To all the choices about which personalized digital experience can best help us make decisions, add a new arrival: the virtual assistant. 

Most people are familiar with one very basic VA—Siri. But, as with all things AI, the technology has become increasingly sophisticated. 

Consumers’ habits are evolving alongside the latest tech developments. You might have paid $20 per month for the premium version of ChatGPT—while also paying $20 per month for Anthropic’s Claude—until ChatGPT started offering the same features so you canceled one. 

A lot of consumers have strong opinions on the pros and cons of different virtual assistants, especially in this early adoption phase. Usage decisions are currently being made on technical criteria such as learning and understanding, interoperability and natural language processing.
But the next frontier is going to be based around more subjective questions: Has a virtual assistant been trained on the data, politics or sensibility that the user prefers? How trustworthy is that data? Are you willing to pay a premium for a specialized, curated experience?

This is where marketers have a huge opportunity. But to find success, some key considerations must be made. 

What’s your tone? 

As with the launch of any new product, the message is everything. 

When Elon Musk launched Grok AI, he said it was going to have a snarky tone and be anti-woke. With that announcement, he opened the floodgates for marketing at every virtual assistant and beyond. 

When Musk threw down that gauntlet, it became apparent that it matters very much how virtual assistants learn. If we are going to use them to make decisions, we want to know that the information they are learning is true and smart. 

Liberal Democrats will never rely on a VA that is proud to be “anti-woke.” This has created a new consideration set on VAs, and purchasing decisions will be impacted by these kinds of marketing strategies. 

What’s your audience?

Grok’s users feel more tied to Musk than those who originally signed up for X, the social media platform formally known as Twitter. 

As VAs are increasingly tailored to whatever a user wants, it’s more important than ever to understand, value and maintain your audience. To do that, you need to look at your existing content. 

There’s already talk about how The New York Times’ case against OpenAI could transform how AI and copyright works, but it could also transform the VA space. We might not be far from a future where media companies concerned with intellectual property will create VAs from their own content and restrict others from using it. 

What’s your branding? 

It is easy to imagine a time when we all pay dearly for a VA that has been trained by all the knowledge that exists at, say, a particular elite institution of learning. That opens up a marketing opportunity, where every important university has a VA brand extension strategy. 

This is similar to the way fashion is marketed. Couture fashion businesses don’t make money; instead, they create a brand halo that sells bags and perfume. 

The same principles will be applied to VAs. An elite university will continue to educate tens of thousands of students each year. But should they decide to develop a virtual assistant—say, the Harvard VA—at $50/month, it could be used by tens of millions of people. 

With that kind of potential, you can expect major brands across every industry to jump into the VA space. 

How will it be distributed?

Just as every brand rushed to build a website, many will rush to build a VA where they can demonstrate expertise and control the flow of information. 

Given the costs of building a Large Language Model, most brands won’t be able to do it themselves, which means that companies including Anthropic, OpenAI, Meta, and Alphabet— which have already built LLMs—may get into the licensing game. 

The bottom line

2024 will be the year that consumers ask, “How is my VA learning” and the brands will respond with marketing campaigns. It’s important for marketers to track the evolution of this major new space, and for engineers to learn lessons on how to customize this exciting new tech to really land with consumers.