Should your brand’s chatbot have a sense of humour?

If you think that the idea of developing a personality for your brand’s chatbot is frivolous: think again.

Many companies have discussions about their brand’s personality, whether brought on by a brand development exercise, or the question of how their brand comes across in TV advertising or, perhaps, how it is seen and heard on social media. Is the brand playful or serious? Traditional or nonconformist? Conservative or outrageous? Does it have a sense of humour?

Often these personality attributes remain somewhat latent. Companies that see their brands as risk takers or eccentric, often find that they don’t particularly want to broadcast the fact for fear of upsetting their conservative customers. Likewise, marketers who feel that their brands can have a little bit of fun on social media, because it is expected of them by other social media users, often don’t use the same sort of fun persona for other communications.

So, where does this all leave us when it comes to conversational marketing?

Conversational marketing is simply marketing using one-to-one communications via a messaging app, virtual assistant or other artificial intelligence-powered voice channel. The conversational nature of Amazon Alexa is now world famous, but there is also a new generation of AI chatbots that use natural language processing (NLP), machine learning and fuzzy logic to mimic human conversation.

The thing is, as chat and virtual assistant technologies develop, the likelihood is that many companies will have multiple editions of intelligent conversation agents, which might appear on their websites, on Facebook Messenger, on Whatsapp or even on Alexa or Google Home. End-to-end conversational platforms are being developed by technology firms such as Amazon, Google, IBM, Microsoft and Nuance Communications, to make this easier and easier to implement and manage. Therefore, it makes even less sense to skew a brand’s personality or mood according to channel.

Your chatbot may be conversing with a customer via a website chat app one minute, via Facebook Messenger the next and via Whatsapp later. Desktop users that are much less active on mobile apps are as much a target for conversational marketing as Messenger addicts. There’s no way of segmenting customer conversations by chat channel for style or tone of voice (if there ever was for any digital channels in the first place).

So, now consumers can chat with, or even talk to, your brand across multiple conversational channels. Doing this, you might reasonably expect that it should be likable and customers should feel happy on the conclusion of their chat, not only because their case has been dealt with, but also because it was dealt with in an efficient, polite and friendly manner. So, how friendly should our chatbots be?

Mumbai-based enterprise conversational Al platform Haptik has been developing chatbots with different personalities to test their effectiveness in different customer service scenarios.

Haptik focused on three personalities for its research: transactional, prosocial and friendly. They tested the three chatbot personalities, asking human participants to rate their experience across metrics for likelihood of re-interaction, their comfort level in dealing with the chatbot, and finally, the conversation’s productivity and efficiency.

As a matter of fact, Haptik’s findings were that user preferences were overwhelmingly in favour of the friendly chatbot, versus its transactional and prosocial counterparts. However, I’d argue that the ideal personality and disposition of your chatbot really depends a great deal on your service and customers. A positively effervescent chatbot may prove to be a smash hit with TGI Friday’s customers, but the same personality may not go down quite as well with, say, J.P. Morgan’s.

It’s natural to assume that customers with a need are simply going to use a chatbot to help fulfill that need as quickly and efficiently as possible, but its now become a popular pastime for chatbot users is to test your AI chatbot to see what it can’t answer and whether there are any hidden ‘Easter eggs’.

Many chatbots will answer questions such as ‘how are you?’, ‘are you busy?’ and ‘what’s your name?’ well enough, but some people just aren’t satisfied until they’ve asked the bot what its favourite colour is, if it can make them a sandwich, or perhaps see how it responds to an invitation to go out on a date!

Given that chatbots are such a popular source of entertainment, even for customers intent on progressing a purchase or other valuable transaction, should questions like these be entertained at all?

Makers of UK financial advisory bot Cleo have given her a chatty, informal tone of voice using lots of emoji in an attempt to inform and entertain, making financial advice a little more fun in the process. The company even rolled out a more flirty chat mode for its users early this year as a Valentine’s Day promotion (with somewhat mixed results). According to the developers, the chatbot has apparently been asked out on a date nearly 2,000 times and asked to send users nude images more than 1,000 times.

Developing a likable, engaging chatbot is not just about fun and games. Many conversational marketing experts now agree that developing a personality for your chatbot is pivotal for retention, repeat usage and driving value from your chatbot channel. Chatbots can engage with consumers 24/7, reducing enquiry response times and customer service costs, while increasing leads, revenue and customer retention.

So, creating an engaging personality and persona for your chatbot should not be dismissed as frivolous. There’s a strong argument for investing time and budget in your chatbot’s personality and committing to its long-term development. However, this may or may not involve scripting jokes for your chatbot. That will depend on its personality. Like some humans, some chatbots just aren’t any good at telling jokes.

This feature was first published by Carrington Malin on Linkedin.

Carrington Malin is an entrepreneur, marketing professional and advisor who has worked across almost every sector of technology. He helps companies, startup ventures and public sector organisations develop marketing strategies, digital initiatives and leverage new marketing technologies. He also publishes a daily Asia AI News digest. You can connect with Carrington on Twitter @CarringtonMalin or via Linkedin https://lnkd.in/furZ3s9