AI in Marketing Workshop - Text generation tools


AI in Marketing Conference at ZHAW

Everyone talks about artificial intelligence (AI) in marketing. But what is the current status quo of such technologies in the creation of concrete content for digital marketing? We tested that at a workshop for the Future of Marketing day at the ZHAW in Winterthur, in Switzerland.

Around 170 marketing managers, employees, and students participated in the event to discuss the question of where the future of marketing is headed. This event takes place annually and brings science and practice together. In a total of six different deep-dive sessions, various topics were discussed. For the artificial intelligence (AI) part, nexoya was invited and was active with a keynote speaker and a workshop.

Marco Hochstrasser at ZHAW workshop

In the subsequent in-depth workshop "Artificial Intelligence in Marketing" by nexoya, four main questions were addressed:

1. What is the development status of text content creation in the field of artificial intelligence?

2. How can images be created with artificial intelligence?

3. Where can artificial intelligence support campaign optimization in the area of data analysis?

4. How far is the area of predictive analytics with artificial intelligence in marketing?



In our next three blogs, we will dive deeper into the workshop and document the different parts in detail. This is the first part of the Workshop and it will focus on Content creation.

Workshop part 1: AI in practice test and Text generation

In the first part, the participants had half an hour to apply state-of-the-art AI applications. The aim was not so much a detailed tool analysis as an initial "experience test" in as little time as possible. Five different groups with around six participants formed and each group was given a specific topic for use as an application example or for marketing.
The objective was to create a landing page based on tools that produce the content with artificial intelligence. It was clear there was not enough time to achieve professional results, but that's not what the exercise was about. We wanted to show within a very short amount of time how far these tools have come development-wise, what can be achieved with them, and where their limits lie.
Each team had a different use case and therefore sold a different product. The goal was to create a website, promote it, and then analyze it and make predictions.


1. Bathing ducks

Bathing ducks

4. Onions

Onions

2. Cat food

Cat food

5. Santa Claus Hats

Santa claus hats

3. Coffee

Coffee

Steps in detail:

1. Landing page as basis with WIX (without AI)

To make things a little easier, we prepared almost empty "WIX - Landing Pages" for each team. Each team then had to finish their landing page with completely AI-generated content. Text generation based on AI.


2. Text generation

The current best AI text generation models are trained mostly in English. Therefore, we focused on the production of English pages. For our AI text generation tools, we selected the following:

Screenshot from Talktotransformer tool
Talk to Transformer is one of the pioneers in the field of content creation. Created by a machine learning engineer Adam King this is an interesting tool to try out.

Screenshot from articoolo tool
Articoolo is a content generation tool that first analyses the text and then creates the content. The tool uses NLP and AI methodologies while creating texts.

Screenshot from AI-writer tool

AI-writer is another cool tool that offers text generation. Moreover, AI-writer offers rewriting services as well. This can be useful if you want to rewrite your existing articles and spin around sentences.



Field Report

As already mentioned, it was not possible to create perfect texts in a short time. All three tools helped to find ideas for text creation, but usable texts could only be generated to a limited extent. The autonomous use of text-generating tools with the artificial intelligence of completely unstructured content still has massive potential for improvement. Nevertheless, the results were impressive.

The second part of the AI in marketing Workshop report will be published in a week. In the next part, we will talk about Image creation with AI and mention a few useful tools. Stay tuned!



What does it take to become a data-driven Marketer?

Subscribe here to our newsletter and get the latest news and tips around the topic :


Share