All Superweeks are great, but this one was Awesome!

by | Feb 13, 2018

At the begin­ning of the month, Tim Wilson and I had an oppor­tu­nity to once again visit Hungary and attend the Super­week Analyt­ics Confer­ence. Once again, we were delighted with the overall confer­ence setup, the people, the content, and the chance to once again record our podcast in front of a live audi­ence. Here are some high­lights for me.

The Global Data Protec­tion Regu­la­tions that will start being enforced in May 2018 were a big topic this year. There is much to consider with these regu­la­tions but what was very inter­est­ing to me was discov­er­ing that I was approach­ing the whole concept with a differ­ent philo­soph­i­cal POV than many Euro­peans. While I like the idea of GDPR, in the context of creat­ing an even playing field between the indi­vid­ual and the compa­nies that would seek to benefit from their data, I think of it as a market­place concept. If a company provides me a good or service that provides value to me, then I am fine with their collec­tion of data. The Euro­pean view is more about the protec­tion of a funda­men­tal human right without a market­place compo­nent. It gave me a lot to consider just in terms of the philo­soph­i­cal under­pin­nings of a set of regu­la­tions that will undoubt­edly make their way to the USA in some form in the years to come.

Machine Learn­ing and AI
So much is happen­ing in our space to embrace machine learn­ing and AI. One thing that is defi­nitely still true is that there is still a lot more talk than action, but that being said many talks at Super­week show­cased some really cool learn­ings and accom­plish­ments. It is no longer out of anyone’s reach to apply machine learn­ing to the prob­lems of their busi­ness, and while there is nothing remotely stan­dard, it is increas­ingly acces­si­ble in ways that I am really excited about.

Also, an insight that Matt Gershoff shared in his talk, allowed me to recon­tex­tu­al­ize the quote “All models are broken, but some are useful.” I think the empha­sis so often is placed on the fact that all are broken, instead of on the fact that some are useful. I am excited to help our clients glean value from the useful models this next year.

The other major take­away for me was from Tim Wilson’s talk on statis­ti­cal think­ing. One of the central tenets of statis­ti­cal think­ing is embrac­ing uncer­tainty. We are fight­ing an uphill battle in digital because vendors often promise us certainty, and frankly some­times so do us prac­ti­tion­ers. But there are larger and larger costs to reduc­ing uncer­tainty, and great analy­sis is about hitting that sweet spot between a great deci­sion and the incre­men­tal cost to reduce uncer­tainty further.

Market­ing Attri­bu­tion
I have long been some­what of a curmud­geon when it comes to market­ing attri­bu­tion in digital. Typi­cally, it worked out as long on cost, and short on any persis­tent value gain. What I saw at Super­week were numer­ous talks that were taking a much more prag­matic approach, and one that I can get excited about. So in 2018, I am allow­ing a shift in my think­ing as I learn more and consider what the next evolu­tion­ary step in attri­bu­tion can provide.

The Digital Analyt­ics Power Hour
This trip to Hungary also was the first time all three hosts of the podcast were able to meet in person. It was so great getting to know Moe, and just cemented how lucky we are to have her as a co-host on the show. We also recorded a podcast there. You can listen to it here. Also, the fine folks at Super­week have created a Youtube video that you can watch as well.

I may be the only person who gets excited walking away from a confer­ence with more ques­tions than answers, but maybe I have just learned to embrace the uncer­tainty.

If you haven’t quite learned to embrace the uncer­tainty, don’t be afraid! Feel free to reach out to us for a little guid­ance. We’re here to help.