This article on data planning is the second in a series of blog posts exploring value in public sector data:
We’ve listed four key aspects of data planning that if carried out in sequence with the Public Sector Data Mapping blog:
Then you’ll have the ingredients to understand and measure what data you have and be able to measure improvement or deterioration in data.
4 Vs of Data
Data courses the world over will very quickly introduce students to the 4 Vs of data: Volume, Variety, Velocity, and Veracity. We need to start measuring these for each source of data throughout an organisation.
Volume is measured by looking at how much data there is measured memory size such as megabytes or gigabytes.
Variety is often two things in one first it is a look at the different formats and types of data. Is it an image, video or text, perhaps a database? Second, it looks at the complexity of data a database that has two columns of data, e.g. Name and Phone Number is a lot less complex than a data set with a hundred columns of data describing a person’s purchase habits.
The velocity of data is exactly that, how fast data is being generated or coming at you. Is it a database that is updated once a month such as budget and actual figures? Or is it Amazon’s credit card transactions which probably just made 10,000 new lines while you read this sentence?
Finally, the Veracity of data refers to the accuracy of the data. An example of how this is measured is seen when you look at a database of names given at a restaurant versus names that are checked against ID at passport control. I never use my real name at a restaurant, I generally go for Pierre Laconte when booking a table.
Open Data Plan
This is as simple as pulling together an Open Data Plan. Now, this probably already exists if you are looking at the public sector in Scotland and the UK, but reviewing that plan and really getting it implemented throughout an organisation is a different kettle of fish.
Essentially data falls into three categories as defined by GDPR: Sensitive Personal Data – sexual orientation and the like, Personal Information – name and location are examples, and Open Data – absolutely everything else. A good open data plan will only collect sensitive personal and personal data if they are required to by legislation. Absolutely everything else unless there is other legislation or policy prohibiting it (secrecy act, commercially sensitive etc.) can be released into the wild. Just think how amazing that would be.
Data Management Plan
Different from an open data plan a data management plan this is how you modify your processes and practices regarding all data throughout your organisation. How is it stored, secured and distributed? Pulling this work together helps you as an organisation to understand the decisions you make regarding data. Ideally, it makes them easier as the work spent to understand what is possible or not possible with data has already been completed.
This is a fun subject. What are the policies and the processes that relate to data? How can we enforce and encourage good data practices? Then how do we build a team to work with this as well as bring along the entire organisation in understanding and action in line with these principals?
These are really the key rules that define what data is collected, where that data is kept and what you can use it for. When this is in place and working well you’ve got a tight ship.
Until next week. Love, the Wittin team!