What is taking advantage of generative AI going to cost you?

Generative AI is transforming digital marketing by embedding intelligence into every part of the enterprise software ecosystem. Software companies worldwide are incorporating this technology into their platforms, enhancing customer engagement, personalization, and productivity. However, in a research article , Deloitte notes, whether these features will drive significant new revenue for the software vendors implementing them, remains uncertain, with 2024 seen as a "transition year" for generative AI's broader adoption and monetization in software. Is the cost of offering generative AI getting in the way of innovative and truly value addition in digital marketing tech? Here's my personal look at how generative AI is taking shape in digital marketing, with some predictions on where it’s headed next.

Scaling Personalization and Efficiency

Personalization has long been a goal for marketers, and generative AI brings it within reach, but at a cost. For example, each generative AI interaction can cost between $0.01 and $0.36, which means providing these personalized experiences at scale isn’t cheap. Despite these high operational costs, AI-driven tools allow marketers to tailor ads, emails, and content more effectively, aiming for “The segment of One” or hyper-personalization. Deloitte’s analysis suggests that even with the cost challenges, the largest enterprise software companies are embedding AI as a “must-have” feature rather than a premium addition.

This approach aligns with data showing that less than 20% of companies are willing to pay extra for AI features, viewing them instead as table stakes. This trend indicates that while AI can greatly improve customer experience, vendors may need to absorb costs initially or find other monetization strategies until AI’s ROI can be clearly demonstrated.

Customer Interaction and Real-Time Insights

AI-powered chatbots and digital assistants are transforming customer interaction by offering 24/7 support, product recommendations, and more. For instance, Optimizely One – the DXP from Optimizely has “Opal” as a “global AI assistant” that unify data from multiple sources within the entire DXP. This allows businesses to make faster, data-driven decisions and deliver high-impact, real-time interactions, which can be a true game-changer for digital marketers.

Optimizely provides Opal as a deeply baked in feature of the platform and using it does not drive any additional cost to Optimizely customers. Personally, this is the way I think it will need to be. Generative AI needs to become commodity, just like cost of sending requests to a database has become somewhat transparent years ago.?

However, as the cost of AI infrastructure – primarily serviced via big cloud vendors - remains high, Deloitte predicts that other software vendors (DXP’s included) will have hybrid pricing models for these features, blending a monthly per-user fee with a usage-based fee to help cover operational costs. Optimizely have been leading the race with early investments in gen AI across the platform, and the operating cost may very well have distracted others from doing the same. So, you could argue that this hybrid payment model may make AI more accessible without compromising profitability for these software companies, allowing more businesses to implement intelligent customer interactions across their platforms. ? From automated blog posts to personalized emails, generative AI empowers marketing teams to produce vast amounts of content faster. Deloitte projects that 100% of the top 50 software companies will offer AI-enhanced tools in 2024 and 2025, many with flexible pricing models that make these features accessible to a broad audience. The ability to generate content rapidly helps brands stay agile, responding to trends and customer needs in real time. However, this functionality comes with operational and strategic challenges— companies may need to prepare for usage-based pricing models and an overall increase in subscription costs to support the demand for these tools.

Customizable AI Models as the Future of Gen AI in Marketing

The next step in generative AI’s journey in digital marketing lies in companies creating and using their own large language models (LLMs). Here Optimizely One with Opal are leading the charge, allowing businesses to implement customized AI models that align with their unique brand identity and data. By deploying proprietary LLMs and combining them with Retrieval-Augmented Generation (RAG) technology, companies can ensure that every customer interaction is both personalized and cost-effective, maximizing the value AI brings to their digital marketing strategies. Optimizely’s approach allows companies to leverage data from various solutions across their stack, making the Opal AI assistant a more integrated and valuable resource.

The ability to use customizable AI models will become essential in digital marketing, and driving the innovation with this type of gen AI capabilities has moved Optimizely ahead in multiple analyst reports. The Opal model address key concerns around cost, scalability, and control over proprietary data. For companies, this trend toward private LLMs represents the way forward: a strategy that combines the power of generative AI with the flexibility to meet each company’s specific needs. This is certainly the reason we are now seeing other DXP vendors announcing very similar gen AI implementations. The question is what the cost will be to use them.

______________________________________________________________________________

Disclaimer: The information in this blog post is based on the author’s personal views and interpretations and may not represent the perspectives or official stances of the companies or brands mentioned. While every effort has been made to ensure accuracy, errors or omissions may still occur. Readers are encouraged to verify details independently and consider that any brand names or products mentioned are discussed from the author’s viewpoint, which may include subjective opinions.

______________________________________________________________________________