Gold mine

Facebook finally lets us tap into its data goldmine

This week Facebook announced a new insight product called topic data, which Facebook says will “show marketers what audiences are saying on Facebook about events, brands, subjects and activities.”.

Up until now, we’ve had a really limited view on what we can get out of Facebook in terms of data, and it’s been nearly impossible to get the full picture of consumer conversation as so much has been inaccessible. Listening tools like Sysomos and Brandwatch do currently aggregate Facebook data – but only public data from Facebook pages and profiles with open security settings. Which meant Facebook was a bit of a black hole for consumer insight.

But that’s all changing…

As of now, Facebook topic data allows marketers and analysts to find out what people are saying, even from private status updates. Facebook has partnered with DataSift to aggregate and anonymise user data, delivering topic data through its PYLON for Facebook Topic Data product.

What we need to know

  • The data is anonymised – you wont be able to see who said what
  • To that effect topic data will only be of value for consumer insight and research – not monitoring and responding, or direct targeting of ads
  • But we will be able to see demographics (gender, age range, self-declared location) to help us segment and interpret the data intelligently
  • At least 100 different users have to match a query for it to get delivered
  • It doesn’t include Messenger – private messaging app data still remains a blind spot
  • It’s currently unclear just how much we will be able to play with the data – according to Facebook “results delivered to marketers are analyses and interpretations of the information, not actual topic data.”
  • It’s currently only available in the US and UK (data limited to these markets too)

What we can do

Now we can finally get our hands on the goldmine that is private Facebook data, but what can we do with it?

Here are three examples off the top of my head of how Facebook topic data can be used to boost social intelligence.

  1. An automotive brand can use it to see which competitor models potential customers are weighing up against their own when they’re talking about buying a new car
  2. A hardware retailer can see demographics on the people talking about putting up shelves to help with content planning
  3. A beauty brand can see what colour eyeshadow is getting people talking this season

It’s still early days, with vendors like Synthesio exploring the possibilities of integrating the DataSift API. I personally can’t wait to get my hands on it – such a huge opportunity to build an even richer picture of today’s consumer.

Image: Andrew Kuznetsov @ Flickr


The moment we remembered what disruption really means

Let it be known that I’m 100% NOT against what we refer to as disruptive innovation or digital disruption. This is my bag, I’m in it, I’m a player, woo! Innovation is what excites me about my job and gets me out of bed in the morning.

But before we get carried away we need to have a word with ourselves and remember the meaning of disruption. According to, disruption is defined as:

“Disturbance or problems which interrupt an event, activity, or process.”

Uber, the taxi booking tech startup, has faced a real backlash over the last few weeks. Since December’s $40 billion valuation the company has faced a barrage of hostility due to its aggressive rollout strategy and “Wild West tactics“. Uber is facing legal scrutiny across the globe – where government regulation is leading to bans in some countries and cities – not to mention questions of passenger safety, privacy and various ethical issues.

The surge pricing model, which adjusts prices based on supply and demand, got Uber into hot water when prices increased in response to people trying to flee the Sydney hostage siege. But it is this controversial model that is one element of Uber’s ‘disruptive’ competitive edge, and something the company is trying to patent.

Tech industry folks (the very early adopters who helped catapult Uber into everyday life) are now deleting the app in a boycott that started after allegations of sexism, misogyny and dubious privacy practices. This prompted a trend of tech bloggers posting about deleting the app, causing The Guardian to ask if Uber is “the worst company in Silicon Valley”.

Boycotting Uber is nothing new – Paul Carr wrote “Silicon Valley’s Cult of Disruption” 2 years ago – but it seems to have taken off in the past month as former advocates flee and the press jumps on every negative story. John Naughton wrote an interesting opinion piece recently suggesting that Uber is actually a poor example of innovation; rather its disintermediation using networking technology to be the middle-man, nothing groundbreaking.

Groundbreaking or not one of Uber’s biggest issues is that, in disrupting an established industry, it has failed to demonstrate an acceptable level of social responsibility. Disruptive innovation means the world is being shaken up. Great! But it also means that the world is being shaken up. While change driven by technology is often magical and exciting it is also painful. This has been true forever.

We should embrace the magic but remember the true meaning of disruption. Perhaps more companies should be prepared and open, and maybe a little more responsible.


Big ideas about big data (sorry I said big data #sorrynotsorry)

Last week I strolled along to Quantcast’s Big Data Summit‘s first European outing to hear from the media and adtech industry on the latest trends in big data. Here are the themes and ideas that resonated.

The appeal of programmatic buying comes with a warning

Programmatic digital advertising is the fastest growing form of digital advertising, currently contributing 25% of total digital ad spend. The exciting opportunities for improved targeting, automation and above all else, increased efficiency and ROI come with a downside. Like a beautiful big brown bear, increased marketing automation looks funky, but if not handled correctly can turn nasty. 12% consumers have complained directly to advertisers, because of sloppy, irrelevant and annoying retargeting campaigns. More thought needs to go into the planning of automated campaigns so that brand reputation isn’t damaged and money isn’t wasted chasing the wrong consumers.

There is growing tension between creativity and data

James Dunford from Cotswold Outdoor said that his relationship with agencies has changed from creative to technical, and the technical relationships are the ones he’s concentrating on. But we can’t assume this means a relationship with technology alone, people are just as – if not more – important.

There is a current debate in the industry about data and insight stifling creativity. I don’t believe this to be true at all (unless the insight is just bad), but I do see the value in striking the right balance.

Attribution needs to focus on outcome over action

I hear a lot about the challenges of attribution and that as an industry we’re way behind where we should be. Attribution needs to mature to demonstrate the long-term effects of brand exposure and engagement.

We need to make sure we’re not forcing ourselves into models that don’t consider the bigger picture. Chasing short-term ROI can be short-sighted, but is often what we’re asked to do; marketing managers are often only in the job for a few years and need short-term results. We should challenge ourselves to think long-term and broader than digital. Like my friends at Attributely who recently proposed a longer-term view of attribution, using a ‘break-even point’.

More isn’t just more – more is different

Getting sick of the term ‘big data’ (I don’t like hypey terminology), I’ve started calling it ‘more data’ instead. But I’m wrong! Kenneth Cukier from The Economist pointed out that with data, more isn’t just more – more is different. Advances in data science and big data are challenging what we know and how we know it. We no longer need to rely on samples – we can just get it all. This opens the door to not just more but different possibilities.

Big data is the new alchemy

Data has become a resource, a new raw material. Not only is it a valuable commodity but it also allows us to create, test and invent. Now that we can collect, store and process data like never before, we can play with it to discover things we’d never expect. Splicing micro data from every farm in America with meteorological data meant a billion dollars for crop insurance and seed manufacture companies. Wal-Mart uses customer data to make sure they have the right products in stock ahead of changes in the weather. Did you know that sales of strawberry Pop Tarts increase 7 fold before a hurricane?