Learnings from presenting at CPA Australia 

Last week I was invited by CPA Australia to a lunch presentation session to showcase analytics to  government organsiations. It was a fantastic session with many departments represented including Aust police, Health, Department of education and many more NSW government bodies.


The purpose was to facilitate discussion on how to best implement analytics to deliver business value.

My presentation can be found here  for anyone that wants to leverage it.

Some of the key challenges include the usual legacy systems but also most interestingly the challenge around people’s mind sets and defining the business problem. Often in analytics many BI managers are fixated on the latest and greatest technology that they forgot that the Tech is meant to serve a business purpose.

Framing the business problem and articulating the value of analytics can be one of the most important steps in leading an analytics team.

Another interesting point of discussion was the number of data scientists required in a team. A lot of people have over hired or under hired data scientists. To those new to data science it’s just a fancy title for applied statistics or maths or physics.

Ensure there is a good ratio of 3-4 business analysts to statisticians (data scientist)

The optimal ratio is really 3-4 business/ data analysts to 1 data scientist. Nearly 80% of a data scientist work is preparing a base table and not modelling and so it’s much more efficient to hire a cheaper resource to help with the task. More importantly than just hiring brute force numbers is to empower business analysts with the right tools like alteryx to help with not just the data portion but also the Spatial, predictive and optimisation parts.

Use best in class tools like Alteryx , Tableau and AWS to overcome technical challenges

Another common misnomer among people is that a data scientist needs to know R, python, Scala etc. in my view it’s more important to be able to work with the business and ask the right questions , framing the problem and then using alteryx to overcome the technical challenge.

This not only reduces time to deliver analytics but reduces key person risk when an analysts moves on the role. Recent developments in alteryx  including alteryx connect ( search and metadata repository) and alteryx promote ( productionising data science models ) have taken the product up a few levels to become an enterprise data science platform.

Start small iterative quickly

Another few interesting points included the need to start small iterate quickly and move away from writing those 100 page RFPs. After all if u really want to know how a product works u test drive it just like test driving a car. No one reads the technical specs and tries to compare each point, instead it’s better to actually try and product and form your own views.

Analytics can be game changing if leveraged correctly – Make sure Alteryx is pack of the technology stack

Organisations early on there analytics journey should also embrace visualisation and great to showcase analytics in this way to get  management  buyin. Later on in the analytic journey however we need to focus on use analytics to deliver game changing capabilities to the business. Just listening to the latest group of customers at Alteryx Inspire Europe reiterated this. Check it out below

https://www.alteryx.com/inspire-europe-2017-tracks

Listen and learn to Alteryx customer presentations

My favourite customer presentations are by Shell, bookmyshow and Asahi. The keynotes by Dean stocker and the product demos were incredible. Didn’t realise u could use alteryx to read in photos and hook up with Microsoft API for face recognition. This is game changing stuff from a business perspective as it can be applied to recognise customers and improve customer experience.

That’s all for now.. Stay tuned for my next instalment..