6 MINS READ
We stand at the precipice of a productivity revolution in the workplace with next-generation AI systems.
These systems can facilitate language translation, tackle complex queries, analyze data, provide personalized recommendations, create brand-aligned user interfaces, and more. They do this by leveraging highly personalized content across channels and customer touchpoints.
Foundation generative AI (GenAI) models like ChatGPT and Midjourney are composable or built into an “enterprise AI platform” that can address many workstreams across enterprise functions. These generative models present a world of possibilities—they can be a creative assistant to humans, providing inspiration, ideas, drafts, and conversational responses. They can also fully automate complex work streams through integration with enterprise and web data sources. While they are already making a huge impact across industries, in this paper we will delve into how generative AI can transform marketing for enterprises and how proven innovation frameworks like the TCS Clay Map can accelerate that journey.
TCS Clay Map
Leverage TCS Clay Map to prioritize innovation and accelerate impact with generative AI.
The TCS Clay Map is an innovation portfolio framework that allows an enterprise to map innovation ideas based on their ability to generate additional revenues through new customers and markets, and their requirement for new capabilities. Our Clay Map for generative AI-led innovation groups ideas by combinations of associated technology risk and business risk across four quadrants, as shown in the illustration below.
Quadrant 1 typically comprises the here-and-now or ‘derivative’ innovations that allow a business to run more efficiently and effectively. These are ideas that utilize current capabilities for current markets and strengthen the core business.
This quadrant involves jobs related to generative AI-conducted research (brand, market, and customer), assisting content developers with AI-generated new marketing and brand outputs, and enhancing customer support through autonomous chatbots. It includes providing hyper-personalized recommendations to customers based on their preferences, past purchases, and browsing behavior.
Generative AI innovation or extension for jobs in Quadrant 1 will need to be assessed from an overall cost-benefit proposition, impact on society, and other concerns. The cost of training, operating cloud models, and operations should not outweigh the cost of people doing these jobs currently.
Quadrant 2 looks at leveraging new capabilities and business models to transform the current business. It utilizes innovative emerging technologies to address existing challenges. Examples of jobs in this quadrant includes developing a platform, enabling AI-generated content aligned with brand guidelines and accessible across the enterprise, streamlining ad creation and optimization, maximizing the effectiveness of advertising campaigns, and creating immersive brand experiences.
Assisting creative designs with AI-generated ideas and variations for marketers is another possibility. Tommy Hilfiger is already doing it—it leveraged GenAI to engage customers in co-creation during the Metaverse Fashion Week, giving them the opportunity to design items in the brand’s signature preppy style.
Quadrant 3 is about extending current capabilities to new customers and new market segments. It could also mean partnering with new ecosystems and technologies to expand current capabilities to new segments, making them more widely accessible.
Organizations that have a niche skill in any area of marketing or marketing tech could consider monetizing by enabling AISkills/AI-Plugins to AI models, thereby creating a new revenue stream. They can also opt for the creation of digital marketing assets at scale and monetization through licensing or creating self-service platforms. For example, Spectrum Reach announced a first-of-its-kind AI-powered platform that helps businesses create high-quality TV commercials, that too with AI-generated voiceover, in five minutes.
Quadrant 4 is about “new capabilities, new markets”, ‘Blue Sky’, and disruptive innovation. Ideas in this quadrant are likely to significantly transform the entire industry, even the ecosystem. It may utilize one or more transformational technology, to develop a completely unmatched offering, or it may assemble a completely new capability from existing and known technologies. Quadrant 4 innovations tend to be longer-term and futuristic, with a potential for bringing in a significant and lasting competitive edge to the business.
The possibilities in quadrant 4 will involve collaboration between AI brand strategists, key marketers, AI-generated products, and AI-generated influencers. They will also be about the creation of highly realistic and immersive virtual reality experiences that engage all senses. Users could explore virtual worlds indistinguishable from reality, interacting with virtual products, objects, and characters in unprecedented ways, which also offer marketers new ways to engage and generate user insights. BMW’s 'The Ultimate AI Masterpiece' that leveraged AI and machine learning to turn a car into digital artwork is one example. Having analyzed and learnt from around 50,000 paintings encompassing the art history of 900 years, BMW’s AI model generated entirely new works of art on the surface of a BMW 8 Series Gran Coupe.
While the above Clay Map is a representative set of possibilities, companies can apply this to meet the unique needs of their industry.
Mitigate generative AI risks with human oversight.
Adopting generative AI comes with its own set of challenges and companies looking to ride the wave of this new technology needs to address them. Here are a few things they need to focus on:
Training: GenAI models require extensive training data to produce accurate results. Obtaining such data can be time-consuming.
Data security: Models often handle sensitive data, including personal or proprietary information. Malicious actors can exploit and manipulate GenAI, leading to cyberattacks.
Bias: Models can exhibit bias introduced through training data, model architecture, or usage.
Privacy: Privacy concerns may arise if user input becomes identifiable in model outputs.
Hallucination: In what is referred as ‘hallucination’, models can produce factually incorrect content confidently.
Misinformation: Can be misused to create fake, inconsistent content and damage reputations.
Inappropriate material: Offensive content can harm reputation and lead to costly consequences, including defamation lawsuits.
Ownership issue: Ownership disputes over training data or generated content can arise, raising concerns about intellectual property rights and copyright infringement.
With proper human supervision, a set of programmable constraints makes apps respond with accurate information and establish connections with third-party applications safely. To mitigate risks, initial marketing experiments should prioritize public or minimally sensitive internal data. Personally identifiable information (PII) must be protected. Human oversight must be maintained on outputs at all times.
The way forward for marketers
The future is humans and AI working together.
Agencies or marketing service providers looking to leverage GenAI need to first conduct an assessment of the potential impact on the fundamental business model, including short and long-term implications. Using TCS Clay Map framework to map Quadrant 1 to 4 activities for the entire marketing function to sub-functionals will help them easily assess the implications.
While assessing each job, it is important to:
The future of work in marketing will be symbiotic, with humans and AI working together. Companies need to prioritize and reimagine marketing workstreams that have maximum value (like augmenting scarce skills) or automate tasks that are resource-intensive and take long to complete. In short, the goal should be to enhance human capabilities while embracing the new capabilities.
Establishing cross-functional teams comprising members with expertise in marketing, legal, AI, design (reimagination), business go-to-market, ethics, and rapid prototyping is important to make the most of AI, ethically. It is crucial to consistently evaluate the balance between risk and value creation, while addressing AI safety concerns such as bias, privacy, data ownership, creative rights, and liability.
A mindset shift
Generative AI is here to stay and marketers should embrace it with an open mind.
The generative AI value chain is developing and evolving. Numerous foundational models, application layer innovations, and infrastructure for extending the value of MLOps/AI ops are consistently emerging. Facilitate a sandbox environment for internal teams to conduct experiments and move towards assembling a marketing AI platform. By approaching generative AI with an open mindset and adopting the tools it offers, marketers can remain ahead of the curve.