Generative AI tools have advanced tremendously in recent years, moving beyond novelties to becoming indispensable tools for companies to leverage. The ability of AI to generate images and videos, not just text, has caught the attention of many retailers, with brands such as Levi’s, Dove and Nike already capitalizing on AI to create marketing and advertising content. These tools allow for quick iterations and refinement, resulting in time and monetary savings by lessening the need for traditional models and photography crews.
The benefits of generative AI in retail advertising are not without their drawbacks, however. Concern has been expressed by some consumers around generative AI’s potential to perpetuate unrealistic beauty standards, harming the mental health of teens and young adults. There is additional concern that AI could erode the progress made through inclusivity in the beauty and fashion industries.
Despite the potential risks, AI is still a powerful tool with significant potential. Companies who seek to integrate AI into their retail operations should define guardrails and polices to govern its use, protecting their customers and the perception of their brand.
Generative AI in Retail Advertising
Retail companies who use generative AI to help create their ads stand to gain an advantage by saving time and money over traditional production methods. AI allows for quick iterations, on the fly changes, and refinements that would normally take a considerable amount of time, and it can lessen the need for costly and labor-intensive photoshoots and productions by generating content from provided images.
The Diigitals, an AI-driven modeling agency, has capitalized on this technology to great effect. They recently started partnering with human models to create 3D avatars from 40-60 still images. They then license the model’s likeness to retail and fashion companies for use in their ads and marketing materials. Retailers can digitally manipulate and pose the model’s 3D avatar to fit their needs and model their clothing. By selling these 3D avatars, The Diigitals aims to help human models monetize their likeness and seeks to encourage the fair use of AI in retail and fashion by integrating AI and human models together.
Similarly, Nike created an ad to promote their brand featuring Serena Williams. In it, they fed an AI model thousands of videos clips of her matches to analyze and create a singular AI-generated video of the tennis star playing against her 16-year-old self.
Potential Drawbacks of Generative AI
Although generative AI presents numerous opportunities for retailers, they should also be aware of the drawbacks and potential risks this technology may pose. Despite firms like The Diigitals seeing success in their use of AI, there has still been backlash from the public on the use of AI models.
After Levi’s announced their partnership with Lalaland.ai and their intention to use AI models to promote inclusivity and diversity, many models and members of the public pushed back. They expressed concerns that the use of AI would be detrimental to the opportunities of real models.
In contrast, Dove began a media campaign highlighting their commitment to using live models to avoid the perpetuation of unrealistic beauty standards in their ads. This ties into a growing concern regarding the potential for AI to perpetuate unrealistic beauty standards and, in turn, harm the mental health of teens. This concern is founded from the issues seen through altered and unrealistic images on social media. In fact, earlier this year, over 12,000 parents signed a petition to compel TikTok to label and police AI-generated influencers who promote unattainable beauty standards for children and adolescents. Despite the tremendous cost savings AI could bring, the downstream effects must be considered to ultimately protect customers and brand image.
Considerations for Retail Companies When Adopting Generative AI
Retail companies looking to capitalize on recent tech advancements have the opportunity to learn from the successes and growing pains of their peers. In parallel with adopting AI, companies should establish standards for its use and implement guardrails to ensure AI-generated content aligns with their values and brand image. Here are a few considerations:
Set Standards for Its Use and How It Will Be Enforced
Before generative AI tools are adopted, companies should decide the types of content that the tools will be creating, how the models will be portrayed, and the brand goals (such as DEI) that must be upheld through its use. Creating an internal group to set AI standards and ensure those standards are met at each stage of development would be a prudent first step. AI-generated content is a direct result of the data it’s trained on, so setting standards early and establishing guidelines for components like training data or prompts is key to ensuring AI content aligns with company goals.
Build an Internal AI Development Team
AI image generators from companies such as Google, Amazon, Black Forest Labs, and Stability AI can all be licensed to further build upon by a dedicated AI development team. The team can then streamline the model for their specific use cases and even train it on a narrower set of proprietary data. Having an internal AI development team affords the greatest control to companies over their AI models to ensure the proper standards are in place and being met.
Govern the Use of AI At Each Stage
There are three major touchpoints that inform the output of a generative AI model: training data, fine-tuning, and prompt engineering. AI imitates the training data it is fed, so as retail and fashion companies leverage leading AI image generators, it’s important to curate additional training materials in the form of image libraries that reflect factors such as diversity, inclusivity, and other aspects they expect the final images to reflect. Development teams can further fine-tune the model by tweaking parameters based on initial results to move closer to the desired outcome. Finally, prompt engineering involves providing iteratively and carefully worded instructions for generative AI models to describe the image that is to be generated. Crafting these prompts specifically and with language that includes key attributes or descriptors will yield a result closest to the desired image. Internal groups tasked with governance of the AI model should be involved in this process as well, working closely with dev teams to craft prompts and vet the resulting images.
Final Thoughts
Generative AI has become indispensable for many company’s operations, and platforms like ChatGPT have made it more accessible than ever. The technology offers many benefits, including cost and time savings through rapid iteration and forgoing the need for previously required resources. AI is not without its drawbacks, however, including risks to public perception based on its use. Retailers looking to leverage generative AI should strongly consider what standards should be set and what guardrails would best govern its use. In weighing the benefits and considerations, companies must ultimately determine the best use cases for the technology and how it would fit into their business model.
For more information on the considerations and applications of generative AI in retail advertising, reach out to Clarkston’s experts.
“At Clarkston Consulting, we make our clients’ purpose our purpose. Since 1991, we have worked with global life sciences, consumer products, and retail companies to advance their most complex global priorities.”
Please visit the firm link to site