Event Report: "For Marketers to Master AI" at Marketing X Osaka 2023
On Wednesday, April 11th, 2023, Woojung Kim, our Solution Architect, spoke at "MARKETING X -23 Osaka" held at the Osaka International Convention Center and discussed the use of AI and ChatGPT in marketing, which is widely discussed today.
Variety of people from a wide range of companies, from companies that are actively using AI to many people who are new to AI, attended the event. This report provides a digest of the seminar's presentations on "AI in Marketing" and "Challenges and Solutions for AI in Marketing".
Use of AI in Marketing
We asked a marketer who attended the seminar to give us an actual example of an AI application.
1. "To whom" Market Search and Segmentation While market size and segments are still often defined based on marketers' experiences and ideas, companies that are increasingly using AI models are analyzing customer and user data to more efficiently and accurately visualize and analyze potential market segments and their trends. AI models can predict demand, reduce unnecessary costs, and identify customer segments that may not have known they existed.
2. “How” Marketers Communicate with Customers and Users Equally important as market segmentation is branding your company. AI enables marketers to identify more effective communication methods (channels, time frames, frequency, etc.) based on past ad and email viewing data.
3. "What" Content Production After determining the potential market and how the company wants to brand itself, the marketer needs to decide what to deliver. There are mainly ad banners, videos, posters, copywriting, emails, campaigns, etc., all of which can be produced using an AI model; AI can be used to produce the most appropriate productions for each target segment based on customer data analysis and detailed customer segments.
The issues raised by the participants during the presentation are challenges that can be found in the use of AI in many areas, not only in marketing. While some aspects of the solutions vary from one situation to another, there are many common aspects as well. This article will describe some of the examples for your reference.
Challenges and Solutions for Using AI in Marketing
1. Data Availability Many companies want to utilize the data but do not have it in the first place, or have data but are not organized enough to use it. The only way to solve this problem is to do it. As the saying goes, "Garbage In, Garbage Out," if you do not collect data that is meaningful from a business perspective and improve the quality of input data, the quality of output will also be low. This is an essential step for AI implementation.
2. Data Accuracy The most frequently cited issue was "We have started using AI, but the accuracy of the output data is very low.” To solve this issue, there are several things to consider, such as reviewing the introduced AI model itself and reviewing the operational environment (e.g., internal organization). If tuning of the input data, the model, or other maintenance is not performed after implementation, various problems can occur, such as a drop in accuracy (proximity between output data and true values) despite little variability (dispersion of output data), or a higher average accuracy of outputs but greater variability.
3. Culture of AI use It is often said that they have implemented AI, but it is not used in the field in the first place. The two main reasons we hear are "we don't know how to use it" and "we don't trust the data. The key to both is to develop an internal culture, but a more concrete solution that can be implemented immediately is the application of Business Intelligence (BI). BI is basically a technology wrapped by UX. Therefore, BI can make the onboarding of AI models and data more smooth, and visualize the effects of measures using AI model output data, thereby improving the confidence in AI and promoting its use in the field.
We hope it will be helpful to you.