Chat Marketing and chatbots, I reluctantly have to admit, are here to stay.
Why am I reluctant? Because I find them impersonal, intrusive and spammy.
Why am I torn? Because I have FOMO and I fear I will be heartlessly abandoned in the desolate fields of outdated marketing, a dinosaur become extinct all because I detest spamming.
A wee bit melodramatic? Maybe not. (Apart from the dinosaur bit.) AI is moving so fast that it’s not unreasonable to assume that marketing and many other fields of work will be quite radically different in just a few years.
In April 2013, the Associated Press Twitter account sent out this tweet: “Breaking: Two explosions in the White House and Barack Obama is injured.”
Within six minutes Associated Press confirmed the tweet was a fake, and only three minutes after that the White House spokesperson confirmed that President Obama was fine and there had been no explosions. But the White House, and the President, had never been the target. The target was the financial markets and the virtual explosion certainly hit them hard. Almost immediately the tweet was posted, algorithms in the trading software identified keywords which told them there was a major crisis. The markets lost $136 billion in approximately three minutes.
This was not a human mistake, indeed the only reason the financial implications weren’t even worse was probably because a human saw the tweet and immediately contacted Associated Press for confirmation. The bots running the trading algorithms react far faster than any human could, and they could not avoid this error which should make all of us wary. AI systems now use deep learning neural networks which predict behaviour and reaction and start to learn from those predictions. But Fintech is just one of the areas banking heavily (no pun intended) on AI. And unless we go completely off-grid, we aren’t going to be able to avoid large swathes of it.
Have you ever discovered a transaction you didn’t make and phoned to cancel your card? Soon you won’t need to. Thanks to the Internet of Things, financial & many other kinds of institutions will already have enough information about us to know whether the transaction was genuine or not.
Transaction made in Chicago? That’s ok, information sent from your phone shows you in Chicago airport the night before.
Transaction made in Lithuania? Alert – you were at home in Sussex and have not visited Lithuania in the last year according to phone records and data from all the other sources connected to the Internet of Things. Soon even your microwave will be spying on you. And your method of payment won’t be a plastic card but something you are wearing which will be connected to you biometrically.
Naturally, this opens up a Brave New World for marketeers. Google and Apple already scan our phone GPS records and we are happy to be able to find our way to a new meeting place using Google Maps. I used it only today to discover a Mexican restaurant in Leamington Spa. And delicious it was too. But do I now want to be targeted with discount vouchers for that restaurant, or for other restaurants in the same shopping centre? And if I bought something secret for my husband on our jointly owned card do I want him to receive targeted messages to his phone from that shop?
The implications for the AI involved in the Internet of Things is so far-reaching that it will completely change the way we live. In 1991 I went to Japan to teach English for the Japanese government, and I ended up staying for three years. Nobody I knew used the internet. Nobody had mobile phones. There was no Skype, no Facebook, no Twitter. Nobody used email. Landline calls were jaw-droppingly expensive. Life would be very different, and much less lonely now.
So that’s a massive change in my lifetime, a change that my children, in this world of devices and instant information find hard to conceive. I was so bored I had to watch the few subtitled video recordings I had (video!) over and over again. This was not good, as they mostly involved Arnold Schwarzenegger who was unfeasibly popular in Japan at that time. Plus Robocop. No Amazon Prime. No Netflix. No streaming. Just Arnie, machines with guns and Sumo wrestling.
The changes that are coming are even bigger, and as humans we don’t like too much change. Self driving cars? No thanks. I’ve seen Terminator (quite a few times, when it was the only film I had in English.), I know this doesn’t end well. Frankly, Conscious AI such as that being developed by Google Deep Mind or the Chinese search engine Baidu scares the heck out of me.
By 2021, the total number of devices connected is projected to be three times the population. Internet of Things technologies are expected to account for more than half of these 27.1 billion devices.
Maybe that’s why I’m resisting Messenger Marketing. I’m seeing it everywhere but for me Messenger is private and a message from a stranger, especially one trying to sell me something, especially one that’s a robot, is the equivalent of cold calling on my doorstep. When that happens on my doorstep at home, the salesperson gets a polite no thankyou (alright, perhaps a grumpy no thankyou) and the door is closed. If a robot was at my doorstep I might well call the police. In Messenger it’s not quite so simple.
We (mostly) trust Apple and Google. We give them access to much more information than we realise. At CMA Live recently in Edinburgh it was the first time I noticed that Uber had intergrated a bot into Google Maps, enabling you to connect with them even if you hadn’t downloaded their app.
We (mostly) trust Amazon and don’t freak out when we receive an email recommending goods based on our browsing history. This kind of predictive marketing is going to expand across every aspect of our lives. In Fintech the rough date we will buy our first house will be predicted and we will start to receive offers. Instead of asking the date our insurance is due, insurers will know and will provide us with a tailored quote. Car dealers will predict when we will need a new car, rather than hoping that we’d like a new model. The list is endless, and some of this chatter is going to be arriving through your Messenger box and will be delivered by bots.
“Despite these astonishing advances, we are a long way from machines that are as intelligent as humans – or even rats. So far we’ve seen only 5% of what AI can do.” Yann LeCun, Director of Research, Facebook
I don’t have a problem with this where there’s an opt-in which makes it clear you are going to be giving the business in question access to your Messenger account, after all that’s just like giving your email address. I do have a problem where sneaky marketeers are tricking people into commenting on a post without telling people that this will mean the business can now reach them in Messenger. I’m seeing increasing complaints in Facebook Groups about instances where this is happening and the person then gets spammed in Messenger. That’s where the cold calling analogy comes into play, and it is the attack on people’s privacy which makes me so uncomfortable about the whole AI situation.
There’s a place for chatbots, no doubt. They can live in any chat platform (Facebook Messenger, Slack, text messages etc) and that’s where lots of us hang out lots of the time. FAQ type questions which get asked over and over again are perfect fodder for them, as long as the conversation doesn’t get too complicated. It’s a way of ensuring almost instant customer service for customers who demand an instant answer.
It can also take the place of actual internet browsing – soon you will be able to message the brand on Facebook Messenger for example and tell them what kind of shoes you are looking for. The bot will replicate the store attendant by ‘bringing’ you relevant shoes. The whole transaction will be able to take place within the messenger platform. Bots can also be used for weather, advice, news, personal finance, grocery shopping. There’s even a bot in China called Xiaoice who will be your friend.
Any business can easily create a chatbot now as part of their digital transformation and use it to automate part of their service, customised to fit their audience’s needs. An astonishing statistic from Gartner tells us 85% of customer relationships will be with Artificial Intelligence by 2020. In the same article we learn that 60% of those who use bots are in the 16-19 age bracket which puts me well out of their target audience. Phew, perhaps I’m not quite the dinosaur I thought. You could visit Chatfuel where they boast you can launch a full-featured chatbot in 7 minutes without knowing how to code. You could build a bot like MyKai, from the research lab that helped create Siri which tells you what you’re doing with your money (including stuff you’d rather not know such as how much you spent in coffee shops over the last year), or Personetics, another financial-savvy bank bot.
By next year 80% of all smartphone users – that’s 2 billion people – are expected to be using Messenger Apps like Facebook Messenger, WhatsApp and Kik. Brand bots integrated with these applications are in a unique position to be selling to Millennials and Gen Zs who want to make a quick, easy transaction on their phone. Converting older people is going to take a while longer. While they might accept that AI isn’t going to destroy the world a la Skynet, they’re not too impressed about it destroying jobs.
So this is why I am torn – between privacy and innovation, between sticking to my principles and being left behind.
If you would like to read some more, or you need convincing why your business needs social media, have a look here.