Duneybe

As someone who once preferred to bake cookies rather than study for statistics exams, the works easy and statistics don’t always go together. However today, at a presentation by IBM and Azendian, the math indeed seemed easy. There are three places analytics are commonly used - new business (above the line), greater efficiency (below the line) and managing risk, This session looked at the operational (below the line ) side, an area not often cited in the examples of analytics in action or practice.

Using an example of HR analytics in looking at attrition, our speakers covered the many non technical areas in an analytics project that are important for success Data governance - who should see what information. Process & policy - how does the use of an analytical model impact existing process and policy Buy in - many senior executives, the ones to sign off and sponsor projects, got where they are from experience. Datadriven decisions, particularly non intuitive ones, may not come naturally to them Communicate with your audience - don’t try to sell analytics with fancy math, sell easy to understand business solutions Business drive successful analytic projects, not IT

At almost no point did maths come into the equation to make analytic projects successful. Citing examples from BWM (reducing cost of brake light errors due to sunroofs), Rugby (keeping your star player free from practice injury https://community.spiceworks.com/topic/553729-novel-use-of-data-analytics-helps-new-zealand-rugby-team-pull-off-epic-upset ) and bike teams http://www.bbc.com/sport/olympics/19174302 that turn 1% improvement to huge success.

Phew, hopefully you got through your proof of concept. If not, don’t worry.  We didn’t fly off the pad on day one either.  In fact it took a while and a few cases to get us going.  But one element that was a big success was getting around our data and interacting with it. 

Tableau as a tool is fast.  Both the processing, the chart changes and the ability write a quick function and apply it.  Vlookup works but is a bit clunky, take a bit of effort and prone to error.  So much better when its on the fly to see results and adjust as funny data spews out errors.

Early on our journey we had a challenge that needed all those tips to save a not insignificant amount of money,  We had some flat data to analyse, a gut feel for the result but nothing to prove it.  So we grabbed a few people, a laptop and the still shiny Tableau tool and got into a room with a projector.  We didn’t really stage the data, aside from setting up the format like the powerpoint in question.  From there though, the questions flowed.  And we could get the answers fast enough, in particular with two of my team in the room to help - one to write functions, drive the computer, one to drive the conversations and the questions.  An hour later, viola , we had enough of the answer we needed.  

Interaction with the data has been a constant through our journey and we don’t normally just stop with one.  Getting people around the data, asking the questions and seeing what develops creates some great insights, and enthuses your audience about what is possible.  

Tips on the interacting with your data

- have some interesting ways to visualise the data handy in your mind, something non Excel, to spark conversation.  Sometimes cutting the data along day of the week and density sparks a conversation.  It also breaks people out of  what is normal with excel

- don’t over prepare.  This is about exploration- know your way around the tool and how to write a function but don’t go in with foregone conclusions that end up shutting down the conversation

- do it in a room, together. This is a pace the room, point at the screen, quite people by looking in the eyes kind of exercise. It may not work well as a packaged presentation on VC where , lets face it, someone is not fully there

- try it out with a friendly audience with no real deep question.  Practice in the flow helps when its time to answer the really questions that may exist.

April, 2016, hundreds of data people gathered at the Marina Bay Sands convention centre for the 2016 Tableau on Tour Singapore event.  It was my first event to join, and I was lucky enough to speak to nearly 100 people on how we’ve taken our journey from those early days in the proof of concept to today, were my team of data visualisers, data engineers and BI specialist support a vast number of data sources, reports, dashboards and users across the globe.   

Personally, it was a blast.  And a learning experience as well.  Getting ready for the event got me doing a lot of background work, picking up some solid tips to make a better presentation (and even more of those from watching the speakers).  But before that, I’d like to take my topics into a bit more detail, and a bit more tangible (aka you can refer back to them) posts online.  So in the coming weeks stay tuned for some sharing on how you can Start small.  Be ready to grow big.  Fast. 

Ahead we’ll look at 

- the proof of concept

- Interacting with your data

- publishing interactive dashboards

- forming a user group

- connecting with a COE

- remembering your audience

- continuous learning

- making it natural

- and the bonus tip. 

In a few short days, I’ll be presenting a talk at the 2016 Singapore Tableau on Tour event.  Really looking forward to the event, to learn a lot about Tableau, connect with some amazing people and with some luck inspire some people in the audience to succeed with Tableau as well.  

There is a lot we do with Tableau, a lot to learn from courses, experience and books to read on the tool, on why visual information is so powerful, on improving communication and building a successful team to make it all happen.  Here I’ll package those experiences up into hopefully bit sized bits that help and inspire others to break away from grids and pages of numbers that are so common in the working world today.