In past times on hearing the word ‘chatbot’ I would feel a cold shiver going down my spine as I recalled all those disjointed customer service experiences, trying to converse with AI and not ever resolving the problem or query. But hey, I will admit when I’m wrong and like anything it’s not necessarily the technology which is bad, it’s just the way its been set up or used.
I have recently been working with the folks at Oracle on the implementation of chatbots on my social messaging platforms.
When you manage four extremely active social accounts, with impressive levels of engagement, with another two potential pages on the way; plus realise that to drive growth you need to develop a business profile on Whatsapp; you know your Community Manager is going to be just a little overwhelmed.
Social media has to a degree reversed from being about brand messaging to being about customer messaging and unless a seamless experience can be delivered across these platforms you run the risk of alienating your fan base. So a chatbot can help? To a degree yes.
What is a chatbot and what can it do? For a start if you do not know your customer, follower or fan then the chatbot will not be able to support you effectively. It creates a conversation based on perceived questions, so first you need to work with your social media and customer service teams to identify the most asked questions or those inquiries which may arise in the future due to new brand activities, for example.
ALERT! Chatbots do not yet work on social post queries, currently they will only run on message channels such as Facebook Messenger and Whatsapp. And this is a good thing. If you really want to demonstrate you care about your followers having pre-programmed responses on a post will kill your engagement, again back to the brand versus customer messaging tool. Credibility will be lost unless audience interactions receive a perceived bespoke response.
Facebook Messenger already has a degree of ‘chatbotism’ built into its functionality. If you run an ad campaign on the platform you can pre-programme potential customer queries which are surprisingly quite effective. I have recently seen customers having a conversation with the responses so if you want to experiment this gives you an opportunity to test on a smaller scale.
Having a tool which is live 24/7 across e-commerce and messaging platforms, can manage multiple languages and which can either answer the query (store location/contact, opening times, price, product availability); or divert to customer services (return, complaint, appointment with a Personal Shopper); or the brand community manager (event, specific product launch) dependent on subject matter; is a massive leap forward in customer engagement. Particularly at a time when brand teams are being reduced but the customer appetite for e-commerce is growing in the region.
BUT P.S. the chatbot will only work effectively if it is programmed to replicate the brand tone of voice – emailing a list of questions and responses to the provider is not enough, remember the cold shiver, the more ‘chattieb’ sounds like a member of the brand team via well thought through, scripted scenarios, the more effective and ultimately believable it becomes.
There are many ‘Chatbot’ providers, Oracle, are just one click here to read more insights from the amazing Nandita Saggu of Blue Logic.
Susan Hewlett 