Enter a search term such as “mobile analytics” or browse our content using the filters above.
That’s not only a poor Scrabble score but we also couldn’t find any results matching
Check your spelling or try broadening your search.
Sorry about this, there is a problem with our search at the moment.
Please try again later.
Nate Silver is one of very few people who is both public figure and statistician.
He started in baseball during its analytical revolution, but became famous in the U.S. for his highly accurate predictions of how Barak Obama would win the presidency in 2008, and accurately calling the outcome for all 50 states in 2012.
Silver spoke at SXSW in a session entitled "Is Intuitive Marketing Dead?"
The talk covered a lot of ground, touching on a number of the themes in Silver's newish book The Signal and the Noise. For marketers, the most interesting (or at least relevant) questions were about data, its limitations and possibilities...and how we create a culture that includes data without losing ourselves to it.
On the session's eponymous question, Silver doesn't suggest that we should seek to find data - based solutions to every problem. On the contrary, he acknowledges that intuition is probably the best way to deal with what he calls "medium sized" problems.
An example he cites is the ability of experienced baseball scouts to outperform models of statistical prediction. But, and it's a big one...those scouts got better at predicting outcomes as they incorporated new thinking about baseball statistics into their "intuition." In other words, our guts get smarter the more data we have, and the more experiments we run.
You can't stop bias, you can only hope to contain it. Bias is like a gravity source...it can bend data until it fits. Every company, team and individual is carrying their own bias whenever they encounter a question.
Silver's advice is to recognize and embrace that bias, because it's most dangerous when ignored.
Models versus failing fast
Digital marketing is blessed and cursed by having enormous data sets.
That's a good thing in Silver's mind. He suggests that when you've got huge data sources and the ability to quickly see outcomes, to bypass modelling in favor of rapid experimentation - if you can see what happens in the real world quickly and accurately, why overthink it?
Big data and corporate culture
Paraphrasing an unnamed philosopher, Silver describes bureaucracy as the opposite of imagination. That indictment often applies to large corporations and part of the problem is separation; teams and individuals operating so independently that the lose track of their peers and larger business goals.
Silver contrasts that phenomenon with the flatter structure of stat ups, where everyone does everything, and can't forget that they're part of something.
Centers of excellence or a distributed model?
When asked about approaches to analytics, Silver is vehemently in favor of a distributed model over the center of excellence. His argument goes back to the nature of bureaucracy...that any advantages you might enjoy by collecting analytical and statistical expertise into a single, powerful team is mitigated by the distance that team has from the disparate parts of the company that rely on them. To his mind, proximity trumps scale in this case.
Brand versus volume
You might expect a data guru to dismiss the ethereal nature of brand, but that's far from the case. Instead, Silver talks about a near future where data and technology are ubiquitous among marketers, where targeting and optimization is closer to science than art (a long way off, but not forever).
As we move in that direction, he suggests, brand gets more important, not less. And he makes a good point...once we've successfully optimized, targeted, tested and personalized everything, what's left but brand?
Having just published a book on the failures of predictive science, it's not surprising that Silver is wary of prognostication.
However, he did agree with the premise that there's huge opportunity in the personal data set...that if privacy issues are properly managed, many successful businesses will spring from the emerging practice of analyzing individual data that's captured first by the individual and then aggregated.
Examples of this type of data include location tracking, diet diary apps, what people search for and where, body metrics and media consumption patterns.