Using Low-risk, High-reward Enterprise GenAI for Marketing

There is a dizzying array of new generative AI solutions promising to optimize customer journeys and experiences. So it’s no surprise many digital marketers are ready to go all-in with the technology. In fact, 63% of marketing leaders plan to invest in Enterprise GenAI tools over the next two years. But taking a frenzied leap without homework could lead to missteps. Amid the high volume of new GenAI solutions, be cognizant that most are designed for efficiency and productivity gains — but not necessarily effectiveness. In short, they may save you time, but they’re not making the enterprise money, when they could.

A new breed of specialized enterprise GenAI technologies has begun to unlock billions in value across industries, and this “value” transcends solely gains in productivity. According to Coresight Research, in the retail sector alone, global revenue generated by AI will exceed $38 billion in 2030, up from an estimated $8.5 billion in 2023.  With so much to gain and so many options, how should marketers assess the landscape of technologies to identify which solutions will drive the most value for their business?

Making sense of GenAI for marketing

In marketing, GenAI is most commonly used for content creation, whether written or visual; Large Language Models (LLMs), a type of GenAI, synthesize massive amounts of data to generate text-based content. Some people use the technology to help develop the first draft of an email or blog post. GenAI is also being used to quickly generate an image to accompany written text, like to illustrate a social post. The key thing marketers need to consider, however, is that most GenAI models are designed to simply churn out content quickly — without any guarantee that the content is relevant, effective or even accurate. This can lead not only to missed opportunities but also to increased risks.

There are two broad types of LLMs used today: open models, which are available for anyone, and specialized models which are designed to meet a specific set of needs or goals, typically created by a business.  Open LLMs trained on the Wild West of the internet, such as ChatGPT, can pose risks to enterprises when used for external communications; they are contextually clueless, can “hallucinate,” and for longer-form copy, can become a copyright concern. On top of that, using such popular LLMs doesn’t necessarily give the brand a leg up. Anyone can sign up for ChatGPT or DALL-E 2, so when everyone is using the same technology for a wide variety of tasks — from large campaign ideas to social media copy — the competitive advantage disappears.

Moreover, these buzzy, publicly available GenAI models are already being used to generate vast amounts of “personalized” content, creating noise. This means your content can easily get lost or fall flat.

The other class of GenAI solutions that marketers should pay attention to if they want both quantity and quality are those that are grounded in proprietary models designed to achieve a specific business outcome. While there are different types of proprietary models, one version is trained on enterprise communications using a specialized and verified knowledge base that enables marketers to meet business goals and avoid some of the challenges of using open LLMs.

Using GenAI for content creation is not only a highly common marketing use case but an increasingly impactful one. According to the Gartner Use Case Prism: Generative AI for Marketing (October 5, 2023), which plots 20 of the most prominent use cases for generative AI in marketing against value and feasibility, the “Content Copilot” use case scores highest for both feasibility and business value.  The report states that the value of the content copilot use case is “productivity, variety, and velocity of content creation that improves customer engagement and response.” This explanation points to the importance of seeking out GenAI text solutions designed for productivity and performance.

Personalizing content where it counts

While it’s tempting to use GenAI to help marketing teams do more of the same tasks in less time, wouldn’t it be better to use it to achieve better results from those tasks? I firmly believe effectiveness is more likely to produce higher ROI. The key to generating not just more content, but better-performing content, is to understand the language most likely to motivate action. To do this, you need a specialized AI solution — with an LLM trained on the performance of real-world messages at scale — capable of unearthing consumers’ emotional drivers and motivators and generating language that isn’t only low-risk and relevant, but also high reward,  driving measured increases in customer engagement and conversions.

Retailers such as Tapestry, the parent company of Coach, Kate Spade, and Stuart Weitzman, have achieved notable business impact using emotion-informed GenAI (called Motivation AI). With Motivation AI, brands can personalize language throughout the marketing funnel — even on the online cart and checkout pages, lessening online cart abandonment rates, which have been a long-standing thorn in the side of online retailers. Rather than presenting generic, static language to customers, such as “add to cart,” or “getting your order,” retailers have begun using Motivation AI to personalize in real time the words and phrases that are most likely to resonate with individual customers and ultimately motivate them to complete their transaction.

For example, personalizing online cart messages based on behavior shown in web session data, with messages such as “you’re really close,” or “we love your taste.” The technology is reinforcing purchase decisions and helping brands achieve up to a 5% increase in sales at the critical cart stage.

GenAI’s biggest test — the 2023 holiday shopping season

With the holiday shopping season upon us, we’re about to see GenAI for marketing face its biggest test — and opportunity.  This is game time for customer acquisition and engagement, higher conversions, stronger customer relationships, and more loyalty.

The next few months will provide a glimpse into what we can expect for the future of GenAI and personalization, but we’re only at the beginning. Risk-averse marketing teams still have time to trial specialized GenAI to see (and seize) the potential of these easy-to-deploy, market-proven solutions. At the end of the day, generating greater revenue and stronger customer relationships must be central to the payoff from any external-facing tech investment. And getting there doesn’t need to be risky.