to a few weeks ago, I used it mainly for reference when getting ready to teach […] analytics classes during my day job […]. Just recently I started reading it from cover to cover (before bed I might add, and no it does not put me to sleep). This book is packed with good info and it is great to find it all in one place. I tend to wave the book around during class … ( I love books), and everyone who has picked it up feels the same way. Same same goes for the “Web Site Measurement Hacks” +100 . (I am a hard core O’Reilly book fan).”

Wow! Nice compliment, thanks!

The second came by way of a friend:

“Sitting in a convoy heading to a client meeting this morning, I noticed that one of my media buyers in the back seat appeared to be reading her book upside down. When I asked her way, sarcastically of course, she replied back in a very self satisfied tone that she wasn’t reading it upside-down, that she was “reading it upside-down and backwards.” Then she handed me the book: “Web Analytics Demystified” by Eric PetersonI opened it and the title page seemed normal – but from that point on the book required you to turn the book upside down, Go all the way to the back and then read in reverse, right hand page first and then the left, before flipping the page to the right. “Upside-down AND Backwards” by western standards at least.”

My suspicion is that this book was one of the orginal books that were printed and shipped by CafePress. I got a number of requests for replacement for “upside-down and backwards” books (a la “Silly Sally” for those of you with little kids) and gladly shipped new books.

If you have one of these “rare” copies of Web Analytics Demystified, if you’d prefer a “right-side up and forwards” copy, contact me directly and I’ll get you a replacement.

Anyway, thanks for the feedback!

Ok, not really. But he does have a pretty good analysis of what I was trying to get at in Web Analytics Demystified when I tried to explain conversion rate. Which reminds me of a funny story …

Did any of you see Jared Spool at last year’s Emetrics ranting about “conversion rate is dead!” I followed him around after his (excellent) presentation asking him how he could be saying that! Didn’t he realize that everyone was just now converging on conversion rate as a meaningful metric?! Didn’t he see that by telling people to not measure their conversion rates that he was just inviting trouble?!?! I was about to become apoplectic (which at the Biltmore, in the Santa Barbara sunshine, is pretty hard to become) knowing that people really listened to this guy and I had to ask him did he know what he was doing?

Yeah, he did.

His point was basically, “Don’t tell me about conversion, tell me about REVENUE!” and he offered “Revenue per Visitor” or “Revenue per Visit” as even better predictors of success than conversion rate.

Ok, that’s hard to argue with.

Maybe instead of The Unabridged History of the Pencil my next book title will be Conversion Rate is Dead!

I mean, they say controversy sells, right?

When I posted last week about Avinash Kaushik from Intuit being podcasted I made one mistake. Intuit is not an Omniture customer. According to Avinash, only the TurboTax web site uses SiteCatalyst; ClickTracks is the official web analytics tool at Inuit used at the vast majority of their properties.

My apologies to Avinash and John Marshall at ClickTracks. No harm was intended. As I told John after Avinash’s presentation at Emetrics last year, the kind of passion and zeal that Avinash has for both web metrics (or “web insights!”) and the applications he loves is rare indeed. Lucky ClickTracks!

Based on a post he made to the Yahoo! group I just realized that my friend Clint is blogging now at Instant Cognition. Welcome to the blogosphere, Clint!

In one of his first post, Clint admits to being conflicted over whether “dashboards” are good or evil. Clint, you should ask Jim Sterne whether dashboards are good or evil when we’re all in Santa Barbara next month. Of course, we both know the answer he’ll give:

“It depends.”

Yep.

If you’re a hardcore data geek like some people I know, dashboards take complex correlations and over-simplify them, almost to the point of being meaningless, just so they’re accessible to a wider audience. Clint wisely points out:

“As web analysts or report designers, our job is to clearly and quickly reduce the overwhelming volume and complexity of data at our disposal to recognizable information that our customers can quickly understand and use to make impactful business decisions.”

Yep.

Clint goes on to say:

“Simplifying that blood sweat and tears of designing an effective display of information into the word ‘dashboard’ both trivializes that work and makes it easier to sell its value to others - so here I sit conflicted.”

Perhaps herein lies the essence of his confusion! “Dashboards” and key performance indicators are not supposed to be the end-product of our labor! These reports, regardless of how they’re presented, are designed to keep the reader connected with the business and drive inquiry; I always say that KPIs that don’t drive some type of action are not KPIs at all!

So fret not, my friend.

All you have to do to resolve your inner conflict is get your corporate masters to understand A) what their dashboards are telling them, B) how much change is acceptable and C) what actions they should take based on any dashboard element exceeding its specific change threshold.

Easy, huh?

Yeah, ok, not so easy. But important. Very important. I propose that no organization will ever be successful with the use of dashboards unless they’re actually prepared to take action based on the information those dashboards represent. Don’t measure your conversion rate unless you aleady have a clear plan, including set expectations, for how you’re going to improve your conversion rate slowly and incrementally.

This is why the tachometer is perhaps an apt visualization for many of our web sites: As Clint points out, when I think “dashboard” I picture the dash of my favorite sports car. When the tachometer in said sports car redlines, I immediately do something–I take action. Failure to do so will cause miserable consequences (likely driving my wife to force me to sell said favorite sports car, rather than have it towed to the shop again …)

Put another way, when people respond to this massive simplification of your blood and sweat and tears with intelligent questions, your conflicted feelings will likely vanish because, well, you’ll have more hard work to do.

Again, welcome to the blogosphere.

Internet Marketing Voodoo has a nice podcast with Avinash Kaushik from Intuit posted on Monday, March 6th. I first met Avinash at last year’s Emetrics event in Santa Barbara and he had me rolling in the aisle, his presentation was awesome!

Some highlights include:

• “Web analytics is dead” and we’re entering into an era of “web insight” — trying to understand how we can get beyond clickstream and log files, etc. I love his positioning, “web analytics is dead” indeed!
• People translate “web analytics” into “clickstream analysis” … but at Intuit the company has been marrying clickstream with outcome data and customer experience data (I wrote about this strategy at JupiterResearch when I described the “new usability framework“)
• Avinash is very much a believer in measuring the customer voice and combining that data with clickstream
• He can take online transactions and tie them back to customer survey data via clickstream, essentially creating visitor segments of “satisfied customers”
• Avinash cops to being a relatively small analytics group inside a very large company (only six people on his analytics team)
• When asked about vendors, Avinash points out that it’s not the package you use, it’s whether you have in house expertise available to take advantage of whatever package you select. He comments that if he only had $100,000, he’d spend $10,000 on software and $90,000 on someone to analyze the data.
• Despite Intuit being an Omniture customer, Avinash again stumps for Clicktracks (somewhere John Marshall is smiling again!)

(Oops! I’m out of time … I need to head out to Web Analytics Wednesday!)

The transcript of the podcast is available at transcripts.internetmarketingvoodoo.com (I’m not linking to it since it is very slow to load and appears to almost crash my browser every time I go to it.) All in all I love this idea of podcasting about web analytics. Maybe more people will begin podcasting, or even better, maybe Sterne will arrange to podcast from Emetrics.

(Props to Eric B. at WebTrends for the head’s up about the podcast.)