Do you think ‘big data’ is
over-hyped? Are you tired of hearing about its potential and want to know how
it affects the bottom-line? Will Hayes (CEO,
Lucidworks), Phil Kim (Cap One), Michael Morrison (CEO, Datawatch), Adam Towvim
(CEO, Trust Layers) share
their perspectives on how companies are deriving value from big data today,
moderated by Andrew Burton (CEO, Logentries).
As the panel
traversed topics ranging from hiring practices in big data to data governance
and protection, there was one major reoccurring theme: big data is not the same
as the big answer. As Adam Towvin eloquently put it, “sometimes it’s not about
finding the needle, it’s about building the right the right haystack.”
Much of the reason that big data as a field feels like hype is
because of the ways it’s being described as a way to find answers and solutions
for a huge range of previously unsolved challenges – and those answers have not
been forthcoming since “data science” became an industry buzzword.
But as our panellists today have outlined, that’s partially
because of the way big data is being misinterpreted and partially because of
limitations in data usage and lack of expertise in the field.
Will Hayes kicked things off reminding us, "don't get
infatuated with technology, get infatuated with results,” a sentiment echoed by
the rest of the panel. Phil Kim agreed that big data is in stage where people
don’t understand what it is and how it can be used – and although there are
great tools for extracting value from data, people aren’t always sure on how to
use them.
The biggest benefit
of big data at the moment, said Towvin, is that data helps companies innovate
faster and improve business in ways it was impossible to do before.
Andrew Burton turned the panel to the question of how “offensive
or defensive” data should be in bringing data science practices to other parts
of the business.
While the group agreed that an offensive approach was important
for educating others about the value and limitations of big data, as well as formulating
questions and helping customer-facing teams build better products, there was
some discussion of the tension that other departments (particularly legal)
might feel when large pools of data across the company are thrown together and
leveraged in this way.
When it came to data governance and protection, the panel
acknowledged the importance of protecting data but Towvin in particular felt
the aggressive ways in which data was encrypted and hidden away undermined
those data true potential. While protection was important, there is a sense we
are selling customer short by using that data to make better decisions in areas
ranging from health to consumer products to personal finances. Kim highlighted
Capital One’s practice of putting risk management and marketing teams in the
same room so that they can jointly solve problems of how to protect and use
data effectively at the same time.
The audience was curious about how to prevent people from
confusing correlation and causation – a challenge the panellists admitted is
almost impossible to solve. Their solution? Hire the right people, experts with
deep insights to help them avoid these biases.
Of course this naturally led to a discussion of how to build the
right team and make the right hires for the field. Kim explained the challenges
he faced explaining to hiring departments what a data scientist actually was,
but all of the panellists agreed that the most important trait for a data
scientist was a deep intellectual curiosity, as well as scientific yet creative
mind that can absorb the quantitative challenges of big data while still
looking for interesting patterns.
Fortunately for the students in the room, Michael Morrison spoke
to the need of these big data companies and teams to hire “the next generation.”
He pointed out that many of these companies need to do as much to market
themselves to talent as candidates need to do to market themselves to
companies. Morrison highlighted an example from Datawatch – they opened an
office in Boston as they knew it was a more appealing location for the type of
people they wanted to attract.
Looking forward, the panel spoke passionately about the benefits
that real time analysis of streaming data will provide to the industry.
Thank you to Will, Phil, Michael, Adam and Andrew for a great
overview of this fascinating and fast-moving field.
