The GetResponse platform has been in the market for over 25 years. We have been implementing innovative solutions in online marketing since the early days, so using artificial intelligence in communication was a natural choice. One of the first AI-based solutions implemented at GetResponse exactly 9 years ago was to analyze the time when a message was opened by the recipient and optimize the sending of follow-up email messages to the recipient at the most appropriate moment. So as to significantly increase the chance of its opening.
Because we saw the power of analyzing large data sets through artificial intelligence, we created mechanisms to classify and evaluate content sent by not always honest customers. All this to eliminate the negative effects of their activity in the very early stages.
At GetResponse, we respond to our customers' needs by providing them with easy-to-use and affordable marketing tools for every industry. Thanks to the possibilities offered by AI, our solutions become even easier to use, more efficient and more affordable. Implementation of such tools is our priority.
What goals did you set before you decided to implement AI in GetResponse?
From the beginning, GetResponse's goal was to enable our clients to develop faster, primarily by automating activities related to online marketing. With the use of AI models, we were able to go a step further and offer users content creation support that, combined with affordable and perfectly functioning content distribution automation, saves them even more time and money.
In order to support our clients in their development, we must develop dynamically as an organization. That's why we implement innovative solutions in GetResponse from the very beginning to automate activities and achieve operational excellence in many areas.
Bulgaria WhatsApp number list
How long did it take you to implement AI in GetResponse and what challenges did you face?
The implementation of the first AI solutions took place a long time ago and was a long-term process, mainly due to exploring a completely unknown area and learning on the job. We didn't have the same competencies back then as we do now, when we have a whole team responsible for AI.
At the beginning of 2021, we released Website Builder, where one of the options for creating a website was to use an AI editor that, based on modest input data such as audience group and industry, designed the layout of the entire website in a few tens of seconds.
In the context of creating content for customers, after OpenAI made the ChatGPT API available in November 2022, we tested the first solution to generate open-optimized message titles after a few days, and after a few weeks we released a message builder that creates full emails for customers. We were one of the first in the industry to offer such a solution, which has already become our trademark.
In mid-2023, we started working on creating our own chatbot. It took us several weeks to train it with our knowledge base Database Sale and launch it for our clients in all languages supported by the CS department. As a result, customer satisfaction with chat is 96% and 89% of conversations start in less than 30 seconds.
One of the biggest challenges recently has been the differences in individual markets. In each, the legal requirements and perceptions of AI services differ slightly. Another challenge is the specificity of many languages, of course including Polish.