GrowthBeat 2015, day two.

Oh, my feet hurt.  It's been a long couple of days, but I did enjoy the conference.  It's interesting to me because it's mostly about straight up marketing, which has definitely evolved from the Mad Men days, but it's not my world.  I'm a techie who got here by accident via the Java community.  Attending something like this is very educational for me, it's literally learning a new language at times.  They were talking about metrics I've never used, behaviors I've never seen.  A bunch of it literally doesn't apply to me - my company sells things, but I don't.  I don't make decisions about ad placements or large scale campaigns.  There's no shopping cart to abandon in my world.  But it's interesting to see and think about how those things can apply to the things I do. 

My favorite speaker of the day was Sid Patil, he's the senior data scientist at twitter.  He's also that guy they wrote books and movies about who counted cards playing blackjack with a group of students from MIT.  We got a little bit of those stories, but his insights about what data is useful and what isn't blew me away.  One of the biggest things marketers track on the web and mobile is where the last click came from, and they tend to target those places for ad spends, but he made a compelling case for that being shortsighted and in some cases totally wrong. 

The example he gave was about a travel company (he didn't name names, but I'm assuming something like Expedia, Orbitz, Travelocity) that noticed a huge spike in purchases around 7pm.  So they went and did a huge spend on marketing appropriate to that behavior.  But it failed.  It actually tanked sales.  After deeper research they realized that customers were researching everything during the work day, and then purchasing after they got home and talked with their family. 

He also argues for picking a path supported by data when you have choices to make, even if occasionally a single transaction fails.  It will fail occasionally, but the overall strategy will win if your research is solid.  He used a 35k loss at a BlackJack table as an example, and a football story that I couldn't retell to save my life (football is even more foreign to me than old-school marketing).  But his point was solid.  Good stuff to think about.