On the 19th of June, Sam Roberts, our Digital Marketing Consultant, took part in the CIM Wales Analytics Masterclass. Alongside other speakers, he took to beating the jargon out of Analytics in the hope to provide people with a better understanding. Here is a round-up of the 5 key takeaways.
1) Machines are not to be feared!
What is machine learning?
Machine learning is a method of data analysis that automates analytical model building. It is a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention.
Why does it matter to me?
Machine learning allows us to analyse data better and on a larger scale. It has become more popular due to things like growing volumes and varieties of available data, computational processing that is cheaper and more powerful, and affordable data storage. Therefore, we can analyse bigger, more complex data and deliver faster and more accurate results. Furthermore, by building precise models, an organisation has a better chance of identifying profitable opportunities or avoiding unknown risks.
2) Cultural changes can lead to big things
To best describe how cultural change can lead to big things, I will hand this one to the other speaker Mans. Mans has a big job to do, he has been asked by the board to use data in all decision making however his team is very small. Therefore, he ‘created a culture of test and learn’ by training 70 web journey managers in split testing and essentially got them doing his job for him. This saw a growth of 3 people in a team to 50.
3) Bad data visualisation can have serious consequences
Not only is bad data visualisation misleading, but it can have serious effects on how people perceive data, which in turn, can lead to bad decisions being made.
How do I avoid this?
Common errors include data duplication, missed data and NA values not marked. Checking that the data is clean at every stage through the data process (collection, sourcing, cleaning, and compiling) will guarantee common errors will become a thing of the past. Ensure that your data is visualised correctly, there isn’t too much information or too many colours, and that your data isn’t misrepresented.
4) A reporting strategy is vital for business intelligence
Without an idea of how you will report, what reports to make, who will need them and why you won’t be able to get the most out of any strategy. It is vital to have a reporting strategy in place to ensure your business intelligence is being leveraged within the organisation.
5) Bad attribution costs jobs!
For those who attended the masterclass, the attribution model of creating a baby will still be imbedded in your brain (for good reasons I hope). For those who weren’t, here is a quick recap:
1. Flow works with John
2. They go on a blind date
3. Love blossoms
4. They get engaged!
5. On holiday they go on a Gondola, the driver is called a gondolier, the gondolier is added to the model
6. They get married adding the Registrar to the model
7. On their wedding night candles and Lionel Richie leads to… well this doesn’t need to be explained. Lionel Ritchie is added to the model
8. And finally, they have a baby! Adding the midwife as the last attribute in the model.
Analytics will only look at the last interaction, therefore making the midwife the person who made a baby. Whilst we can all agree the midwife helped incredibly, this does not give an accurate representation. This is called ‘last click attribution’.
In order for attribution to be correct and effective, you must create your own custom attribution model, due to built-in models not giving an accurate representation.
View Sam’s full presentation hereCIM Attribution Pres.compressed
Want to learn more about Analytics?
If you want to learn more about any of the 5 key takeaways or any other aspect of digital marketing then get in touch today to find out more about our training services.