Wednesday, March 8, 2017

The Ruthless Efficiency of Algorithms is Advancing Digital Frontiers

We recently caught up with Alistair Croll, Visiting Executive at Harvard Business School as well as our Marketing Analytics & Data Science Conference keynote speaker, to discuss the state of marketing analytics and data science, and where it’s going in the future.

Today, Croll helps to accelerate startups, and works with some of the world’s biggest companies on business model innovation. As an entrepreneur, he co-founded Coradiant; the Year One Labs accelerator; and a many other startups. Not to mention, he’s a sought-after speaker, and has launched and chaired some of the world’s leading conferences on emerging technology, including Startupfest, Strata, Cloud Connect, and Pandemon.io. Croll is also the author of four books on technology and entrepreneurship, including the best-selling Lean Analytics, which has been translated into eight languages.

What is the state of the data science and analytics industry in 2017?

Croll: There is a realization that data itself doesn't lead to answers. This is really maturity: It's asking the right question that's hard. Big data is replacing business intelligence, but most of it is still being used to run reports and batch processes—rather than to find advantage or insight.

At the same time, feeding the corpus of data into learning algorithms holds promise. Those with the authority to do so are pointing machine learning at their data seta to find correlations, then testing those for causal relationships they can exploit.

What have been the biggest changes data science and analytics since you started your career?

Croll: I'm not an analyst by trade. But the biggest change is clear: once, we first defined the schema, then collected data. Now, we collect the data, then define the schema.

In other words, "Collect first, ask questions later." This is a huge difference, but it has sort of snuck up on us. It means we can iterate more, answering questions and adjusting our lines of inquiry.

Have the influx of social media and mobile made your job easier or harder?

Croll: More data sets mean more potential insights, but also more spurious correlations. So it's a two-edged sword.

How is data science and analytics transforming every industry right now?

Croll: The simple, and somewhat terrifying, truth is that AI gets unreasonably powerful, very quickly. Whether driving a car, or playing a video game, or diagnosing a disease, or optimizing the design of an aircraft part, algorithms are better than humans. They don't get tired; they make fewer mistakes; they don't take breaks.

And what do we feed such algorithms? Data. There is no industry that will not be changed by the ruthless efficiency of algorithms advancing its digital frontiers.

Why is data science considered the “sexiest job of the 21st century?”

Croll: Data science is the intersection of statistics, critical thinking, and engineering. It requires a sense of narrative, and the ability to build something. It's that element of engineering that distinguishes it from simple analytics, because it builds things that become products, or processes. Rather than running a report, it improves the report's results.

If big data is oil, data science is the refinery that makes it usable.

What is the biggest challenge in data science and analytics today?

Croll: We are still, sadly, trying to replace opinions with facts. My good friend Randy Smerik argues that there's no such thing as big data: An airline that knows you're running late fails to update your hotel; false positives about in credit card management.

His point is that while we have tremendous amounts of data, we seldom apply them to significantly improve the business or the customer experience because doing so means making fundamental changes to the organization, job descriptions, customer policies, and so on.

Where do you see data science and analytics moving in the next 5 years?

Croll: Democratization, with the help of smart agents. Pundits have been saying that for a long time, but in the last couple of years tools like Cortana, Google Now, Siri, and Alexa—as well as various chat interfaces like Slack, Sophos, and Skype—are going mainstream.

I also think that insurers will put significant pressure on companies to implement better analytics and algorithms because it will be too risky to do otherwise. If the organization can know everything about itself all the time, it will be expected to do so. "We didn't know this was happening" will no longer be an excuse. And consequently, algorithms that can parse all of that data and reduce risk will be mandatory.

Hear more from Alistair during his keynote session, “Don’t’ Get Duped by Data” at the Marketing Analytics & Data Science Conference April 3-5, 2017 in San Francisco, CA.

Data science and marketing analytics are transforming every industry. There is a reason why it is being called the sexiest job of the 21st century. Calling all professionals that want to harness analytics and data science! Do you realize how critical you are to the future of your organization? Learn more here: https://goo.gl/CbYosj


Use our exclusive Blog discount code MADS17BL for $100 off the current rate. Buy your tickets here: https://goo.gl/CbYosj

No comments: