Big data has no role in digital marketing
Big data has a time and place, but not when it comes to digital marketing strategy, explains Michael Laps in the guest post.
“As accurate, then, as big data can be while connecting millions of data points to generate correlations, big data is often compromised whenever humans act like, well, humans.”
That quote is from Martin Lindstrom, author of ‘Small Data: The Tiny Clues That Uncover Huge Trends’. And I think he’s bang on the money.
According to the Advanced Performance Institute, humans generate the same amount of data every few minutes as they did between the beginning of time and the year 2000 combined.
The majority of the organisations that can collect enough data to be considered “big data” are huge. Banks, insurance, technology or large retail are some that come to mind. Most initiatives that these companies undertake are generally slow moving, taking years to roll out. So, my question is this, is it the companies/company structures that make big data unusable, or the data itself being too large to segment and action?
What another shocking article. Pick up ya game mumbrella.
I think the issue is what constitutes big data. IMO it is the stormwater drain of what we do online, it isn’t structured and gives little real info into anything unless it can be attributed or connected to a person…in which case it isn’t big data.
In basic terms, big data simply says we did stuff online, we posted a pic, we watched a video, we read a story – all things we can assume are now givens.
What about the software supplies who host the banks, insurance companies’ data?
Saleforce must have a ton of data! No wonder Microsoft bought Linkedin.
#datawars
suppliers*
That’s true, agencies that use the corporations data. That leads to my question, it’s the companies, not the data, that makes it hard to work with.
There is an article floating around the web about how the Brexit and Trump campaigns effectively used big data to run effective campaigns. To cut a long story short they employed some of the smartest people on the planet – people working at the edge of whats possible. They created profiles of 220 million people and they implemented a campaign genuinely aimed at delivering exactly the right message to the individual. Then they wrote a big cheque. Trouble is most “big data” starts with the wrong team, not enough money, inertia at the client and an advertising agency without the skills. The issue is not that big data can’t be used it’s just that you need the right people.
Agreed, you couldn’t even get your meme right – the name of the ad is right there in the filename!
Hi Simon,
I agree with you. Talent and resources are critical to using data effectively – whether it’s be big data or agile data.
And yes, the Trump campaigns used big data to their advantage, but they had over a year in which to run that campaign – as opposed to short-burst campaigns based on changes in behavioural trends. They spent a long time building voter personas, understanding voter needs and motivations, as well as identifying areas in which they could capitalise. It wasn’t an overnight thing.
My premise is that for companies who are trying to roll out new marketing campaigns – and specifically digital strategies – that effectively communicate with rapidly changing consumer needs, they need something more flexible and digestible than big data.
Sure, if you had Apple’s budget and Google’s team of data analysts, you might be able to turn big data projects around more rapidly. But unfortunately that’s not achievable for 99% of companies.
Hi Dave,
That’s a great question! I personally believe that it still comes down to the size of the data set and the nature of the projects themselves.
Unfortunately having an organisational structure more conducive to rolling out big data projects won’t necessarily reduce the time it takes to actually complete the analysis of the data. Because you’re analysing an enormous number of individual data points and trying to find patterns and correlations, it’s going to take time. There’s no way around that – unless you have the budget to have an enormous team of data analysts (like Google, for example). But that’s not achievable for 99.99% of companies. And that’s the issue.
The fact that a stronger organisational structure would reduce roll-out times makes it a change worth making, but if it’s going to take 6 months to generate the insights and strategy first, then my original argument still stands – it’s too slow. Companies should invest in running more frequent, more agile projects that are able to respond more rapidly to consumer needs.