So, who am I? Firstly, someone who likes to ask pseudo-existential questions on the internet. Beside that, I am a “Data Person” (certified Data Scientist, previous Analytics Architect and Analytics Lead) having hands-on experience with Web Analytics, Online Marketing Technology and Big Data Systems. Also I have a degree in Psychology (i.e. I sometimes ask “how does that make you feel?”).
Starting in Psychology
The last one seems out of place, doesn’t it? It is, so let’s start with that. Foy my whole life, I have always been fascinated with computers and large scale systems. During school, I learned HTML and C++, taught myself Visual Basic, PHP, and SQL and built some applications and websites for myself and others. So naturally, once I finished secondary school, the next logical step for me would be to study Computer Science.
At that time, Germany required you to do some obligatory military service. The only alternative was to go into civil services. I always liked the idea more to help people instead of being trained to kill them, so I went to my nine months of civil service, working in the operating room of a hospital in Frankfurt (sounds awesome? It was!) I learned a lot about medicine but more importantly, I worked with people in extreme situations for the first time in my life. And it worked! And it was fun! So why not study humans instead of computers?
Of course it was a bit more nuanced than that. Regarding Computer Science, I already knew everything I needed to build computers, as well as program and maintain websites and desktop applications. So studying CS didn’t seem so attractive any more, at least compared to “human sciences”. That lead to the next decision: should I study medicine or psychology? That one was easier. I always loved playing around with hardware and building stuff with my hands but loved software quite a bit more! And what else is psychology about, if not “human software”? So that’s what I chose, studying psychology and specializing in clinical psychology and organizational diagnostics. At the end of that, I even got the chance to publish some articles about my thesis.
Venturing into Web Analytics
During that time of study, I earned my living in customer service for some automotive companies while keeping my existing website customers happy on the side. At some point, I felt like there was nothing left to learn about customer service at my employment level so I started looking for something new. Luckily, my university was looking for someone to help out in the IT department (wrangling Excel sheets, im- and exporting data, managing accounts and databases, etc.) so I took that opportunity. Once they realized I know how to code, they moved me to the team that built and managed the self-built university management systems. We built the whole infrastructure, front- and backend, and integrations to other systems, including Salesforce (and some Apex stuff in there too). A lot to learn in a very exciting time of my life!
When I then finished studying, I was once again confronted with choosing the next step. My company back then wanted to keep me as a backend developer, but Unitymedia (nowadays know as Vodafone) was looking for a Digital Analyst to build up the inhouse web analytics team. Web analytics with Google Analytics and Piwik (or Matomo, as the cool kids say) has always been part of the package when I was building websites and, for myself, complemented reading logfiles server side. Together with the knowledge I gained about statistics and scientific working, it didn’t seem to far off of what I know. Sure, they used “Adobe Analytics” (that I never heard of), but how hard can that be? Unitymedia it was!
Moving to Data Science
Over the next years, I became an expert for Adobe Analytics while falling in love with the tool at the same time. It feels like everything is built around a few basic principles combined in very a clever way. There is so much more control over how and which data is collected and displayed to the user compared to other tools. It supports all levels of user-proficiency and gives me everything I need to deliver value to the business in the best way possible. I can’t count the times I thought to myself “yeah, that makes total sense” or even “I would have built it the same way”, raving about it whenever I can.
At the same time, I kept things fresh by extending my range to the other Adobe Experience Cloud Products, namely Target and Audience Manager. That lead to me being part of our Digital Conversion Taskforce and a company-wide initiative establishing omnichannel re-marketing. Outside of my day job, I started building things with the Elastic stack (Elastic Search, Kibana, Logstash), Grafana and the Hadoop ecosystem, including NiFi, for use cases like real time dashboarding, forecasting, and anomaly detection. I used that to build some nice real time dashboards for Adobe Analytics data, forecast our sales and traffic data with Python, and other fun stuff. At some point, we included IT monitoring data, collected by Dynatrace, to give a 360° view on our technical and UX performance. Besides the Dashboards running on some large-screen TVs, we used LaMetric Displays to give an easy access to our realtime data. It was a wild time and some of the best fun I’ve ever had.
All that work brought some attention to our department. With the next corporate reorg, our team grew three times (!) in size, allowing me to refine my role to the Digital Analytics Architect, emphasizing the ground I conquered so far. While I included more and more Big Data systems in my toolbox, I also worked on the knowledge and skill set behind them. This lead to me being certified as a “Data Scientist advanced level specialized in Machine Learning” by the Fraunhofer Big Data Alliance.
Moving on to the media industry
So everything went perfect, right? Well, yes and no. While the reorg brought some objective advantages for our team, it had a very bad impact on everything surrounding the team. To make things worse, Vodafone announced to acquire us. All of this lead to changes that spoiled what was fun about the work. So, I tried my best to fix things and decided to move on once that has proven ineffective. That was a very, very tough decision.
Because the job market for analytics in Germany is what it is, it didn’t take long for me to find a new job when I decided to leave. Looking for a new industry to work in, I was cast to take over Digital Analytics at a market-leading TV broadcaster that also operated an assortment of websites, mobile apps, and also apps on other platforms. That sounded like fun!
It was an exciting time, but for other reasons that I have imagined. While I have made a lot more impact than I have planned (preventing a move from Adobe Analytics to Google Analytics, building my own team of four people, creating a data strategy based on self-service and enablement, creating a Business Intelligence practice with Microsoft Power BI, etc.), the amount of buy-in and support I would receive left a lot to desire. In addition, all of that great progress was met with only moderate levels of interest or even opposition from the rest of the company. Even worse, innovation, learning, and iteration was actively fought and discouraged. That wasn’t fun at all. It only took two weeks at that job for me to say “I feel like a consultant who was brought in to fix the analytics practice and then leave things in a better state for the next guy”.
So the real challenge turned out to be the company culture outside of the (then formed) analytics team. Part of that was caused by the company’s history as a TV broadcaster, where actual user data and feedback were unavailable and the capability to work with that was therefore very limited. My team and myself worked very hard for two years to change that attitude and culture (supported by an awesome head of product) and made some tangible progress. Unfortunately, driving the change with only a few supporters while fighting against a bad habituated culture burned me out to a point where going to work wasn’t any fun anymore.
There was only one way I could handle that very bad situation: Leave the place as soon as possible, but in a better state than I found it. So that’s what I did. Within two years, I’ve successfully built a team, processes, tools, and a bit of culture where I could, at least up to a certain point. As the company reached a certain level of where people understood why they should be doing analytics, I started to plan my exit and brought in my old mentor to hand things over to. Then I left.
It’s always darkest before the dawn
I know, that all sounds a bit dark. In that, it’s very authentic to how I felt at that time. I was heartbroken because I had to leave my team behind and burned out from two years of struggling with the wrong culture. To my own surprise, things got better a lot faster than I could ever have hoped for when I joined the Digital Analytics team at DHL. Suddenly, work became fun and rewarding again. The people are great and the projects are diverse and exciting. I’m in a much better place today and can’t wait for the amazing things we will do in the future.