GYoung Digital Blog

A little while ago I did some reading about how AI is going to affect social media community management.

Social media community management is basically where the brand behaves as another social media, or digital user that participates in online conversations with its customer base. When there’s a conversation on Twitter, or Threads, or Bluesky, or Reddit, the brand will be there that contributes to that conversation as well as taking note of any hot topics that arise in that conversation.

There was a time when I worked in community management. The first role in community management I had was with Microsoft many years ago, a short contract role that turned into something of a career for a few years which eventuallly took me from Seattle all the way to New York City where I joined WPP, specializing in social listening, analytics measurement, and general media monitoring, all of which are pretty much advanced community management skills.

With the recent spike of interest and adaptation of AI, I’ve been thinking about how AI might affect community management. I’ve already spoken to agencies in the past that have talked about how they’re going to use AI in community management. At the time, I was probably dismissive of AI, looking at it in the same way that I looked at the metaverse, of crypto, or NFTs, or whatnot, speaking of AI as a cute little thing that is worth entertaining but not worth taking so seriously as to worry about or current industry shifting in any profound way.

Boy, how naive was I?

AI seems to be injecting itself in all areas of marketing and communications. I still find myself to be a bit resistant to AI, thinking that it’s not nearly going to be as useful as some might try to lead us to believe that it will be. However, my resistance is showing that I am a big dumb dummy that’s not keeping up with the times, as the entire industry that I work in seems to have adjusted itself so drastically that giant companies have merges to make one supermassive company, and so many jobs have been cut in (the apparent) preparation that AI will lead the way. Or, to put it another way, many see AI as a way to save money for companies as well as work more efficiently.

So I was wondering, how much more efficient community management can be?

Turns out, quite a bit. Though, not completely.

It actually makes sense what AI can do for community management. The easiest and most obvious is one of the most time consuming and repetitive tasks that community managers need to do, and that likely means responding to the most common questions and comments that users leave for brands. Depending on the brand, such as a SaaS company, users might as you the exact same question over and over, and the community manager will give the same answer over and over.

Suppose there’s a cybersecurity company that uses an app that checks on your computer’s passwords and makes sure nobody suyspicious us trying to log into your computer, but every day there are hundreds of users who sem to have the same issue with their app, that it’s suddenly not working, or there is some weird issue that’s not letting the app work the way it should. It turns out that the best way to fix this issue is to delete the app and reinstall it, nearly every single time you tell that to a user, the problem is solves.

Imagine responding to a hundred comments every day with that exact same comment?

Granted, that is a bit of an extreme example, though I image there are many companies out there that probably deal with similar issues.

Then, there’s the matter of identifying posts that are worth resharing. There could be so many influencers talking about your brand, posting content that’s getting a lot of engagements.

Basically, depending on the volume of posts that a community challenger has to work with, the community manager would spend great deals of time just answering questions, and in a repetitive manner, taking away the time that the community manager could take in developing strategies and finding ways to bring together the audience.

That’s basically the main purpose of AI in community management; It will take those super boring and repetitive tasks, like common customer service, and answer those questions automatically, allowing the community manager to spend more time developing the strategy and developing more content ideas for the audience.

Though, there is a very funny caveat that I find when it comes to implementing AI into community management. While it will likely take over those boring tasks away, there’s still the likelyhood that one will need to monitor the responses that AI makes in order to make sure they’re aligned with the brand voice, and making sure that they’re responding accurately. AI is not sentient (despite what many people might say), but it works on the rules and precedence that’s set before it. After all, AI still makes mistakes. Community manager will need to monitor the responses from AI to make sure they’re accurate, and fix or rewrite when necessary.

The community manager will still need to monitor AI responses, thus creating another boring, repetitive tasks to do. It kind of makes you wonder what’s the point of implementing AI at all?

But I digress. In the end, it looks like that AI will certainly make community more efficient in the end, but community management is a very human, personable task, and there will be those complex, more nuanced situations that will require a person to take control from AI and respond to the person directly.

It would seem to me that many brands expect community management to become a glorified chatbot, where social media will respond to users the same way that a chatbot will respond to users on the bank app when asking questions about their account. Though, it will be very interesting to see how AI evolves when responding to users on social media. Social media cannot control the direction of comments that come in. On chatbots for bank apps, for example, the chatbot that take complete control of the conversation. However, those conversations on social media are much more unpredictable and much more fluid, companies are unable to control the conversation. There will be posts that are sarcastic, and the chatbot will take them all at face value.

So, I do see some use for AI in social media community management, but social media conversations are so fluid that it would still be wise for brands to employ a person to monitor the conversations that come in. It would be simple to apply rules to social management tools like Sprinkr to filter certain conversations into a certain comment stream and allow AI to respond to those comments automatically, while other nuanced comments will be pulled into other comment streams that the community manager will respond to manually and individually to each comment.