By Lebo Moerane, Head of Social at VMLY&R South Africa
The intersection between artificial intelligence and social media management is an interesting one. Social media is where brands go to behave more human; AI is synthetic humanity – a machine’s ability to perform the cognitive functions we usually associate with human minds.
But AI can also do things humans can’t, like create at scale and process large volumes of data in a short space of time. And this is where bringing the two together creates magic.
AI and Social Media: A Match Already Made In Cyberspace
The combination of AI and social media is nothing new. In fact, the two are so closely intertwined that we don’t even see the AI at work.
The personalisation we’ve come to know as a hallmark of social media – like friend recommendations on Facebook or freakishly well targeted content on TikTok: all brought to you by AI. And these recommendation engines are getting smarter and more powerful. TikTok’s gravedigger algorithm, for example, is a nifty piece of code that finds old content buried in the depths of creators’ feeds and breathes new life into it by serving it to new audiences.
Similarly, the tools brands use to create targeted ads – from gathering platform audience data to identifying matched and lookalike audiences, and even dynamic content optimisation – are all powered by AI.
Deep text is an AI mechanism that helps platforms filter out spam, hate speech and fake news and understand slang, while facial recognition is the AI that allows you to be recognised and tagged in unflattering pictures. A sobering reminder to review your privacy settings.
More recently though, AI’s relationship with social media has evolved.
Faster, Bigger Creativity
AI allows brands to create with speed and at scale. It also means personalisation becomes a lot more accessible. Smart brands are using it to create campaigns that would have been impossible or prohibitively expensive without the tech – like the Dutch bank ING, which gave clients an AI-powered tool to generate life-like images of their ideal retirement, inspiring them to invest using the power of visualisation.
When it comes to connecting with audiences, brands can use AI to be more human than ever by interacting at the speed of culture. No more sad, after-the-fact attempts to be relevant: Brands can use social listening to stay on top of what's trending, then generate genuinely relevant content before the conversation dies. In an extreme example, the Woods Art Institute in Germany used AI to identify trending topics and then turn them into art, creating the world’s most contemporary art exhibition.
What makes these effective campaigns, as opposed to quirky gimmicks is that in each of them AI was the enabler, not the idea. That’s the key – whether it’s AI or a new social media channel, we should never use tech because we feel we should be using it or so that we can appear cool and current. Tech is just the tool that brings great ideas to life. Always ask, “What’s the objective?”
24/7 Community Management
People engage with your brand on social media at all hours of the day and night, but your community managers may not be keeping the same odd hours. Chatbots help fill the gap, but up until now, they’ve been limited in their capability and lacking personality, leading to a lacklustre customer experience.
AI can make chatbots more interactive. An AI-powered chatbot can engage with customers like a human community manager, responding to queries, resolving service issues, processing refunds and recommending products and services – in different languages and dialects. Tech as seen in AI companion generator Replika could allow a brand to embody an avatar that's available to customers 24/7. But exercise caution: like any human staff member, bots can go rogue. Train them well.
Share of voice as a metric has major shortcomings; it largely depends on how having an open API or keywords that the system can pick up. A mention and a piece of content are very different levels of engagement, but the traditional SOV metric doesn’t allow us to separate the two. Now, using AI-powered technology like image recognition, brands can finally move beyond SOV to more meaningful metrics, like share of user generated content or share of experience.
P.S. Read the Ts and Cs
One last thing: We’ve all claimed to have read the terms and conditions on a website with the casual dishonesty of a teenager sneaking into a nightclub, but AI tools are not the place for white lies and deception. The law is still lagging behind the tech, so brands must be careful to use tools that stipulate they will own what they generate to avoid legal spats.
This article first appeared on Media Update.