This post originally appeared on the Wunderman Thompson blog.
Conversational technology is omnipresent. It’s becoming clear that constant, real-time, personalised communication is dominating customer-centric services.
Indeed, Forrester research found that using chatbot technology leads to a 71% increase in customer satisfaction. After optimising chatbots for customer service, the sector is now positioning chatbots as an outlet for commercial messaging. Yet, at first, we weren’t convinced.
At Wunderman Thompson Belgium, we launched the Ask Yugo chatbot for Telenet a couple of months ago. This FAQbot helps customers during their whole journey: from product discovery and first try, to daily use and for questions or guidance. Through talking with customers, the chatbot grows smarter and improves its answers. When the chatbot can’t answer a certain question, it will automatically direct users to the client services team.
After delivering this state-of-the-art chatbot assistant, we decided to dig deeper and look into the current capabilities of chatbot technologies. One of the latest innovations is chatbot ads, which are, according to Facebook, “relevant promotions [sent] to the people your business is already talking to in Messenger.”
The benefits? Chatbot advertising could enable you to send highly relevant commercial offers.
That may sound great, but as these messages are sent in a one-to-one channel, we’re hesitant about this new advertising format. For one, simply by interacting with a brand’s chatbot, the user consents to receive offers. Other than a paragraph in Facebook’s user agreement, users are not made aware of this, which we suspect will lead to irritation when they realise. Indeed, chatbot offers could easily be perceived as intrusive, which is something we believe a brand should avoid at all costs.
“CHATBOT OFFERS COULD EASILY BE PERCEIVED AS INTRUSIVE, WHICH IS SOMETHING WE BELIEVE A BRAND SHOULD AVOID AT ALL COSTS.”
Will chatbot ads irritate consumers?
To date, we couldn’t find research investigating what determines the acceptability and effectiveness of chatbot offers. So, we did it ourselves.
We partnered up with the University of Antwerp to look into this new phenomenon. And last month, we published our findings in the journal Computers in Human Behavior in an article entitled, “Chatbot Advertising Effectiveness: when does the message get through?”
In our study, we let 245 participants interact with a small Facebook chatbot designed specifically for our study. We asked them to book movie tickets at a made-up chain of movie theatres. Afterwards, the chatbot sent them a message offering a range of snacks they could purchase for the film. Two results stood out:
1) As expected, we found the effectiveness of the chatbot offer is, to a large degree, dependent on whether it manages to avoid being perceived as intrusive.
2) In turn, the intention to act on the offer was found to be partially dependent on the degree of involvement with the brand.
How can a chatbot ad be used effectively?
Our findings pointed to the crucial role of qualitative initial chatbot service. An initial chatbot experience that’s evaluated as helpful and useful will make people more receptive to follow-up communication.
In short, we recommend that chatbot ads are only used on involved audiences who’ve already had a pleasant experience with the chatbot.
Before retargeting users with offers, brands should first get their basics right. At Wunderman Thompson, we’re convinced that, when done right, the unique characteristics of chatbot communication can be leveraged to provide real value both to the customer and to the brand.