Web Analytics Wednesday Toronto - November 10

Shaking the next Web Analytics Wednesday, at Bar Wellington - second floor.


You're invited to bring 25 copies of a single sheet of paper that contains:

• Three bullet points of analysis, preferably with an actionable recommendation or finding
  

• Data that supports your analysis

• A reference to the data source

• Your name, company, contact info, website, blog, twitter handle, and so on
  

It's not a dashboard.

It's web ANALYTICS wednesday.

Why?

You'll leave with something in your hand and feel smarter for experience.

We rarely get a chance to share our craft with other practitioners.

Why shouldn't practitioners have a chance to put up?


The agenda

We'll start at 6:30pm and we'll get up and move around with our sheets. You're mostly analysts, so you'll start off with the people you already know and then you'll radiate outwards.

Distribute your sheet and talk to it within each group.

Be brief, be brilliant, be moving on.

We'll go on like this for 90 minutes and the night will carry on like it normally does from 8pm.


What kind of data?

Do not share data from your company unless it's already publicly available. I cannot stress this enough. Be professional in your selection of data source.

If you have your own website, company, or have a major forthcoming paper - and it's your data - you can share from it. Quote it.

If you do not have access to your own set of web analytics, there are alternative sources. You are, after all, living in the world of big data.

Open sources include:

• Google Trends

• Wolfram Alpha

• TweetAlyzer

• Quantcast

• Compete

• Facebook Insights

You are not short on data sources and interesting stories to tell.

If it's of interest to you, you'll be able to talk with interest, and others will be interested in it too.


Do I have to?

Please do.

The November edition of WAW is traditionally long on analysts. Our space can accommodate 25 people nicely, and another 20 on the other side.

I'm planning for 25 people to show up with 25 sheets. If there's more, the merrier.

This is a really rare opportunity offered to people who want to try it out.


I'm not an analyst, do I have to?

Please do.

If you're not a practitioner, the exercise of putting something together will give you an idea of what it's like.

We always welcome new people into the community.


Where do I sign up?

http://www.webanalyticsdemystified.com/wednesday/list.asp?event_city=Toronto


Questions? Reply on back to me.

More data than we know what to do with

It was only towards the tail end of the second year of university when anybody tells you about the pearson tables. It's a glorious thing, all alone in there, hidden away in PASW. You run it for a number of variables, and it gives you a beautiful matrix showing the strength and direction of relationships among them.

It can be disastrously misleading. Violence can dull sensitivity. Still, it can be used to rapidly validate mental models and rule others out.

I cope when I'm confronted with a large dataset.

I identify what is it that I'm trying to figure out. Then lay out all the independent variables that I think might relate into explaining that variation in something dependent. I keep a few notes. I'm about to go in deep and if I'm not careful, I risk being fooled by randomness.

(The longer you randomly look for a connection, you more likely you are to find something randomly significant. The risk is real.)

Then I run pearsons to quickly rule out the ones that don't look promising. Then validate the two that do make sense. Chances are good that I'll come out with a much better understanding behind the drivers of something.

It's a coping mechanism when confronted with too much data. And it takes time to do.

There are other ways of coping, too. The average, the median, the mode. Standard deviation, skew and kurtosis. For the visual: the histogram and the normal curve.

The Key Performance Indicator is always supposed to be coping mechanism.

We're confronted with more data than we know what to do with. It's so much even with a very high degree of statistical literacy. It's akin to knowing how to read and being confronted with a Library of Congress. Even reading the back covers is time consuming. Possibly unrewarding. Why would you want to?

I think most cope by simply tuning out.

And so do I.

I cope by selectively tuning out.

Instead of wanting to know everything, which is implausible, I simply try to know something useful. And something useful typically relates to a competitive advantage. I actively practice selective ignorance.

I need something better than that. Because somehow I don't think the banner IGNORANCE IS STRENGTH is a winner. Do we like amnesia?

Perhaps the construction of new mental models mandates temporary amnesia or selective intelligence? Sometimes, to move forward with a fresh perspective, to innovate in a big way, you need to temporarily forget much of what you know?

It's more data than we know what to do with. But someday it won't be enough.

Web Analytics Wednesday Toronto Recap

Fun was had by all - there were quite a few people from out of town who came out - and a very diverse group appeared to get along well. The venue of The Counter went over well, and will probably be the preferred place for crowds in the 35 person range.

There were a couple of inside-track discussions, and I think we're in for quite a shakeup in the next couple months.

One theme that continued to come up are challenges in overlap and general effectiveness of our programs. Complaints about unresponsive institutions continue to roil, yet some wins were celebrated. We toasted Mike Sukmanowski's victory in what appears to be a clean install of Omniture over at the Globe and Mail.

I'm increasingly concerned with the unification of CRM methodologies and thinking with Branding methodologies. Emma Warrillow is too. There was a lot of discussion around that, and something we'll continue going over in the lead up to December.

Discussion of the UBC Course and Certification was also dominant at the tables I happened to join. It's really great to see so many new grads of the program out at the event.

Thanks to all who came out and made it a productive and fun evening. An announcement about the November venue is forthcoming.

Web Analytics Wednesday Toronto: Oct 6.

Tonight's edition of Web Analytics Wednesday Toronto is taking place at The Counter. A new venue that we're going to try out.

I've invited web analysts to get together with IA's, Data Miners, Dev's, and, for the first time in a long while, Creatives. Only good can come out of that sort of interaction.

The openly visible agenda is all about forming weak ties. People getting together over beer and wine, talking to other people with a totally different perspective, in small groups, in an honest exchange, is valuable. It doesn't happen very often because of the tendency to keep to ones own.

I look forward to meeting everybody.