What does a babies bottom and a dataset have in common?


One of the joys of grandparenting is that I get to spend time with and indulge the little people in my life and then hand them back, after all I've done the hard yards already. Occasionally I am known to change a nappy, and it is still sometimes a marvel or surprise to see what joy awaits, the sniff test can give an indication, but the truth is never known until the package is unwrapped. Years of practice can mean that in a short time, the little person is cleaned, covered in sudocrem, dressed and sent on it's wriggling crawling way much happier than when we started.


You may be quite rightly asking why I am wandering down this strange path, well the reason for my mental exercise on this topic is that last week I spent time with two of the little people in my life and bless one of them (they are no doubt pointing fingers at each other) brought the winter sickness bug back from nursery for us to revel in, one by one. So as you can imagine babies, hygiene and potential impact has weighed heavily on my mind and elsewhere this week.


Anyway, possibly whilst in a hallucinatory state I was thinking about mess and cleanliness, and started to think about data modelling, don't ask, sometimes my mind just works that way.


When the request is made "Can I have a report in Power Bi that shows me....", the answer is usually "Yes" followed by "What data or information have you got for me to work with and where is it?". This when things get interesting. The 'where' isn't usually a challenge, there hundreds of connectors and ways to gather data from a whole variety of sources, SQL, Salesforce, Xero, Quickbooks and on and on, one way or another the information can be collected ready for use.


The challenge really is getting the data into a shape or format that can be used in the most efficient and consistent manner. This is called data modelling and data cleansing and this where the majority of time is spent, in preparation for the report itself.


Think back to the nappy, you have sniff and tentatively unwrap, it's a mess, it needs cleaning, soothing and preparing before the toddler goes on it's way. The same with the data, chances are it's a mess, numbers, columns and fields splattered all over the place, method has to be found, the data cleaned and prepared for use before we can move onto the next stage.


As a business grows the information stored and used grows, usually without a grand plan just a need to get what's needed when it's needed. However when presented to me, if I fail to spend the required time here with the data model I will be regretting it and creating work further down the road. Bit like that nappy again, fail to do your job properly and a sore bum and tears await....


And there you have it, good data cleansing is just like changing a nappy, time, cleanliness, order and attention to the task in hand are important to ensure a happy outcome.



Recent Posts

See All

It is now three years since I left my 'proper' job to add other and varied interests into my life. I loved what I did, I just reached a point where I didn't want to do it everyday, change things aroun