MeasureCamp – A slice of an in-demand analytics event
Measurecamp is a mighty popular global ‘unconference’ for analytics professionals.
In 2012, a small number of the analytics community decided to meet and share their knowledge. Five years on, this has become one of the hottest events in the analytics industry, with tickets selling out in under a minute. This event is free and held on a Saturday from 9 -5pm – it’s that popular that people are happy to give up a big part of the weekend for it.
It consists of sessions from analytics professionals to discuss what techniques they are using in their company. And it has provided a platform for over 300 digital analysts to generously share their ideas with each other, learn from others and network.
One of the recurring themes during the March 2017 MeasureCamp – we were lucky to be there - was that, the digital analytics landscape is rapidly changing. Ecommerce has been growing at a fast pace – recent statistics from Statista found that in 2015 global online sales totalled $1.55 trillion dollars. Emarketer forecasts that this will only increase – with ecommerce sales estimated to reach $4.058 trillion in 2020, amounting to 14.6% of retail sales. The increasing number of people shopping online means that technical analysts are now able to capture more customer behaviour data than ever before. However, it also means that as more people shop online, website usability needs to continually improve and adapt to keep up to date.
So what’s the impact for digital analytics?
It means that as more people shop online, there is a greater volume of data to analyse and provide actionable insights from. Whilst reporting on a website’s performance is still very much important for businesses, analysts are now able to incorporate data science and machine learning techniques in their analyses. There were seven sessions that covered the area of data science at MeasureCamp Data science uses statistical modelling to make predictions about future events. One powerful study that shows the importance of statistical modelling was discussed at MeasureCamp, whereby American psychologists John Gottman and Robert Levenson used the power of science to predict relationship longevity based on a short observation of 78 newly-weds. This study revealed that by using an algorithm to predict divorce, they were 93% accurate in which marriages would end in divorce.
Now, utilising this in digital marketing is an effective technique as you can understand how customers are likely to respond and react in real time. You can use predictive modelling to measure the optimal time customers are likely to return to the site (useful for marketing attribution) or you can look at whether people are leaving your site due to page load time or another time component. Data is growing in volume and by modelling this data can help to improve our understanding of customers and their journey. Not only do data science techniques provide practical importance for ecommerce sites – they also help businesses make confident decisions.
The Digital Intelligence team at Salmon is now using data science to make data-driven decisions for its clients. New software tools and fresh innovative ideas are proving successful.