The phrase 'big data' has been on the tip of everyone’s tongue
lately. It’s the kind of phrase that’s gone from the boardroom to TV
commercials seen by consumers. It’s also seen as a magical elixir for
marketing executives – an expensive initiative that looks and sounds
good in presentations, but may not have actually solved anything.
The truth is that too many big data initiatives today are put in
place without any specific outcome in mind. They can amount to $50m
investments that allow executives to pat themselves on the back, even
though they failed to move the bottom line.
That’s all preventable though. Marketers just need to build their
data initiative with some clear ideas in mind. There are really two
pieces to successful big data strategy – asking the right questions, and
then putting the answers into action.
The first step with any big data initiative is figuring out the
purpose of the initiative. It’s the big 'Why?' As stated above, big data
efforts can cost upwards of $50m. Clearly, these are not designed to
deliver quick fixes. If the chief executive or chief marketing officer
is asking a big data question at the moment they need it, such as why
they are losing customers, then it’s already too late (and that question
is likely going to be a very expensive one).
Big data is about using numbers to change, improve or add – it’s for
accomplishing something. Data crunchers don’t necessarily understand
marketing strategy, so they need direction. Before even building
infrastructure, begin with the fundamental questions about outcomes. Is
the purpose to turn the brand into a market leader? Is it to change the
brand identity? Is to build a new brand? A big data initiative can only
give the answers it’s designed to provide.
Once the questions are in place, brands must assess if those
questions can even be answered by the data on hand, or if more is
needed. The data they already have represents what a brand knows. Brands
must then identify what’s missing – and what they don’t know – before
building their data team and initiative.
Another important question is 'when?' When a big data operation is
built around a set of goals, those goals must be accomplished in a set
time period. Yet most initiatives I see don’t have a set time, putting
those brands in danger of simply spinning their wheels. Data scientists
will continue to probe ad infinitum without a direction, out of sheer
intellectual curiosity. Giving them a direction is crucial.
The final question to ask is if the data is actually usable. The
central question to any data initiative is figuring out the difference
between what a brand can do and what they should do. In short, which
techniques, answers and insights are actionable and will help accomplish
the stated goal?
Even once big data initiatives produce insights and answers, it’s
important to question those answers. Academic analytics are not the same
as marketing analytics, and data scientists hired straight from
graduate school often only have experience in research-based work. They
may end up implementing innovative analytics that aren’t practical in a
commercial business, solely because that’s their only experience to
date.
It’s important to be critical before building big data
infrastructure. Failure to do so could result in a big investment on
something that doesn’t work, and could jeopardize the whole initiative.
While throwing around the phrase 'big data' certainly feels good,
actually knowing the answers to valuable business questions feels that
much better.
No comments:
Post a Comment