Monday, 25 May 2015

Ahead of the game

In Vincent Granville PhD's book Developing Analytical Talent: Becoming a Data Scientist he writes an interesting section on the history and pioneers of data science. He highlights past developments and predicts what will happen in the future. I quote from his book.
  • "1988: Artificial intelligence. Also: computational statistics, data analysis, pattern recognition and rule systems.
  • 1995: Web analytics. Also: machine learning, business intelligence, data mining, ROI, distributed architecture, data architecture, quant, decision science, knowledge management and information science.
  • 2003: Business analytics. Also: text mining, unstructured data, semantic web, Natural Language Processing (NLP), Key Performance Indicator (KPI), predictive modeling, cloud computing, lift, yield, NoSQL, Business Intelligence (BI), real-time analytics, collaborative filtering, recommendation engines and mobile analytics.
  • 2012: Data Science. Also: big data, analytics, software as a service (SaaS), on-demand analytics, digital analytics, Hadoop, NewSQL, in-memory analytics, machine-to-machine, sensor data, healthcare analytics, utilities analytics, data governance and in-column databases.
  • 2022: Data engineering. Also: analytics engineering, data management, data shaping, art of optimization, optimization science, optimization engineering, business optimization and data intelligence."
After I've read this I realized that Autolytix Data Science is ahead of the game. For the past 5 years we have been working on solving optimization problems, which Granville notes will only be available in 2022.

Optimization in any subject matter can be difficult, but our approach was to build the gaps between analytics and profitability.  

In order to bridge this gap we developed what we call the data science maturity curve, to start understanding the link between data and profits.

The maturity curve was a good start, but we still needed a link between the analytics and profits. This is how we solved it.

In order to make the analytics work for you in your business, you need to quantify the impact a change in the variables may have on the business. For example: if you spend $1000 per month on
stationary versus $100,000 on raw materials, a 10% change in both these expenses will have a significantly different impact on your bottom line. This is where prioritization will come into play. You will prioritize a 10% saving on the $100,000 raw materials above that of stationary.
To execute this saving, you will need to plan for this change and we use a strategic initiative to assess the variables that need to change and how it will be changed to affect the result you require.
For example: In a strategic initiative we will recommend which suppliers to eliminate and/or which suppliers need a pricing review, to renegotiate contract pricing. There are many ways and means to affect this change.
Last step will be to implement this change. Lastly, when drafting your strategic initiative, also be cost conscious. Sometimes, it will cost you more than the benefit you will receive.

I hope this short article has given you some idea of how to turn data into profits.

1 comment: