Meet the nextgen of chatbots: Personality-based AI
The way in which people engage with brands has changed. Gone are the days when it was a one-way interaction; younger demographics especially love a value-add exchange with brands they can identify with.
Chatbots have been heralded as one way of opening up the channels, but as with every new marketing tech innovation, you need to get it right. Brands need access to tech that can do the job well and not make the experience clunky and robotic.
One company that recognises this issue is Israel/US based imperson. The founders have been working on AI since 1999, forming a company, AI, that turned into a research facility once it realised the market for chatbots wasn't quite ready. imperson itself was founded in 2014, when the market became ready, harnessing the power of the previous company, AI.
In 2015, imperson graduated from a Disney accelerator, and has since used its unique idea of bringing movie characters to life through interactive chatbots to work with major media agencies such as Disney itself and Universal.
And just last month, the company launched the chatbot platform for enterprises to create and manage their own personality based bots for platforms such as Messenger, Amazon Echo, Skype and Twitter.
Co-founder Eyal Pfeifel and marketing director Avital Rabani spoke to MarketingTech about imperson’s clever bots.
What does it do?
imperson’s tech is not just an automated keyword based response AI; it’s focused more on working with people to write personalities for the bots it uses.
So far, the tech has been used create bots for already existing fictional characters such as Miss Piggy and Back to the Future’s Doc Brown.
imperson has also developed personalities of its own for the bot to integrate with, such as its ‘intern’ Artie, which sits on imperson’s Contact Us page and has quite a distinct personality. Engaging with it, it’s very easy to forget it’s a bot.
While this may seem very novel, it’s worth noting that the format is working at engaging people. It’s seeing quite a lot of success, with nine minute length average conversations and a 20% rate of re-engagement.
Bots ‘coming of age’
When the company was founded by four co-founders in 1999, it established capabilities in the area of conversational interfaces, where computers could handle naturally flowing conversations with the users. “What we realised at that time was although we had developed the tech to do it, we couldn’t convince people to use it,” Pfeifel explained.
People weren’t used to using the same messaging apps as we have today, and the notion of wanting to engage in conversation with a computer was seen as “crazy”’.
So, the company was kept on low fire and relaunched two years ago, updating and upgrading its tech. This was, Pfeifel added, due to a feeling that the market had changed, thanks to widespread use of messaging platforms and virtual assistants like Siri and Cortana.
imperson’s founders also decided to give people a reason to engage with bots, so it aimed itself at the media market, where there were already existing famous characters that had a fanbase who wanted to engage - despite being fictional.
It managed to secure some interesting projects involving high profile characters, using the bots as promotion for films.
“It was quite unlike 1999, today it’s socially acceptable as we can deliver the right experience. People don't have this ‘computer feeling’ when they chat with bots and this enables their experience; they feel they are speaking with a famous character. It's authentic and a fulfilling experience for them, entertaining and fun.”
Creating a bot personality
The technical aspects of creating a bot are one thing, but how do you ensure you have the personality right?
Each character imperson creates a bot for is a separate content project, Pfeifel said, and there is a team of writers who build a personality profile in a similar way to how it’s built for a movie/tv series.
“Each one has a profile, we document it, know everything about the personality and we build its language in terms of how it phrases itself and what it will and won’t use. This must all be very clear and accurate. We do this with the customer, so for example in developing Doc Brown we worked with the original script writers of Back to the Future,” he explained.
This is then implemented on top of the technology to create a “naturally flowing conversation which is an accurate simulation of a personality and feels exactly like the original in the movie.”
The future: Personality bots for brands
As mentioned, imperson has opened up its tech to partners to create their own personality-driven bots for themselves. Imperson thinks that personality driven bots, rather than ones that merely answer simple customer service related questions, can play a big part in brands’ marketing, Pfeifel explained.
“Every major brand has a ‘voice’, and a voice is a personality so we’re saying, create a branded bot that will speak your brand language,” he explained.
Text bots aren’t the only kind that’ll change the landscape of customer service though, with voice-based also increasing in popularity. As co-founder of Naked Communications Will Collin wrote recently, “With the advent of digital intelligent assistants like Amazon’s Alexa, Apple’s Siri and Google’s Assistant, artificial intelligence and machine learning programmes are finally starting to appear within everyday consumer products.”
In the long run, text chat bots may ultimately have more scope owing to the proliferation of messaging apps and is where imperson is really gunning for.
“This is a channel is far from being tapped by brands and marketers and where we believe our tech can grow,” they concluded.
Personality based bots certainly have an appeal, but like all forms of marketing tech, you need to evaluate if it’s something you need - or a ‘nice to have’.
Would you consider implementing one? Let us know!
- » Three key ways predictive analytics can improve social performance
- » Are you ready for the “access on demand” economy?
- » Three interesting ways that AI is changing the hiring process
- » Moving from AI to ‘applied intelligence’ to identify revenue sources and deliver business value
- » Improving CX strategies: Get the balance right between technological and staff investment