analytics - Servant of Chaos

The Known Unknowns – Small Steps Big Gains with Watson Analytics

Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know.
— Donald Rumsfeld

My morning starts with a river of data. First up, there is email – a quick run by the inbox alerts me to urgent issues, questions, client inquiries and news. Every email I open, every link that I click and every article that I read is marked, tagged, tracked and collated. And then it’s over to LinkedIn’s news feed to check what is happening in the industries I care about. Having built a substantial network of well-connected and insightful connections over the years, I get a very quick sense of what is trending globally, what is vital locally and what needs to be reported or responded to. Finally, I switch to Facebook. I am blessed with opinionated and smart friends who share and increasingly, the richness and quality of news and insight available through that network is out-pacing all other channels.

But what am I doing here? I am working with Donald Rumsfeld’s mantra. I am looking to these networks to tell me the things that I don’t know that I don’t know. Essentially, this data is helping me look beyond my radar.

The same approach can be applied to marketing. In fact, for marketers to remain relevant and responsive – we need to be looking beyond our own radar.

For as long as marketers have been marketing, we have used media to reach our customers. We’ve equipped our teams with megaphones and messages and marched them to the perimeter of the business compound. To reach our customers, we buy and create media – after all, as the adage says, “fish where the fish are”. But this is the inside-out marketing model from the 20th Century – and social media has turned it inside out.

Turning to sites like Twitter or Facebook – or even LinkedIn for the B2B marketer – can feel like facing a firehose. The torrent of information coming through is astounding. Just take a look at the recent social media statistics for Australian audiences:

Over 13 million using Facebook
13.5 million using YouTube
2.5 million using Twitter

But it’s really not the big numbers that are important here. It’s what you do with them.

The known unknowns of our customer data

Just as I do each morning, marketers need to be thinking selectively about their customers. Rather than aiming to speak – or “connect with” all 2.5 million Twitter users in Australia, why not start somewhere easier? Why not start with the “known unknowns”? Why not start by figuring out WHICH of our customers are using Twitter or Facebook (ie a lot of them), and improving our understanding of those people?

Most modern CRM platforms have fields that allow you to collect the Twitter handles or Facebook profiles of your customers. Why not start by understanding how many of your existing customers have this information included in their profile? This gives you your known unknowns:

Run a report on your completeness of social media profile data in your database
Use Facebook custom audiences to match Facebook profiles to your existing email database
Do some profile matching via Twitter to do the same
Run an email campaign asking for that one additional piece of information. Provide a useful incentive. Make it worthwhile.

For most organisations, taking these small steps would take less than a day. And it paves the way for much deeper exploration.

Combining Voice of the Customer and Analytics as your over the horizon radar

Once you know who your customers are on social networks, it opens the door to a much richer experience for both you as a marketer and your customers as consumers. And what you’ll find on social networks is not the well-manicured conversation of corporate marketing – you’ll find the very direct “voice of the customer”. Generally, however, our “social listening” platforms are built around our own keywords. Our products. Brands. They are only sometimes built around understanding our customers pain points, needs and expectations. This means we are again working from the inside-out – working with the known knowns.

Thankfully, powerful natural language processing is beginning to provide the analytics horsepower we need to decipher social streams. I have explained previously about the way that IBM and AusOpen collaborate to transform customer experience at the Australian Open Tennis events – but analytics is no longer the domain of big business. With platforms like IBM Watson now available at an affordable rate (starting at around $50 per user), you don’t need to be “Tennis Australia” nor a data scientist to understand what’s going on with your market. You just need to understand your business.

Take a look at this video to see how you can use Watson and Twitter data to analyse retail sales. Look at the way that the language in the real time reports is structured around the way that marketers work. Rules are setup and then data populates accordingly. But most interestingly, because Watson works with natural language – it works with the language of the marketer as well as working with the language of the customer.

For marketers, this means that Watson does the hard work of identifying the most interesting facts contained within your data sets, letting you focus on making the right decisions about what happens next. For example (at 2:13), “sales by state” is flagged. Watson chooses the best representation of the data (in this case, a map) but also provides a “ribbon of data” that can be used to interrogate and analyse at a deeper level. Typing in a search related to what you need to know (eg “tweets by hashtag”) turns that data into a report that lets you see immediately what happened to your sales data and why.

Suddenly that river of Twitter data becomes understandable. A connection can be made between your business results and the social media data coming from a particular channel in a particular location.

And in case you need to tell the story of your digital marketing and your analytics to your executive team – or to your customers – with a few clicks, the visualised data can be compiled into ready made infographics. Now you not only have a custom radar to understand your customer – you can link your customer and business value together. Will this make you a better marketer? It will certainly make you more relevant to your customer – and that is a win all around.

https://www.youtube.com/watch?v=7yEDUCc4qUg

Data is Eating Marketing: Digital, Social and Mobile in 2015

Facts & Figures, Mobility, Social Media

Data. It’s out there. And there is plenty of it. We create data with every status update, photo shared or website viewed. Each search we make is being monitored, sorted, indexed and analysed. Every purchase we make is being correlated, cross-matched and fed into supply chain systems. And every phone call we make is being logged, kept, passed on for “security purposes”.

There are so many kinds of data that it is hard to keep up with it all. There is the data that we know about – the digital items we intentionally create. There are digital items that are published – like books, websites and so on. There is email which creates its own little fiefdom of data.

There is also the data about data – metadata – which describes the data that we create. Take, for example, a simple Tweet. It is restricted to 140 characters. That is the “data” part. But the metadata attached to EACH and EVERY tweet includes information like:

Your location at the time of tweeting (ie latitude and longitude)
The device you used to send the tweet (eg phone, PC etc)
The time of your tweet
The unique ID of the tweet.

But wait, there’s more. From the Twitter API, you can also find out a whole lot more, including:

Link details contained within the tweet
Hashtags used
Mini-profiles of anyone that you mention in your tweet
Direct link information to any photos shared in your tweet

There will also be information related to:

You
Your bio / profile
Your avatar, banner and Twitter home page
Your location
Your last tweet.

There is more. But the point really is not about Twitter. It is the fact that a seemingly innocuous act is generating far more data than you might assume. The same metadata rules apply to other social networks. It could be Facebook. Or LinkedIn. It applies to every website you visit, each transaction you make. Every cake you bake. Every night you stay (you see where I am going, right?)

For marketers, this data abundance is brilliant, but also a distraction. We could, quite possibly, spend all our time looking at data and not talking to customers. Would this be a bad thing? I’d like to think so.

The question we must ask ourselves is “who is eating whom?”.

In the meantime, for those who must have the latest stats – We Are Social, Singapore’s massive compendium is just what you need. Binge away.

Digital, Social & Mobile in 2015 from We Are Social Singapore

From Big Data Science to Big Data Action

Big Data, Digital Strategy, Disruptive Technology, Mining the Gold, Social Business, Social Media

From the dawn of civilisation through to the year 2003, Google calculates that humans have produced 5 exabytes of data. That’s a lot of stone tablets. But with the explosion of mobile devices, 3G and 4G networks and social networks, we now produce 5 exabytes of data every two days. That means that every photo you upload to Flickr or Facebook, every video you share with friends on YouTube or Vimeo and every one of the billions of tweets broadcast on Twitter is contributing to the avalanche of data.

But add to this the fact that each of these items comes with contextual data. At the same time that you update your profile or publish a photo, you may also be sharing your geolocation, your likes and preferences, your upstream and downstream behaviours, and your attitude to topics (based on sentiment). You may also be sharing your trust network of on and offline friends.

And this is just the tip of the big data iceberg.

The rise of big data is a blessing and a curse for CMOs

While analytics have been available to businesses for decades, but it has largely been the domain of business analysts and researchers. The rise of big data now places analytics firmly in the marketers court. Earlier in the year, a CMO Council and SAS report indicated that only 26% of marketers leverage customer data and analytics to improve decisions, targeting and personalisation.

The blessing of big data is that it is readily available to most organisations in the form of structured business data and the publicly available unstructured data coming from social networks. The curse is that in-house skills and experience with big data is scarce – with a number of marketers now looking to bolster their teams with big data scientists and data analysts.

Marketers don’t need data they need action

It’s not data scientists that marketers need, however. Already we are seeing software vendors emerging who are able to tap structured and unstructured data sources to produce business-ready dashboards. Mapped to best practice business processes, these dashboards and analytic tools promise to release marketers from the fear-inducing data tsunami that looms on the horizon.

Platform players such as Anametrix, for example, transform the science of data into actionable business knowledge for key business processes. This means you can spend less time and resources understanding the data and its various relationships, and focus instead on making decisions that impact the top and bottom lines of your business.

A great example of what can be achieved is the BrandWatch US Electoral Compass. Drawing on Twitter data and press discussion generated since July 2012, the compass matches structured information (location, policies and dates) with unstructured information (tweets, sentiment etc) to reveal the topics that are important to American voters. Now, this is not data from focus groups – it’s stated intention as revealed via status updates, commentary and attitude.

And as business analytics packages get better at mapping business flows, these reporting systems will become ever more granular. They promise to revolutionise the way that businesses engage with their customers – and that will bring another set of challenges for CMOs. The question is – are you ready for this new form of customer engagement?