The Power of AI Voices: Hitting the Right Notes for Marketing Success
By Fotis Efthymiou
As AI voice assistants like Amazon Alexa and Google Home become increasingly integrated in various aspects of our daily lives, an intriguing question arises: How do the distinct characteristics of these voices shape our perceptions and influence our behavior?
A key factor that influences how we perceive voices is the Vocal Tract Length (VTL) - the anatomical distance between the vocal folds and the lips - which determines the frequencies a voice can produce. Longer VTLs result in deeper, more resonant voices, whereas shorter VTLs produce higher, shriller voices. Thanks to advancements in AI-generated speech, we can now digitally mimic these effects. By leveraging this technology, AI voices can be precisely tailored to products or brands that share similar characteristics, which in turn can trigger quick associations in consumers’ minds.
Our research shows that voices with longer VTLs (and thus deeper voices) are a better match for products traditionally seen as more masculine, such as a pork steak or a large truck like the Ford F-150. By contrast, shorter VTLs align more with products that are often coded as feminine, like a green salad or a small car such as the Fiat 500. When such voices are used for advertising purposes, we observe that consumers respond more positively when there is a match between the voice and the product’s personality, leading to higher engagement, increased click-through rates for ads, and reduced advertising costs.
This phenomenon aligns with the broader principle that sensory stimuli trigger automatic associations in our minds. Just as certain colors or shapes influence our perceptions and expectations, the sound of an AI voice can evoke associations with specific product or brand attributes, creating a more intuitive connection between the voice and the product or brand. Our research also suggests that the mechanism of this phenomenon is tied to how we perceive the physicality (i.e., height and weight) of the entity “behind” the voice. Simply put, people tend to link deeper, more masculine voices with larger and heavier entities, while higher, more feminine voices conjure associations with smaller, lighter ones.
For businesses, these insights hold significant potential and unlock a world of possibilities for their marketing and advertising purposes. While many companies currently use arbitrarily chosen AI voices for marketing, advertising, and customer service, tailoring these voices to reflect specific, context-relevant characteristics can significantly boost customer engagement and enhance speech interactions with AI. Such digital voice engineering gives brands a new level of control, allowing them to fine-tune their marketing strategies with higher precision.
Looking ahead, AI voice technology could become even more adaptive. Beyond just specific voice characteristics, AI voices may soon adjust to a consumer’s current emotional state or psychological need during the purchase journey or customer support. For instance, an AI voice could adopt a calming tone during stressful decision-making moments or sound more enthusiastic when presenting exciting offers.
Ultimately, our research highlights that the voice of an AI assistant is more than just a medium for delivering information—it is a crucial element in shaping how products and brands are perceived. A well-suited voice can make a brand stand out and resonate deeply with consumers, and the technology to create these tailored voices is already within reach, ready to be leveraged.
Article details
The Power of AI-Generated Voices: How Digital Vocal Tract Length Shapes Product Congruency and Ad Performance
Fotis Efthymiou, Christian Hildebrand, Emanuel de Bellis, and William H. Hampton
DOI: 10.1177/10949968231194905
First published: November 15, 2023
Journal of Interactive Marketing
About the Author
