Artificial intelligence (AI) is one of the bigger things that’s just starting to hit mainstream marketing. The early adopters who have been a bit ahead of the game have been integrating it into their activities. But what exactly is artificial intelligence in the marketing world?
AI can take many forms in marketing:
- Predictive analytics
- Recommended purchases (via machine learning)
- Social media algorithms
Basically, these overarching functions help shave time off the mundane parts of marketing – just like marketing automation is supposed to do. But, a little differently than automation, of course. It can even help save social media marketers some time.
But social media is supposed to be the antithesis of AI, right? It’s supposed to be human at its core. Well, yes, you want social media to be authentic above all else. But social automation has some fingers dipped into AI as well.
So, here’s where things get tricky. Because we know that social media automation has some great uses:
- Scheduling posts
- Social listening
- Initiating campaigns
But social media automation has some big risks, too. Things like sounding robotic in automated replies, or automated replies being inappropriate to the situation. Adding AI into this mix, when done well, can help mitigate some of those risks. But it can never, ever replace human interaction. Be prepared to respond personally to any conversation you start on social media, no matter how you start it.
How AI can boost social automation
AI has already snuck into social media, very subtly. Social media marketers might not even consider this AI, but in reality, it is. And the bigger pieces point to some huge potential for social automation.
1. Finding the “best time” to post to social channels.
This is already here – how else does your social scheduling solution know when the best time is? But there’s even “smarter” scheduling than you might see in your favorite social scheduling software. Meet Yala. (Hi, Yala.) She’s a clever piece of software who can schedule your social posts, but also learn when to schedule your posts to when your audience will be most engaged… and she’ll adjust your posting schedule as your audience evolves. She uses machine learning to improve your scheduling with every post.
2. Predictive analytics.
Predictive, really? Yep. Wouldn’t it be amazing to have a good idea of which posts will perform the best for you? The data team at the New York Times build a slackbot (called Blossom) to do that – it predicts which stories have the most potential to go viral. A part of the Social Bakers analytics suite can predict which organic content to promote based on predicted future performance. This is all based on an analysis of what kind of posts have done well for your company in the past, coupled with what kind of content your audience likes, so the software can predict how different kinds of content will perform.
3. Assistance with writing posts
What’s one of the things that we humans can bring to marketing? Creativity! So we can count on AI never putting creative types out of a job… but even we need help with understanding if a piece of creative is going to be effective or not. Enter AI. CoSchedule has an analyzer for both headlines and social messages (two of my go-to tools) to help write engaging copy for the relevant media.
There’s also a tool called Pi, from Post Intelligence that learns how you post, what your audience likes the most, and then finds that kind of content for you. It can even start to write short posts for you in your brand voice.
4. Facial recognition
From a user perspective, this can sound creepy, but from a marketer’s perspective, it’s pretty cool. Say you’re a brick & mortar store of some sort, like a restaurant. A guest posts a photo at your place of business, and tags you or checks in there on social media. Imagine the facial recognition feature of Facebook AI giving you the opportunity to retarget this user with a coupon, to help inspire repeat business. Or, looking to the future, using the emotional detection and content delivery mechanism that Facebook is working on – to deliver specific content to your audience members based on their moods. There are lots of ways to capitalize on an audience mood – from buying things based on something you’re celebrating (adopting a pet, having a baby), of if you’re feeling a little down (retail therapy).
5. Surfacing the most important conversations
Say what, now? Well, this is best described as acting like a spam filter for social media. Lots of conversations happen, but how many of them are genuine conversations that deserve response? And how many of them are just posts from bots? Who’s looking for help, and who’s entering your contest? And AI can help you prioritize those legitimate users. But AI cannot help you actually respond to them – it’s still our job to do what we do best: be human.
AI can be a tremendous tool for social media marketing. Pair it with good ol’ human-powered innovation and creativity, and you’ve got a powerhouse ready to work for you. The possibilities are endless.
Have you started using AI to boost your social media efforts? Tell us about it in the comments below.