Henrik Werdelin

View Original

Branding in the Age of Chat Bots: How do you create an emotional bond solely through a chat window?

As AI chatbots become increasingly prevalent, we face a new challenge: rethinking the nature of branding. Traditional branding strategies may no longer work when customers increasingly interact with AI-powered chatbots instead of human representatives or nicely designed apps or websites. This is becoming increasingly clear from experiments we have been doing at BARK with “Conversational Commerce” and our AI incubation platform Audos where we help hundreds of new founders connect with their customers.

When text is the only tool you have to create an emotional connection, it becomes essential to how the AI-based agent “talks.” Is it funny? Are there things it shouldn’t talk about?

One new concept we are using is "neighborhooding," an approach to branding via AI prompting to create a more personalized, welcoming, and community-oriented customer experience. Obviously, most people like to interact in their native language. So that's an initial step. However, “neighborhooding” goes beyond the mere localization of content and services. While localization involves adapting content to align with local slang, cultural references, regulations, and cultural norms, “neighborhood” takes it further by fostering a deeper sense of familiarity and belonging. Just as in any actual neighborhood, where every element—from the local dialect to the visuals in advertisements—is tailored to make the users feel at home, an AI-powered neighborhood aims to create a similar sense of comfort and connection for users. We distinguish “neighborhooding” from personalization tactics. While both aim to enhance the user experience, “neighborhooding” focuses on creating a sense of community and belonging without being too invasive of its customer’s privacy. It's about making customers feel welcomed and valued in a familiar setting rather than feeling they are about to be tracked and targeted across the internet.

We are learning via Audos that we get better results when we are super purposeful in designing the tone of the AI bots. A few examples:

  1. We get better response rates when we ask “What may I call you?” instead of “What is your name?”.

  2. We create a pretty unique character with a personality to express the brand position via tone of voice. (Think, if your brand were a person, who would it be?) This also has a fun effect on bots inside an organization. Eg. BARKs HR/People bot is called Bossy and is being trained to have a bit of a snarky attitude.

  3. We define what the bots should avoid discussing and where they might take a non-conventional position. For example, the BARK user-facing bot “Sassy” is not fond of cats…