Starting an Analytics Revolution

WITH AXIS GROUP BEING THE GOLD SPONSOR FOR THE SOUTHERN DATA SCIENCE CONFERENCE, WE FELT IT WAS 100% APPROPRIATE TO HIJACK THE MOCK-UP MONDAY CHANNEL TO TALK ABOUT ADVANCED ANALYTICS. VIVA LA REVOLUTION!

You say you want a revolution

Well, you know

We all want to change the world.

-The Beatles

Analysts, data scientists, and enterprising managers all over the world are trying to start revolutions. And some are even trying to change the world.

Most would-be-revolutionaries are finding that it’s difficult to revolutionize an existing organization. More often than not, what they are missing is PURPOSE.

LET'S LOOK AT TWO HISTORICAL POLITICAL MOVEMENTS TO SEE WHAT WE CAN LEARN

First - American Revolution

Driven by sense of purpose and unity, the American colonists (with some help from the French) stood up together in the late 1700's and successfully rebuffed the British and established independence. Their purpose was simple and well known - independence from the Crown. The colonists were able to rally around that objective and come together and revolt.

Second - Occupy Wall Street

Regardless of your opinion on the movement, Occupy Wall Street is an interesting study. OWS began in 2011 as a collection of people protesting global income inequality. In general, the protesters were fighting against influence of corporations on politics, income inequality, and the power of institutions. Despite having outlines of things that they were opposed to, the movement lacked a specific goal or piece of legislation to rally behind. Some leaders of the movement went so far as to reject the idea of creating a set of specific goals in general. In the end, one of the reasons that the movement fizzled out was due to a lack of a concrete purpose to move toward that the members could agree upon and rally behind.

While admittedly being oversimplifications of historical events, these two examples still demonstrate the importance of having a PURPOSE. Here are three ways this applies to your analytics revolution.

  • “Build it and they will come” is a bad idea

    As consultants frequently performing autopsies on analytics programs, we see this approach more often than you would imagine. 9 times out of 10, this strategy is a recipe for failure. The leading cause of unused BI applications and failed analytics projects is the lack of business buy-in from the start.

    To understand why, think about the how a typical product from an analytics organization attempts to realize value. Inherently, to gain value, the business will have to change an existing business process. Put yourselves in the mindset of a VP of Marketing (insert any business unit) when someone they’ve never met approaches them and says:

    “I’ve done some analysis on your data and you’ve been doing it all wrong. You should be doing this instead. Here’s a new dashboard.”

    There is no better way to make someone defensive than to tell them that they have been doing their job incorrectly. It is also more than a little bit presumptuous. To be successful, a better strategy is to partner with the VP at the beginning with a conversation like this:

    “I’d like to talk to you about some ideas I’ve had that can help you blow your last year’s number out of the water. Let’s partner to add value to our organization.”

    By securing buy-in at the beginning, you’re enabling the other side to have a stake in the process and your adoption rate will skyrocket as a result. Your final product will be stronger as well.

  • Prove value quickly to keep your momentum

    Nothing stagnates an analytics initiative faster than long timelines. If you offer an executive a time-to-value in months rather than weeks, you’ve already lost them. By the time you finish, they are already on to something else. The key is to show value quickly. As this is one of the biggest problems we face, we’re always seeking new ways to increase our speed to results. We recently formed a strategic partnership with an analytics software company called Emcien specifically for this reason. Their toolset allows us to turn around insights in days. That’s how important this piece is.

    Start with putting out simple reports that show them new insights about their business. Don’t be afraid to ask them to commit weekly or bi-weekly 15 minute catch ups where you can demonstrate the value that you’re creating. If this sounds like a sales pitch, that’s because it is. Analytics practitioners must always have a sales approach when dealing with skeptical internal stakeholders.

    At the end of the day, what you’re trying to accomplish is getting the business to buy in to your way of doing things. Do this by constantly proving value to keep their interest.

  • Leadership cares about the dollars and cents

    If you start a project pitch to an executive and mention technology or algorithms or dashboards before you layout the ROI, you’ve already lost. The old cliché of having 30 seconds in an elevator with the CEO to pitch your ideas is a cliché because there is some truth to it. The people with the power to fund your ideas and initiatives are busy people who are responsible for ensuring that the company is profitable.

    As a practitioner or manager, you should be able to articulate the business value of your idea in under a minute. For example, if I were trying to convince the head of Accounts Receivable that I had a new approach that would help his or her organization and I only had a few seconds I would lead with something like this:

    “I’m Nathan from the Data Science team. I can help your team improve your collections by 30% over last year. When can we meet to discuss further?”

    Instead of:

    “I’m Nathan from the Data Science team. I saw a webinar last week about applying predictive algorithms to accounts receivable. I think there may be a way we can use Python and some other open source tools create a machine learning algorithm that models the behavior of our customers and helps your team.”

    See the difference? The first one is focused on the business value. Methodologies can be discussed at a later time (if at all). The second one focuses on the process and puts the value added to the team as secondary to the creation of an algorithm. Before you pitch a project, spend a few minutes to think critically about what the recipient of your pitch actually cares about. Do you think it’s their bottom line or your algorithms?

    Finally, here’s a simple checklist for pitching new analytics ideas:

    - Understand and document the business case and expected ROI up front
    Create small initial wins quickly to build momentum and support
              - Have a vision that you can evangelize (read: sell) to other stakeholders                            - Parade your successes around the organization to find new stakeholders

You say you want a revolution. Make sure someone outside your team buys in.

The Contributor:

 

Nathan- CircleNathan Hombroek is the Data Science Practice Lead at Axis Group. He is an experienced analytics leader and practitioner heading a team of Data Science Consultants that solve complex business problems with data and analytics. He is passionate about architecting creative business solutions and mentoring the next generation of analytics professionals. Nathan has a BS in Economics from Georgia Tech and an MBA, Business Analytics from Georgia Tech.

TAGS: Data Science, Mock-Up Monday, Analytics

Related News

In a large business, overutilization of supplies and expenses can amount quickly if you are not...

Read More
Read More