I'm just back from the Superweek digital analytics conference in Hungary, held high up in the mountains in a secluded spot with stunning views. Think The Shining with a couple of hundred web analysts and digital people spending the week together sharing ideas, techniques, and war stories.
Going to an event like this you always leave with a whole heap of ideas and thoughts, many of which get lost forever after you've been squeezed into an aeroplane for twentysomething hours on the way home, but I wanted to get a bit of a digest together about some stuff that went on.
So here, in no particular order, are some slices of what happened. I apologise for the photo quality.
Avinash Killed The Funnel
You can't get a global conference of analytics people without someone announcing that something was dead, but nobody was expecting this. Avinash doesn't like funnels, and as other presenters and opened their laptops to update their slides to reflect the new funnel-less world that we all now inhabit, there's actually some pretty sound reasoning behind the headline.
The funnel model is pretty old (1898, if I recall correctly, was when the AIDA funnel was first suggested). The lovely Jim Sterne held his hands up and said that in his first Emetrics white paper in 2001 he waxed big about the funnel -- and whichever way you look at it, consumers are very different in 2016 to how they were in 2001 (and, presumably, different to how they were in 1898 too.)
Purchase and conversion is non-linear, cross-device, and we behave less less as though we're moving through a funnel. As analysts, we look at movement through the funnel but as customers of goods or services online, we think of ourselves as wayward, unpredictable, individual little butterflies.
It's not quite Dead until there's a new model to replace it, but it's a thought-provoking idea.
Avinash's First Step And The Last Mile
Superweek had two Avinash sessions (not including an Ask Me Anything) and he structured the first around The First Step, and the second around The Last Mile.
I'm a fan of these as the two pivot points for successful analysis; deployment is important, but strategy is more important and the Last Mile stuff around logical and clear visualisation and display of data is an area that all analysis can benefit from.
The dataLayer Is The Only Way
Statements like "if you're not using a dataLayer, go home now" seemed to crop up earlier this year than they do now.
During chats around and outside the main conference stream, analysts are less tolerant of on-page dependencies with Tag Managers; last year it seemed acceptable but this year the dataLayer is king.
Attribution Is Mature
There were a couple of talks on attribution, and one panel (which I had to miss thanks to aforementioned squeezage onto an aeroplane.) No matter whether you think it's a waste of time or a vital component of successful analysis, the techniques and procedures are devastatingly mature now and what has historically been airy-fairy is now starting to become increasingly bedded-in.
Privacy Is A Timebomb
Australia is, thankfully, quite insulated from what's happening in Europe around data privacy and collection. It almost reminded me of On The Beach, where Frankston is the only part of the world that's shielded from worldwide nuclear devastation.
Led by concerns in Germany and the wider EU, there's a massive gathering of a storm around data collection, which is likely to stretch very far beyond IP Address Anonymisation. And while we're shielded in this country. what happens over there tends to happen over here and I'm expecting some serious changes to privacy to impact what we do over the next five years.
Online-To-Offline Is Here
Two Googlers, Jos Meikerhof and Oliver Borm, did an unforgettably entertaining demo of using the Measurement Protocol to fire sales and refund data from an offline store to Universal Analytics.
At scale, this si still a very big challenge for retailers but the technology demonstrated itself as clear, simple and reliable (even on spotty conference WiFi).
The Golden Punchcard
Each year, Superweek holds a Golden Punchcard Prize which gives a prize to the audience-voted best example of an analysis technique, technical miracle, or other cool stuff.
This year the winning technique was using predictive analytics to improve remarketing performance, but my personal favourite was a mechanism for inferring an organic search keyword and passing missing SEO data to Google Analytics as a custom dimension.
Other notables were a library to cut down on GTM implementations, and using Chromebits for $30 beacons to track user movements and check-ins via Google Analytics.
I Didn't Want To Go Home
I love my home very much, but this year I didn't want to leave Superweek in the slightest. It was like being in Disneyworld for a week, except without the creepy overtones and rabid capitalism.
The only thing rabid about Superweek was the appetite for analytics and while their upcoming conference in Jamaica is a bit hard to get to from Australia (35 hours travel time, including transits) Hungary 2017 is already in my calendar.