By John Egan

From Larry Smith’s vantage point, some fear mongers and mainstream media outlets are giving artificial intelligence (AI) a bad name.

Broadly speaking, general-purpose or strong AI is characterized by “Star Wars”-style or “Terminator”-type technology that’s designed to replace humans, according to Smith. The reality, he says, is that businesses are adopting special-purpose or weak AI aimed at taking on fairly narrow tasks, such as customer relationship management (CRM), sales prospecting and performance optimization.

“It is important to distinguish what kind of AI you are talking about and what problems it will solve,” says Smith, a partner at IT consulting firm Thematix Partners.

During DMA’s &THEN marketing conference Oct. 8-10 in New Orleans, Smith will offer his insights about AI during a session titled “State of Play: AI-Enabled Personalization.” Joining him will be Jake Bennett, chief technology officer at digital agency POP, and Kate O’Hare Mercier, vice president of marketing at clothing retailer Charlotte Russe.

From a marketing perspective, Smith says AI will bolster both performance and ROI.

“Properly fed with all types of data, CRM systems will be able to uncover and highlight opportunities to prospect for new customers, identify current and lapsed customers, and serve the most profitable customers,” he says.

“Every person will have a profile and a personal plan for relationship interactions. A wide variety of triggers will automatically spawn reactions and solicitations tailored to that person; over time, the AI will learn and improve according to KPIs [key performance indicators] and ROI.”

To fully embrace AI, Smith suggests marketers:

Monitor their competitors’ AI strategies. “Are there significant changes to major KPIs?” he says.

Track the AI innovations of CRM vendors, and ask for demos to see whether there’s value in adding that technology. “Consider doing a proof-of-concept test,” he recommends.

Be wary of AI applications that are overhyped. “These are the early days,” he says, “and many so-called AI solutions are just algorithms in a black box and lack the ability to learn.”

When looking at AI technology in marketing, your brand should conduct tests or undertake proof-of-concept initiatives in areas like marketing automation, lead generation, programmatic advertising, CRM profiling and social media, Smith says.

Keep in mind that a lot of AI technology is being added to existing software, sometimes seamlessly, so you might not be immediately cognizant of a product’s range of AI capabilities.

“AI is like salt or sugar for food; it makes everything better,” he says.

If AI isn’t incorporated into an existing product, remember that AI development demands a considerable amount of time, data and training, Smith says.

Companies seeking to take the customized approach to AI “need to start the journey now in order to realize some benefits in the future,” he warns.

This article is brought to you by &THEN, DMA’s annual event. Click here to join the leaders of the marketing community and explore how data transforms marketing in New Orleans, October 8-10. Save over $300 when you register online.