Thoughts about media in Canada

Data’s Human Ecosystem


Everything is data. Human behavior, attitudes, spoken words, sentiments, the weather, traffic jams, past experiences with a brand, web behavior, media usage…

–          All this is captured by different tools & databases, coded in different ways.

–          Those who manage to link the databases together and analyze them will find trends and correlations, discover deeply hidden insights, build predictive models… ultimately giving them a significant edge over their competition.

Big Data is a symptomatic expression. It tells you that business leaders have a growing appetite for more insights about their consumers, about their industry, that they recognize the need for bigger and better business intelligence, they acknowledge it’s an enormous and complicated task and that they’re willing to pay the price to get it.

As a digital media shop, we understand the need of providing data services to clients and when done right, it can be a real cash cow both on the short and long term. How do you make it happen? In my (not so) humble opinion, it’s a matter of having the right mix of people/skills and knowing how to use them properly.

<strong>The Data’s Human Ecosystem</strong>

Two years ago I started at The-Agency-Formerly-Known-As-Media-Contacts. Not long after, I was mandated to launch the Canadian Data department. Not an easy task for a Jack of all trades media guy with a somewhat above average analytical brain but who’s by no means a superstar. I sweated (and still sweat) a lot. That being said, it helped me grow a lot and helped identify my strengths, sure, but mostly my weaknesses (read: where I plainly sucked and needed backup). Through this ego deflating process, I ended up understanding I can’t do everything myself and subsequently identifying the different types of people that were needed to have a successful data business.  I thought I’d share that knowledge with you. I’ll do this through a not-really-media-related example.

There’s a community bicycle service in my city called Bixi. If you are a Bixi member, you can pick up a bike at any of their stations across the city, use it at will, drop it back at any station when you’re done, and your credit card is then charged based on your usage time. There’s this obviously clever programmer that always ended up having problems with Bixi such as arriving at the location of an important meeting and facing a 100% full Bixi station “parking”, forcing him to go out and look for another station to drop his bike off. It irritated him… understandably.

Bixi provides a semi-live feed of bike &amp; parking avails per station, so the guy just programmed a robot that captured this data and created a heat-map mobile app. Over time, he recognized patterns in usage and was almost able to predict usage for any given time of the day, or day of the week. The predictive model was far from perfect though. Weeks of analysis helped him identify 3 variables that were messing up the predictions: Weather, Outdoor Events/Festivals, and Subway failures. Once these variables were identified, he programmed more robots to capture current weather, the subway’s Twitter feed and the city’s events agenda and his predictive model became much more accurate.

This app is useful for Bixi’s customers as it allows them to avoid future irritation. It is also useful for Bixi operators who can know where and when to move bikes from one station to another.

The programmer was talking about the app in a conference. To his dismay, he hasn’t been able to sell the app to Bixi, and has had little success getting customers.  Someone in the audience said: “Sell it to taxi companies…” The programmer’s eyes widened. Of course! Cab drivers would love to know where to position their cars as they anticipate transportation needs. Genius!

This example illustrates the 4 people/skills you need in your team.  Should these people be hybrids or pure specialists? It doesn’t matter, as long as you have them all.


<li>The Visionary. You need someone that will turn a business challenge into an idea, an opportunity, a concrete plan to measure something and draw insights out of it, just like this guy who thought of a simple solution for a concrete need of his. How often do you attend a client meeting to see them desperately looking for a solution to a business challenge, one they’ve had since forever, and no one in the room has ever been able to think of an idea to address it. That’s a huge opportunity loss. And then…  Someone on your team, someone meeting the client for the first time, goes: “We should address the problem THIS way.  Shall we craft a proposal?” Make sure to have someone on the team that just sees opportunity everywhere, and make sure that person gets briefed often / on everything.</li>

<li>The Programmer. I’ve often not pushed an idea just because I had no idea how to execute it, or if it was even executable. Then I’ll talk to a programmer/developer and he’ll say: “easy peasy, dude. I can do that in 2 hours”. It’s not because you’re clueless.  It’s because, just like me, you don’t know what technologies might exist out there to make your idea possible.  If you have an idea, ask a developer.  He’ll be better geared to confirm feasibility, quote a potential cost on the project, and you’ll be better equipped to solve the client’s riddle. Ask this programmer a lot of questions; ask questions all the time.</li>

<li>The Analyst. Creating an app with the perfect data set isn’t going to lead you anywhere if you don’t know what to do with it. When this guy talked about the 3 variables messing up his Bixi data, he hadn’t even finished pronouncing the word “variables” that I knew what they were. I could’ve saved him weeks of headaches and tests if I had happened to be friends with him. That’s the reason I have a good job: I see things instinctively… Rule of thumb: You need someone on your team that just sees correlation everywhere, and that instinctively finds methodology mistakes, false positives, and so on. Give that person the time and the latitude to dig: they usually hate filling insight-less surface dashboards that they know will lead to bad business decisions, but love being told “I can’t seem to find the reason behind XYZ, can you?”</li>

<li>The Sales pitcher. As good as the rest of the team is, if you can’t communicate what the real value of an idea is, and most importantly if you can’t get the client on a certain emotional level, you will never get the client to sign off, which means you will not get the money, which means you will not develop cool projects, which means you’ll eventually lose your qualified staff and clients to the competition. Some people are just awesome at sales pitching.  Find a geeky luxury car salesman, and give them the challenge paired with the proper retribution/commission for getting the signatures. Let them apply the pressure. They know when and how to listen, and when and how to talk.</li>


Now think of your Data department and ask yourself: Is the department big/effective/profitable enough? Is it maximizing results? If it’s not, do you think you lack one of these elements? Can you change or shuffle the existing ingredients a little? You might get a whole new set of answers. Food for thought…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: