Interview Christina Rettig: AI in communication

Christina Rettig is head of communications at Schott and co-leader of CommTech Working Group 3 ‘Reporting& KPIs’. As a PhD student at Antwerp Business School / Universiteit Antwerpen, she is currently researching cyber-physical systems and the impact of CommTech on communication departments.

The interview was conducted as part of the webinar “How Artificial Intelligence Democratizes Media Analysis,” hosted by IMWF on March 23. Here, she answered the following question, among others:

Thomas Mickeleit: Christina, you are not only the head of communications at Schott, a thoroughly data-driven company, as we will learn, but also the co-head of CommTech AG3 “Reporting”. Do you see yourself spearheading data-driven communications, or how best to describe your role?

Christina Rettig: In fact, data-drivenness at SCHOTT refers to the fact that we talk a lot about product and production data, and like most companies, we are managed based on numbers and data. That’s when the question matured in my mind of how we as communicators could present ourselves in the company in such a way that our value contribution is understood. And whether we could not even be a pioneer in the use of modern technologies. That was my motivation to get involved in the CommTech WG, together with Antonia Dieterle (Sennheiser), Oliver Loenker (Siemens Healthineers) and David Willmes (SCHOTT).

Thomas Mickeleit: It certainly doesn’t hurt to understand what goes on in the engine room. But does that mean you have to be an IT specialist if you want to use artificial intelligence in communication?

Christina Rettig: First of all, you just have to remember: AI means that computers are able to solve problems on their own. Machine learning is a subset of this and means that systems learn from sample data and recognize patterns from a large amount of data. So I have some form of self-learning system that I show it input and target data, or reward it when it shows me correct results. Or I simply let the AI search freely in the data for patterns to build customer segments, for example. For a first overview, by the way, I recommend the YouTube series Startup Teens on AI. What you need above all is an interest in getting involved and trying things out. You never know exactly what the AI will work out of the input and what insights you can derive from it. That’s what’s so exciting about it!

Thomas Mickeleit: Not everywhere that says AI on it is AI in it, but the reverse is also true. Not everything says that AI is in it. Can you give us real-world examples of this?

Christina Rettig: At university we were always told: If it’s written in Python it’s probably Machine Learning, if it’s written in PowerPoint it’s probably AI. In other words, a lot of things are hype, terms are mixed up. But what you call it is really secondary – it’s what you do with it. Right now I’m experimenting with SSML (Speech Synthesis Markup Language) for a read-aloud feature of our new online magazine. One finding was that speech output works better when the AI automatically translates the text beforehand and then speaks it in the target language. No one tells you this beforehand, you only find out by trial and error. I also find AI-based content analysis of videos interesting. For example, I could train the AI to show me all the video sequences in our internal database that feature our CEO. At the moment, this is purely experiential knowledge of the team members; in the future, an AI can take over this.

Thomas Mickeleit: What barriers do you experience when implementing IT projects at SCHOTT. Tell us concretely about your experiences

Christina Rettig: We are lucky to have great contacts in our IT. In general, however, I notice that communications departments often need special solutions that no one else in the company uses. And especially in the case of AI applications, the providers are often startups that are difficult to reconcile with the purchasing conditions of a full-blown IT organization. What’s more, digitization initiatives initially focus on core processes within the company – production and sales, for example. As a comparatively small department, communications falls through the cracks. Digital transformation requires that you transform the entire company and not just parts of it. And the communications department can better support the process if it lives and breathes data.

Thomas Mickeleit: How do you see the future of “data in PR”, also with a view to the much-invoked “data culture”?

Christina Rettig: Data is the key to two things: First, that we make communicators better understand our value proposition in the company. Second, that we generate exciting content from the data itself. At SCHOTT, we are currently using Google Data Studio for this purpose, mainly to evaluate our social media performance. That’s where we want to dive deeper in the future.