http://www.aquilainsight.com/wp-content/uploads/2018/11/Amanda.jpg

You’ll remember from school other students stopping you from seeing their answers, by putting their arm around their exam paper.

It’s the same at work, people are secretive about their ideas as though there’s a limited supply of original thinking in the world, and a constant need to protect every innovation.

Except at Merkle Aquila, where we believe the more you give away, the more you get back.

It was in this spirit of sharing that we headed to London to present at the Enterprise Applications of the R Language Conference. EARL is a cross-sector conference focusing on the commercial use of the R programming language. Held over two days (11-13 September 2018), the conference is dedicated to the real-world usage of R with some of the world’s leading practitioners.

Our presentation was on an internal project, focused on learning about novel techniques and identifying ways to improve work efficiency. Here’s the blurb for any R aficionados:

Understand your customer base in a matter of hours with the power of Shiny and AWS

Increasing product relevancy is a common challenge that businesses face and one of the first approaches to addressing is customer profiling. As a consultancy firm specialising in consumer analytics, it is also one of the many services we provide to our clients, and often the first step of an analytics consultancy process. Whilst the task is frequently carried out, the data structure and information tends to vary greatly. To address this issue, we as a team decided to create an analyst tool using R Shiny, intended to achieve three things: (1) provide interactive and semi-automated data visualisation of customer level data, (2) easily enrich our data with 3rd party data, (3) and produce accurate insights consistently and efficiently. In this presentation we will share our approach and the process – from building the dashboard to exporting analysis results. Furthermore, we will also cover the challenges associated with building scalable complex R Shiny application for corporate use, such as the use of AWS and optimization tricks implemented to provide performance gain.

Being able to give a presentation is an important part of sharing our work and getting recognition from our peers in the industry. But it all starts with communicating the relationship between complex analytical solutions and business benefits, to our clients. If we can analyse data, but we can’t communicate the story well, well what’s the point? No one will act on our analysis, and no change will occur.

Some people have a gift for data storytelling, and it’s a hugely valuable skill. A lot of data analysts, in my opinion, are good at the analysis part, but giving a presentation may not be one of our strengths. At Merkle Aquila, I’m one of the technical leads of the internal innovation team, specialising in data mining, predictive modelling, segmentation and clustering, A/B testing as well as customer profiling using R. The conference was an excellent opportunity to practice communicating our insights from this particular project.

The challenge we faced was to summarise everything, concisely. Explaining why we wanted to do this project and give an update – in only 20 mins – was difficult. But, we often have shorter client meetings, so it was good discipline. In my experience, transforming business data into actionable insights is the easy bit. Successful solutions are often made by people who are good at data storytelling. I’m still working on the last bit.

If you’d like Merkle Aquila to speak at your next event. Get in touch


Back to News