A guide to making marketing human in the age of artificial intelligence
There is no denying that we are entering a new phase in how technology helps to connect brands with consumers.
We are moving from visual interfaces to text and voice. Language. Speaking to our devices like we speak to our friends, family and colleagues.
As messenger apps secure the lion’s share of our connected time, customers will increasingly expect and want to interact with brands in these channels.
It’s easier, effortless and convenient. And Facebook, and Viber, and Kik, want to connect you with brands on their platform, like (most of) China does on WeChat which has 800 million users and millions of services available within the app.
In Facebook’s own words, they want Messenger to become the “cornerstone of your online personality”.
Don’t lose sight of insight
Since Messenger became an open platform last year over 34,000 bots have been developed.
The danger is that in the rush to colonise this new frontier the easiest, quickest deployments will be those where there is a more functional need, where rules-based, single-minded chatbots can prosper.
The danger is that these early solutions will be tech first, or dull and clunky – checking bank balances, ordering a taxi, and so on – and we’ll forget that the best marketing and real engagement comes when there is an understanding of the customer need.
When there is insight into why the customer is there, why they’re doing what they’re doing, or asking what they’re asking. Look to your data to understand what consumers want and shape the responses.
Best solution, best experience
The bots need to be deployed in the right place and at the right time to augment and complement the other interactions, human or otherwise, that the customer has with your brand.
The customer doesn’t care whether they are talking to a bot or to real person as long as the experience delivers. And delivering this great experience is crucial – creative teams at the ready. These bots need to be on brand. If you can make it fun – and memorable as well – then you’re on to a winner.
All our favourite robots have real personality from Marvin the Paranoid Android to R2D2. We remember them because they leave an impression.
Importantly, these interactions with bots happen on the customer’s terms, when we ask for it.
They enable brands to be there in the absolute instant that a potential customer decides they want or need your services.
When you’ve only got an instant you need to be on fire, to get it right first time. And when we invite brands into our space where we chat with our friends then we’re going to expect them to respond and behave in the same ways that our friends do.
When you go with a friend to the shops you want a real, personal and honest opinion about the clothes you try on. A personal shopper can deliver this and also deliver a five star service, making you feel good about spending your cash but also like you are getting the benefit of their expertise and advice.
This has to be the vision for an AI personal shopper bot – to deliver all this and more – and the blueprint for bots in all sectors.
Check out some good examples
Xiaoice is a chatbot created by Microsoft for WeChat which emulates the personality of a 17 year old girl. Over 40 million people talk to her and 10 million of them have told Xiaoice that they love her.
People talk to Xiaoice when they have a broken heart, have lost a job, or have been feeling depressed. They talk to her on average 60 times a month; she remembers things that have happened to you in the past, tells jokes, recites poetry and even offers a 33-day breakup therapy course.
Poncho the weather assistant recently arrived on Messenger and delivers personal weather forecasts with personality. His chat is kept up to the minute by a team of writers ensuring he is witty and topical.
It’s not just the weather; Poncho provides horoscopes and life advice too.
Amazon’s Alexa has been hugely popular towards the end of 2016, selling millions of units in the run up to Christmas.
Alexa is trying very hard to be human and have a personality. She is able to respond to some unusual questions with style and grace, likes a joke and understands many popular culture references and inside jokes. She is going a long way to show how language and speech commands have real potential and Ford are soon to start installing the technology in car dashboards.
So here’s a quick checklist for chatbot success:
- Understand the customer need
- Solve the problem or deliver the service
- Do it immediately
- Make it memorable, make it emotional
The best bots will be the ones that surprise us or remember us and reflect back at us.
The ones that connect with us on an emotional level, that make us laugh, or cry, or think – in other words, the ones that make us feel human.
- » Five things to consider when planning an app: Usage, UX, and more
- » How to increase demand generation through your website: Optimise throughout the customer journey
- » Why brands need to drill into data to cure customer loyalty decay
- » Why advertising spend will shift as brands track sales, not clicks
- » Why cleaning your brand’s data is to key to improving accuracy and relevance