Generative AI in Fashion Advertising: What Fashion Brands Should Prepare For

In May, Meta, the parent company of Facebook and Instagram, made an announcement regarding the introduction of the AI Sandbox in July. This sandbox serves as a platform for testing new tools and features, including ad tools powered by generative AI. Among the initial offerings is the capability to create background images using text inputs, similar to text-to-image tools like DALL-E Open Ai and Midjourney . Additionally, the AI Sandbox now includes image outcropping, which automatically generates various aspect ratios from the same content to suit different formats such as Stories and Reels. Another tool provided is a text generation feature that produces multiple variations of text, allowing advertisers to customize their messages for specific target audiences. These developments have the potential to revolutionize how fashion marketers operate and may also have an impact on how we perceive advertisements.

Shortly after Meta's announcement, 谷歌 also unveiled its own set of similar tools, which will become available in the United States at the end of July. These tools, collectively known as Product Studio, offer various functionalities such as generating lifestyle imagery from product shots, removing backgrounds from existing product images, and enhancing the resolution of small images. In addition, Salesforce , a software company, recently introduced MarketingGPT and CommerceGPT (abbreviated from "generative pre-trained transformer"), which will provide capabilities like generating personalized emails (starting in October), creating contextual visual assets for multi-channel campaigns, and generating tailored product descriptions for buyers (starting next month). Furthermore, Shopify , an e-commerce firm, is implementing ChatGPT-generated product descriptions, while clients of Amazon Web Services now have the ability to construct their own generative AI bots. These advancements signify a significant shift in the advertising and e-commerce landscape, as industry players adopt generative AI technologies to streamline and enhance their marketing strategies.

In contrast to previous AI applications in text and video that have been showcased as online experiments, often with noticeable flaws such as distorted faces, inaccurate content, or repetitive clichés, the aim of these new tools is for customers to be unaware of the involvement of generative AI in the creative process. This initial introduction sheds light on why some individuals are both optimistic about and apprehensive of the potential for generative AI to disrupt marketing roles, serving as a mere starting point.

These new tools introduce a faster and more efficient way to generate a substantial amount of marketing materials that can be utilized across multiple platforms and purposes. This has the potential to greatly impact the advertising ecosystem, e-commerce, and personalization by transforming how brands test different assets and enabling them to create a greater variety of content personalized to various customer segments.

Furthermore, these tools facilitate the testing of personalized emails and web copy on a scale and speed that was previously unattainable. Salesforce's CommerceGPT tools leverage what they refer to as "harmonized first-party data," which combines data from diverse sources owned directly by the brand and from the customers themselves. This empowers retailers to automatically generate product descriptions tailored to each individual buyer. For instance, if the system is aware that a customer recently purchased a specific handbag and is planning a summer vacation, it may include text that suggests complementary items or pairings.

Samantha Porter McCandless , the chief merchandising officer at luxury resale marketplace The RealReal , emphasizes the increasing expectation for personalization among the majority of people. Consequently, tools that facilitate scaling and improving personalization are poised to become crucial for numerous brands. The RealReal, in collaboration with Salesforce , has already been utilizing AI to generate e-commerce product descriptions, recognizing the value of such advancements.

A survey conducted in May, commissioned by Salesforce and targeting marketers in the United States, United Kingdom, and Australia, revealed that over half of the marketers are already experimenting with generative AI in their work. Furthermore, 71 percent of them anticipate that generative AI will enhance their productivity, estimating time savings of approximately five hours per week. Ranjan Roy , the VP of strategy at lingerie brand Adore Me , estimates that an internal tool used for generating product descriptions has saved employees around 30 hours per month in copywriting tasks.

The demand for generative AI products is projected to generate roughly $280 billion in revenue for software tools, including specialized assistants, infrastructure products, and copilots that expedite coding processes, as stated in a June report by Bloomberg Intelligence. Big tech companies are expected to be among the primary beneficiaries of this growth. Digital ads, fueled by generative AI technology, are predicted to be the second-largest revenue driver, accounting for approximately $192 billion. Overall, the generative AI market is estimated to reach a value of $1.3 trillion in the next decade, a significant increase from $40 billion last year, according to the report.

The advent of generative AI, like any new technology, raises legal and ethical considerations that have yet to be fully defined. Common concerns revolve around the potential for computers to copy the work of others, introduce errors in materials used by brands, or replace human workers on a large scale. In fact, in May, tech leaders even cautioned about a worst-case scenario that could result in threats comparable to pandemics or nuclear wars.

However, these messages of caution have not significantly dampened the enthusiasm surrounding generative AI in the short term. Brands are enticed by the allure of enhanced marketing efficiency, while tech companies see the potential for increased value in their platforms and tools. Adore Me, for instance, has already been utilizing and developing its own generative AI tools internally. However, integrating these tools into a core platform like Meta would offer greater convenience and efficiency, according to Ranjan Roy, VP of strategy at Adore Me. It would enable the brand to swiftly create and launch multiple targeted advertisements, even up to 20 at a time, catering to specific demographics.

Exploring Creative Iterations in Fashion and Beauty Brands

In the context of fashion brands, the application of AI tools holds particular significance, according to Rachel Tipograph , the founder and CEO of the social commerce analytics platform MikMak . In the past, creating a few visually appealing lifestyle images and a single compelling copy was considered sufficient. However, the landscape has evolved, and brands now require a multitude of options, especially with the increasing diversity of ad formats and the growing specificity of customer segmentation.

The shift toward generative AI tools enables brands to cater to the demand for a higher volume and wider range of creative options to effectively engage their target audiences.

Nicola Mendelsohn CBE , from Meta, emphasizes the early-stage benefits of such tools. Brands can gain insights into the potential appearance of an advertisement before investing in a full-scale photo shoot. By expediting the brainstorming process, advertisers can expedite decision-making and approve ads that have the greatest impact.

Rachel Tipograph, CEO of Mikmak, underscores the opportunity for generating numerous asset variations by combining different texts and images. This enables brands to explore a wide range of possibilities and determine which combinations are most effective. Once successful approaches are identified, ad spend can be reallocated accordingly, optimizing marketing efforts.

Balancing Personalisation and Privacy in E-commerce

For a long time, marketers have grappled with finding the right balance between personalization and privacy. The same automated processes that enable the creation, deployment, and testing of ads can also pave the way for a more individualized experience on e-commerce platforms.

Kelly Thacker, the SVP of product marketing at Salesforce and CMO of retail and consumer goods, believes that generative AI will bring online shopping closer than ever to replicating the enjoyable aspects of the physical store experience. This advancement is expected to provide fashion marketers with increased opportunities to be creative, establish personal connections with customers, and instill greater confidence in their purchase decisions. The potential for generative AI to deliver a more tailored and personalized shopping experience holds the promise of enhancing customer satisfaction and engagement.

According to Samantha McCandless, Chief Merchandising Officer at The RealReal, current personalization in e-commerce follows a "cohorted" approach. This means that a website might recognize that a customer has a preference for expensive handbags and prioritize showing those products while allowing the customer to further refine their search through filters. However, there is potential for a more unique approach that leverages customer data to generate highly specific imagery, recommendations, descriptions, and web copy.

Solutions as GardeRobo A.I. 's Cyber Stylist instantly understands end-customers shopping preferences and provides personalized product recommendations in real-time in aesthetic carousels or collages. GardeRobo also provides content managers with the ability to customize recommendations according to their marketing goals. The automatically generated content can be utilized in email newsletters and social media platforms. GardeRobo marketing platform is easy to set up and allows for any changes, such as adding a non-sellable product to a Total Look. Book your demo and learn more about the solution.

McCandless acknowledges that the current limitation lies in the availability of design resources to generate the extensive amount of personalized content desired. To achieve a higher level of personalization, more conversations around privacy will naturally arise. Brands and consumers will need to navigate and establish a comfortable level of personalization that respects privacy boundaries.

The potential shift towards a more individualized and tailored e-commerce experience raises questions about the balance between personalization and privacy, emphasizing the need for thoughtful consideration and open dialogue between brands and consumers.

The Complexity Behind Automation

While the immediate benefits of using generative AI tools for fast image editing and copy creation are apparent, there are various ethical, legal, and long-term considerations that need to be considered. One concern is the potential for unintended consequences when a tool is instructed to create content in the style of a specific brand, which could lead to unusual outcomes. Another issue arises in copyright infringement, where a brand might unknowingly publish copyrighted material. These challenges reflect the ongoing debate surrounding the responsibility for content on tech platforms, which has been an important topic for big tech companies.

To address some of these concerns, Google has implemented safety measures such as offensive keyword filters, and AI-generated images created through Product Studio undergo existing content reviews for products to meet content standards. Google is also developing models to include watermarking and metadata on AI-generated images, providing additional context and preserving information even after modest edits.

At present, brands and their creative teams are primarily responsible for any potential missteps or legal issues arising from the use of generative AI tools. It is expected that tech companies, like other industries, may attempt to avoid taking full responsibility in these cases. Consequently, some larger companies remain hesitant to fully embrace this technology. However, it is anticipated that the coming years will bring about litigation and subsequent regulations in this area. Advertisers must be accountable and ensure that the creative content produced is original.

Maintaining originality extends beyond visuals to the conceptual level. As brands gain easier access to optimized content, they will need to intentionally customize and personalize it to maintain their distinctiveness.

Despite the advancements in generative AI tools, there is a consensus that they will not completely replace creative jobs in the near future. Rather, they are seen as augmenting human capabilities, enabling marketers to become more efficient, intelligent, and faster in their work. The human touch remains crucial, and professionals in creative roles should embrace these tools as a means to enhance their skills and focus on tasks that require human expertise. While generative AI may pose a threat to creatives, the key is to leverage the power of software to create the next generation of jobs and find ways to enhance one's professional abilities through their adoption.

#AI #generativeai #aifashion #fashion #ecommerce #ecommercemarketing #chatgpt