Monish Datta: “It was the best WAW yet!”*
By Tim Wilson on in Social Media, Web Analytics with 5 Comments
Another month, another Web Analytics Wednesday (WAW) in Columbus. We had two sponsors — both the Web Analytics Wednesday Global Sponsors and Lightbulb Interactive, which was nifty. And, we headed back to O’Shaughnessy’s Public House because, by golly, we just knew if we gave them enough chances they could get up to batting .500 when it came to screwing up our reservations. They succeeded by having no record of our event in “the book.” We made do nonetheless.
The turnout was slightly below normal — we wound up with eleven people all told — but we tried a new format for the discussion that worked out well! Although the group was small, it was a good mix of people: web analysts from major financial institutions, web designers, SEM and web analytics types in online retail, a horse racing marketer, and a slightly-crazy-but-always-entertaining developer from an interactive agency:
The format for the formal part of the discussion was going around the table and asking everyone (who was willing) to describe the report or type of report that they felt was the most worthless or irrelevant, and then to also describe the type of report that they could not get or that was unduly difficult to get that they felt would be most useful. In other words — cheap and easy blog fodder for me! The results…
Most Worthless/Irrelevant Reports
“Hits” reports — we agreed that two cases where this wasn’t a worthless metric were: 1) error logging (e.g., missing images), and 2) server load monitoring; a late arrival proceeded to state how many hits she had to her company’s web site last year. Doh! She actually had a good recovery by proposing a third valid use: when your site is selling sponsorships and you need the biggest number you can find. Okay, so “valid” is a stretch here. Marketers. Yecchhhh!
Overlay reports — great eye candy for the vendor when they’re selling a web analytics product, but notoriously inaccurate, can’t handle links in Flash, require a lot of very careful link creation on the page that’s going to run the overlay to make sure all links are unique (which hurts SEO), and don’t work for pages that have their content updated with any regularity (when trying to look at an overlay from “two weeks ago;” Bryan provided us with an amusing medley of impersonations of business users asking questions about this sort of report
Average time on page — this prompted some debate, but the general agreement, I think, was that the problem with this report is that many, many people use it without understanding its shortcomings (which Avinash covered in detail early last year in a blog post).
Path reports — again, we had general agreement that it’s the persistent myth that a significant percentage of visitors to a site will follow the exact same path through the site that is the killer (I call that the “people are cows myth“); we walked through the various alternatives that do have value — single-level paths to/from a page, bucketing of types of pages, looking at combinations of pages visited but not worrying so much about sequence, etc.
Geographic overlays — they have their uses in some very specific cases, but they really don’t warrant being on the main page of any tool’s dashboard
My favorite from the “worst” discussion, though, was this: “Any report provided without context.” That one from the aforementioned slightly-crazy-but-always-entertaining developer.
Most Wanted or Wanted-With-Less-Work
These reports got a bit more philosophical, but it was a good list nonetheless, with some common themes:
- Several people brought up the need to marry web analytics data to other marketing channels as a biggie; they provided examples of where they had or were in the process of doing this in some fashion, but the beef was with how painful it was; this also headed down a tangential discussion of “attribution” — siloed marketing channels lead to each channel vying for as much credit as possible when they “touched” someone who converted to a sale at some point; I think we all ordered another drink in the midst of this discussion, and the lively discussion took a slightly maudlin turn. But the drinks arrived, and we recovered.
- Forward attribution combined with segmentation — this was actually related to the prior one, and I scribbled it down as soon as Scott threw it out…but now realize it went totally over my head. Maybe he’ll elaborate in a comment on this post (after he nails down the venue for next month’s WAW, of course).
- Form abandonment — this was one where it wasn’t that it’s not doable, it’s that it takes a lot of work to pull it off effectively. Well worth the effort, but would get more use if it was easier to set up.
- Onsite search — this is akin to the form abandonment one, in that it’s a really useful set of data to look at, but, all too often, is tricky to get set up in a way that makes it practical to use
- Social media integration with web analytics — this one is a result of the decentralized nature of social media, so much of what we’d want to integrate isn’t happening on sites that we “control.”
Other discussions/topics/mentions of note from my end of the table:
- Dave went from being a social media skeptic less than a year ago to being an active user and evangelist. He’s even speaking on the subject in Cleveland next month (although he hasn’t yet plugged that on his blog)
- In that same vein, Dave has also become a Gmail convert. Now…if I can just get him off of Blogspot and on to WordPress, my work here will be complete…
- I found myself talking up Techrigy’s SM2 in two separate conversations — encouraging people to sign up for a freemium account to explore social media tracking, and plugging Connie Bensen as someone to ping on Twitter with questions.
- I wound up talking about many of the people I met (in person or via social media) last fall when I moderated a panel on social media for nonprofits
- We had a few chuckles about the <political> “Leaving Us in Great Pain” video </political> that I helped produce with some friends from Austin
I almost passed my notepad around asking people to put their Twitter usernames on it…but I decided against it. Feel free to add yours as a comment here whether you were in attendance or not if you’re interested in Columbus Web Analytics Wednesday. And/or, you can join our Facebook group. I was struck by the difference 10 months makes. We talked about Twitter during the first couple of WAWs last year, and the number of users were in the distinct minority. Some people had not even heard of it. Everyone I talked to last night uses Twitter, and uses it enthusiastically. The times they are a’ changin’!
* While quotation marks would ordinarily indicate that this was a direct quote, those in the title of this post more indicate paraphrasing of Monish Datta’s take on the evening. Actually…”paraphrasing” is an overstatement. In other words, I totally made the quote up. But, Monish was smiling and laughing, so I don’t feel too bad about it. I really just needed to get his name in the title for SEO chuckles.
- Columbus Web Analytics Wednesday — Jan 2009 Edition
- Monish Datta: “I can’t believe Sasha skipped WAW for the US-Mexico World Cup Qualifier!”
- Win a WASP v1.09 for Analyst license at Web Analytics Wednesday
- Web Analytics Wednesday: A Segmentation Experiment
- Old School Online Community Leads to a Dozen Data Geeks and Drinks