Facebook Doesn't Get Small Data Right (And You're Worried About Big Data?)

Posted by

**There was a lot of grumpy folks complaining about [Facebook](https://www.facebook.com/mitchjoel “Facebook”) (again) lately.**
I can understand why some folks might be upset. Someone in their family passed away or a pet got lost this past year, and suddenly this image is showing up in Facebook’s recently released The Year In Review as a way to celebrate that moment with your social graph. The Year In Review is a little app on Facebook that allows users to share with their friends a collection of photos – month by month – to summarize the year that they had in 2014. The thing is, that Facebook took the initiative to pre-populate the app with photos that they thought would best illustrate our year. So… yeah… that led to all kinds of [awkward and bad photos being a part of the shareable slideshow](http://mashable.com/2014/12/27/facebook-year-in-review/). While it is easy enough to customize, remove and add whichever photos might be better, we can be a grumpy bunch around the holiday season. So, seeing photos that make us sad or angry probably wasn’t the best way to get users to use it and talk about it.
**Pushing beyond the Facebook user issue.**
Facebook is so big. It doesn’t matter what it does. It will be met with all kinds of reactions (some good, some bad… some neutral). And, while many marketing and communications professionals are quick to analyze how Facebook is reacting (and apologizing) for offending some users, there’s a bigger question that marketers *really* need to be asking: *if Facebook can’t even use simple tags and basic semantic tools to figure out if a picture was good or bad for a user, how do we really think that our brands are doing when it comes to handling data?*
**The big data myth explored.**
Of course, big data is a real thing. There is no doubt that there are engines of marketing technology that are using previously unavailable data sets, and combining them at speeds previously unimaginable to create new and interesting marketing opportunities. With that, are you telling me that Facebook’s technology wasn’t strong enough to figure out what really were the best photos to share? Forget that, are you telling me that your brand would have been able to do any better than what Facebook pulled together? Big data is real. What’s more real is that there are very few brands that are currently able to fully harness, leverage and act on the isolated data sets that they have in regards to consumer usage, email addresses, search engine marketing results and customization of website based on usage. With that, there are probably even fewer brands who are able to take the data sets mentioned in that last sentence and meld them together to get a clearer perspective of what the actual digital consumer looks and acts like. In 2014, every brand talked about the need to better leverage big data, in a world where almost every single one of those brands are still fundamentally living with a bunch of data ghettos (h/t [Bryan Pearson](https://twitter.com/pearson4loyalty “Bryan Pearson”)), this Facebook experience validated something very important: for most brands, big data is still – sadly – out of reach.
**A better day for data.**
Don’t take this commentary as a shot against big data and what it can (and will) be. Marketing automation – and the output of it – is one of the core disrupters to the marketing industry. Understanding the value in automating many of the data-based initiatives that we are currently running is going to change lots of the ways that we score *”success.”* Once more automation is in place, the data sets are going to become increasingly more interesting and relevant. Once that level of hygiene and efficacy is achieved, just imagine how powerful the consumer data becomes. At that point, a brand’s ability to make that data sing with new and untapped repositories of data from other parts of the business will create unprecedented opportunities. It will, without question, also be the real moment in time when marketing will not be perceived as a cost center at the c-level, but rather a core function of business operations. If this sort of thing seems far off in the distance, don’t be fooled. It’s available today – and it’s not all that expensive – for those that have the culture within the organization to make this all a part of the core. With that, isn’t Facebook all about the data? So, is this one experience gone wrong, or is it just another illuminating example that we have a long road ahead of us?
**The questions that we need to ask.**
It feels like a contradiction. On one hand, big data is here and accessible to all. On the other hand, big data is here, but we still have a long and complex road to head down for brands to truly generate any kind of real and true economic value from it. If anything, Facebook just showed the world that data – even when it’s not all that “big” – is still funky. Filters, tags and other tools are still liable to fail. So, we’re back to a simple – and always used – ethos: *let’s get great at the basics*. Mistakes happen. Small mistakes can lead to a lot of painful attention. Big mistakes can be fatal (and hard to undo). The excitement of big data is not to be diminished. Let’s just make sure that before we go ahead and let the technology do what it can do, that we are in the best place possible with the data that we’re currently using.
**Are we all cool with that?**