Faster, More Productive and More Individual
Text robots for classic customer communication
Texts are increasingly being generated by computers. Notable media companies such as Forbes, the New York Times and the Washington Post have already begun to automate parts of their news production – that means getting programs to write them. But how can texts written by a robot’s pen be used for customer communication?
As the website www.radaronline.com reports, Jessica Biel is expecting a baby! This page is based on the statements of the US actress’ close friend.” This text was written by a robot. And that is no longer a surprise. More and more texts are now being generated by computers. The news agency AP uses such software to create around 4,000 sports and finance reports each quarter. In the US election campaign, too, readers and users received an increasing number of texts that had been created by algorithms and robots. Renowned newspapers such as the New York Times and the Washington Post, as well as broadcasters like CNN and NBC, are among the media companies delivering customized news via chatbots. Why do more and more media companies rely on text robots? They have three key advantages: They are cheaper and more productive than their human colleagues, and don’t make any careless mistakes. A software can aggregate data for a bulletin and report about events that you couldn’t otherwise read about in the media. The bases for this are computer linguistic processes, syntactical rules and structured data, which the text robot adds to templates and then puts together in a logical, coherent text.
New opportunities for classic customer communication
But everyone who praises text robots for their speed and nearly faultless quality should also take into account their limits and further applications. A text robot is, after all, nothing more than an automated text software programmed and trained by humans. It can only generate its texts on the basis of structured data which, for example, could come from the product data sheets of a merchandise management system. It doesn’t learn and think “independently”, rather it completes the series of tasks that it was programmed for quickly, precisely and without mistakes. That’s why text robots are used for time critical and serial tasks. This includes not just the sub-sections of journalism (weather, sports and finance); in the area of e-commerce (product descriptions in fashion retailers’ online shops) and marketing, as well, text robots will be employed even more in the future. Why? Although the text robot doesn’t answer directly and quickly like a chatbot in customer service queries, it does offer classic media channels new ways and opportunities to communicate with customers. For example, by individualizing and personalizing serial and high-volume media such as business reports, newsletters, mailings and catalogues. This is because, with the help of software, texts can be created which along with a personal address, contain individual product proposals and proactive purchase recommendations. These are generated according to the customer’s previous order history and their sociodemographic data. In order to do this, the text robot must be “fed” with the respective customer data from the merchandise management system and from past marketing communication. It can then link these together and cluster them “independently”. Yet another positive outcome: the data that a company collected in the past about its customers, which may be spread across different departments such as sales, marketing and logistics, will be brought together, activated and animated for communication with customers. By doing this, the text robot will become an extremely powerful tool in marketing and in discourse with customers if it is implemented correctly – and this on all channels!
Author: Editorial team Future. Customer.