Digital Innovation Digital Publishing Top Stories
4 mins read

Generative AI tools can fetch up to 5,700% ROI: Key insights for publishers from INMA

Getting your Trinity Audio player ready...

New INMA report takes a deep dive into Generative AI technology and the latest tools being used. The report shows via research and experiments how publishers can use these tools to automate news subscription marketing.

“Generative AI is a new technology that has the potential to revolutionize the way we approach marketing, particularly in the realm of news subscriptions,” according to a new INMA report. It finds that an average publisher can expect up to 5,700% return on investment if using generative AI to personalize its marketing. Additionally, on average, such a tool needs to save a marketer as little as 48 minutes per working week to pay for itself. 

We are now only scratching the surface of what generative AI can do for organizations and the people within them. It may soon be standard practice, for example, for such systems to craft most or all of our written- or image-based content.

Thomas Davenport, Professor of Information Technology and Management, Babson College

The report, “AI Guide and ChatGPT Promptbook for News Marketers,” is authored by Greg Piechota, Readers First Initiative Lead, INMA. It looks into how publishers can use generative AI tools for marketing news subscriptions to save time, improve scalability, and increase the effectiveness of reader engagement.

They can be used to assist with a host of marketing tasks such as executing nurture campaigns, creating compelling advertising, designing and optimizing websites, and more, according to the report. It also takes readers through the step-by-step process of using an AI tool for:

  • Researching a new product,
  • Optimizing subscription landing page and advertisements, and
  • Creating an email onboarding campaign

“As simple as texting your friends”

Generative AI creates new content based on a set of inputs and algorithms. The tools available now are easy to use. All one has to do is register or subscribe to an AI tool such as ChatGPT, Jasper AI, Writesonic and chat with it. It can be “as simple as texting your friends on a mobile device.” Some of these tools are free, others require a subscription. 

Here’s an example from the report which shows how AI can be used to research a product. ChatGPT was given this prompt: “Act as … a product manager for the Financial Times. Plan a research project aimed at segmenting audiences based on their preference for breaking and in-depth news as well as willingness to pay for online news.” 

It returned with the following plan of action.

The researchers followed this up with a prompt to draft the research tools it proposed, like a survey on reading habits or a scenario for in-depth interviews or focus groups. The tool responded with a 12-question survey. The prompt indicated some questions that the tool included along with suggesting completely new ones.

In another experiment, the bot was prompted to draft a welcome email for the Italian newspaper Corriere della Sera. Here’s its response in comparison with a human-drafted email.

The bot can also create versions of an email tailored to different reader segments making it more relevant for each. This considerably reduces the workload of people working on such projects freeing them to do more important tasks.

“The quality and relevance of generative AI’s outputs depend on writing effective prompts,” suggests Piechota. “The demand for professionals with this skill is likely to increase.”

He recommends the following best practices for using generative AI tools effectively:

  • Write clear and specific prompts. 
  • Provide context and be explicit about goals. 
  • Provide examples for the AI to learn. 
  • Refer to established marketing theories or frameworks. 
  • Think in terms of line of inquiry and not single prompts. 
  • Role-play with the AI. 
  • Create versions of outputs for different audiences. 
  • Carefully review outputs. 
  • Provide and ask for feedback to revise the outputs. 
  • Add human touch to the generated content.

“Advanced and productized enough to be used even today”

“The landscape of generative AI tools is rapidly developing,” writes Piechota. “With advancements in machine learning and natural language processing (NLP), new tools are constantly being released and existing ones are being updated and improved.”

INMA reviewed the top 15 generative AI tools for marketing copy. Here is its ranking:

And here is its ranking of the top copywriting tools. Only three tools were chosen for this—ChatGPT, Jasper AI, and Writesonic—based on the size of capital raised by the companies and the average customer satisfaction score in G2, a software review database.

“All three tools could successfully assist marketers of news subscriptions,” writes Piechota. “They are advanced and productized enough to be used even today. Any upcoming improvements, such as a new GPT-4 model trained on more data, will just make them better.”

“There is a learning curve to using the tools,” he adds. “It is important to spend time with the tools and put effort into understanding and learning how they work. The ride becomes easier over time. In general, we would recommend trying a number of tools in practical scenarios to see how they perform and which fits you best.”

The report recommends beginning with hiring AI as an assistant, promoting it to a personal coach, and eventually accepting it as a teammate to build trust and understanding between the human and machine teammates.

Writing with AI assistants is like riding a bicycle. Everybody can learn it but only some become professional cyclists.

Greg Piechota, Readers First Initiative Lead, INMA

The full report is available at INMA:
AI Guide and ChatGPT Promptbook for News Marketers