Just as I’d started getting used to the ‘Big Data’ square on my buzzword bingo card, the world of number crunching is moving on.
Through much of 2012 and the first few months of this year, big data was touted as the cure for almost all the connected world’s problems - from solving civilisation’s worst social ills to fixing publishing’s audience engagement challenge. Sitting at the tail end of 2013, the promised data miracle seems a little less miraculous.
The most memorable tweet from last month’s Media Marketing Strategies conference illustrates the developing shift in attitudes wonderfully well - “Big Data is Bullshit”.
The first time I saw big data described as male bovine excrement was in coverage of a speech by the painfully cool Harper Reed, CTO for Barrack Obama’s data-driven 2012 re-election campaign (pictured right).
The ‘Big’ there is purely marketing… This is about you buying big expensive servers and whatnot.
Not much room there to see big data as a “process that has the potential to transform everything”. That was the view of US computer scientist Josh Kleinberg reported in a New York Times article explaining how big data got so big, just 12 month’s earlier.
Smoothing the digital transition
So where does the apparent collapse of the big data revolution leave the publishing fraternity’s search for a Grand Unified Theory of Everything to smooth the digital transition?
Luckily, as a publishing pro closing up your strategy for 2014 and beyond, there are a few points of clarity emerging, most advising a tight focus on the ‘data’ half of the proposition, few caring much for the ‘big’ part.
The story behind the “Big Data is Bullshit” tweet from the Media Marketing Strategies conference was that publishers generally don’t have anything near what could be called big data. They have small data; multiple sets of small data, according to Adestra’ Parry Malm, whose #MeMS13 presentation inspired the tweet.
Reading past the ‘Bullshit’ soundbites, the big data backlash isn’t anti-data. It’s anti-hype, a welcome reminder that there are no silver bullets.
Data quality matters more than quantity, and proper analytics matter more than almost anything else. The question now is ‘Do you want big data or good data?’ and there’s clearly only one way to answer that question.
Despite being quoted as saying “never give an editor a spreadsheet”, I know that data is an integral part of strategic and tactical decision-making for any publisher; actually, for any business working in this connected world.
I wrote a blog a couple of weeks back about similar data-angst in the pharmaceutical market. Understanding how to encourage patients to take their drugs probably matters more than figuring out how to convert container app downloads into subscriptions. But the concerns are the same.
The Key Perspectives on Customer Databases survey from healthcare CRM vendor Cegedim asked senior pharmaceutical company executives what makes data valuable. Accuracy was the lead factor in determining whether data was useful, with completeness and coverage – big data stuff – rated much less important.
The survey also showed phone verification as the most trusted data validation methodology. Phone verification? Big data doesn’t use the phone. Big data measures itself in terabytes and petabytes, not caller minutes.
Find the signals
Whether its pharmaceuticals or publishing, it’s the data that matters not the scope. And, as Parry Malm says, the key is finding signals in the data.
At this year’s Defrag conference, a future-tech event billed as the conference for big thinkers, big data took a bit of a kicking with speakers lining up to challenge the general understanding of the concept: it doesn't have to be big, it's not about data, it's about insight.
The strapline on a Defrag report from The Upstart nails it:
Big data is worthless until a scalable and accurate process for generating insights is improved and refined.
What we’re really seeing is an end to technology-fuelled ‘my data’s bigger than your data’ bravado and a re-focusing of attention back on the value that data brings. As Graham Oakes wrote on Econsultancy in August, the ‘Big data backlash has its origins in the mismatch of expectations and delivery. The ‘bullshit’ commentary is a welcome reality check on the idea that the science of number-crunching was somehow going to solve all our problems.
None of this means big data is going away. Forbes ran a report in September, asking ‘What big data backlash?’ where the author cited a survey of Fortune 1000 C-Suite executives that reported 68 percent of respondents were planning data infrastructure investments valued over $1 million, up from just over one third in last year’s survey.
But even here, where the trend continues, the burning issue was analytics; 70 percent of survey respondents highlighted the need to “accelerate analytical processes” and develop “more sophisticated analytics”.
So as you wrap up your plans for next year and beyond, thinking about how to use data to build a single, multiplatform view of your customers or automate your marketing processes, bear in mind that getting the analysis right and acting on it is way more important than how big your data is.
And if your servers and software don’t give you the quality of data you need to make informed decisions, don’t worry, just get back on the phone.
Peter Houston runs Flipping Pages Media, providing training and consultancy to magazine publishers working to leverage print success on to emerging digital platforms.
Image via Flickr courtesy of Just A Prairie Boy used under a Creative Commons licence.