For CRM, AI Advances More Than Just Contact Center, Marketing Goals

Artificial intelligence is the new linchpin driving innovation that enhances customer relationship management and e-commerce. Once a concept for science fiction, AI is now a reality transforming sales, marketing, and many other business functions.

In CRM, AI revolutionizes how businesses store and analyze contact information, track customers through the sales process, and automate critical sales and marketing functions, streamlining workflows.

Generative AI, leveraging natural language processing (NLP) and large language models (LLMs), can comprehend and generate human language text, making it a significant force in technology and business this year. The insights generated by generative AI optimize business processes and are accessible to office workers across various roles and experience levels for both professional and personal use.

Marketers, advertisers, and customer experience (CX) specialists use generative AI to enhance campaign effectiveness, improve hyper-personalization, and build customer trust and loyalty. It allows them to process and understand large volumes of text and apply sentiment analysis to gain insights that enhance customer success and boost CRM effectiveness.

AI also helps marketers, advertisers, and call center agents gain insights into consumers’ brand sentiment and perception. It determines their likes and dislikes, according to SAS Head of Martech Solutions Marketing Jonathan Moran.

“These insights help them adjust strategies and improve customer communications, service, and support,” he told CRM Buyer.

Pros and Cons of Conversational AI

One of the top marketing trends is AI-powered conversational marketing. Brands and companies use this strategy to communicate with customers through personalized, real-time interactions. These exchanges often occur through a combination of web, email, chat, messaging, and social media channels.

“Conversational AI delivers cost, efficiency, and productivity advantages to marketers,” noted Moran.

Automating consumer responses across digital channels lowers costs and saves money from a resource and staffing perspective. Combining conversational AI with other process automation technologies — such as robotic process automation (RPA) — can increase efficiency and productivity by scaling concurrent customer service and support interactions, he continued.

Moran acknowledged that the main disadvantages of conversational AI are customer misunderstandings and the resulting frustration. As automated technologies are used to converse and interact with customers, certain dialects, syntax, sentiments, and emotions cannot always be understood by those technologies.

“This can result in a subpar customer experience but can lead to customer frustration, attrition, and churn,” he admitted.

Gaining Better Understanding and Acceptance of AI-Assisted Agents

Misinformation and misconceptions about AI are common among workers and consumers. We asked Jonathan Moran to share his expertise on what the new technology provides and what is needed to keep its use controlled.

CRM Buyer: How can conversational marketing be useful with a solid purpose?

Jonathan Moran: Conversational marketing is most effective if an organization or brand has strong data management and a scalable and flexible analytics platform. The former ensures the availability of relevant, high-quality data, while the latter helps transform that data into better decisions.

How does AI marketing differ from digital marketing?

Moran: I think conversational marketing, ML-based marketing, or algorithmic marketing [all AI-based marketing] are just components of broader digital marketing strategies. As digital marketers mature their practices and processes, infusing AI technologies and capabilities into those processes becomes a natural next step in their digital marketing maturity.

Can AI be integrated into existing digital marketing without eventually eliminating human marketers?

Moran: My colleagues and I recommend following the concept of human-in-the-loop (HITL). While people may use technologies like generative AI to get a head start, human oversight is still required.

AI helps marketers be more effective and efficient. With it, organizations can optimize processes and enhance competitiveness.

So, you do not see these AI-powered advancements as a threat to human workers’ jobs?

Moran: Just as automation in manufacturing led to new roles, AI will spark new jobs and new careers. Of course, as AI evolves, marketers and advertisers will too. Adding AI skills to your portfolio will grow in importance. Those who integrate AI into their roles will have an advantage over those who do not.

Ultimately, AI will not replace people. Rather, AI helps people to do more.

Do you think that AI training will ever surpass human capacity for creativity?

Moran: AI, for all its promise, has limits. It cannot feel nor express emotion as a human can. It cannot innovate, invent, or add new knowledge. ChatGPT, for example, relies on or summarizes existing information.

I believe AI will never surpass the capacity of human creativity. But it will be a powerful and important tool in augmenting and supporting human creativity.

What ethical problems are inherent in using AI in marketing?

Moran: The main ethical issues in marketing relate to the bias AI machines can impose. AI is making decisions based on data and algorithms, not human intuition. Certain situations arise whereby a decision is made in an automated fashion that can offend based on age, gender, race, cognitive ability, or some other qualifying factor.

Years ago, marketers and, to some degree, content providers used subliminal advertising techniques. Is using AI for predictive analytics a similar subterfuge?

Moran: I don’t think AI is intended to deceive customers in any way. The majority of consumers are aware that AI technologies are being infused into digital marketing practices.

AI should be — and is — mainly used by brands to automate processes for both efficiency and effectiveness. The result is typically a more satisfied consumer because they get their service and support concerns addressed quickly.

AI-related technologies like natural language processing — a technique founded in sentiment and text analytics — help marketers and advertisers gain insights about consumers’ brand sentiment and perception, as well as likes and dislikes. These insights help them adjust strategies and improve customer communications, service, and support.

What are the challenges in integrating AI into the creative aspects of content production?

Moran: The main challenge is authenticity. While there is no doubt that AI-based content creation increases the efficiency and productivity of marketing and advertising departments, AI techniques such as generative AI can create images and content that include plagiarism, unintentional bias, and content not meant to be repurposed.

Often lacking the genuine emotion and unique perspectives of human-created content, AI-created content can appear inauthentic to both the brand and the consumer.

One way to avoid this inauthenticity is to keep humans in the loop, particularly creative professionals. With human oversight, errors can be minimized, and AI can fulfill its great promise to help marketers and advertisers hyper-personalize messages and offers, build brand loyalty, and enhance campaign effectiveness.

What is the biggest threat AI imposes on e-commerce?

Moran: The biggest threat today is fraud supported by AI technologies. Brands must ensure that cybercriminals are not using AI to create fraudulent customer profiles that steal from brands via e-commerce platforms.

Additionally, if brands use generative AI to create content for e-commerce sites, they must ensure approvals and guardrails are in place before that content [images and descriptions] goes live.

How can AI developers ensure ethical results from AI?

Moran: Three key considerations can help to keep AI’s conduct in check: quality data, oversight, and flexibility.

AI models are trained using data. Insufficient data or no appropriate learning reinforcement can lead to inaccurate or unethical decisions.

Organizations need a system of governance with clear owners and stakeholders for all AI projects. Then define which decisions they will automate without human input. AI models should be monitored and audited regularly to identify bias creep and unexpected operations.

To enforce policies, people must be able to select and adjust training data, control data sources, and choose how data is transformed. They must have the ability to modify AI models when they are incorrect or operating outside of ethical boundaries.