Latest / Social Media Marketing with Fexingo: Instagram, TikTok, LinkedIn, and Platform Strategy / Why Brands Are Using AI Avatars to Sell on Social Media
Transcript
- Lucas: So Luna, I was scrolling through my feed this morning, and I stopped dead on this video from a skincare brand I actually follow. The presenter had perfect skin, obviously, but something felt... off. And then I realized — it wasn't a person. It was an AI avatar. Luna: Oh, I've seen those. They're getting creepy good. But are we talking about like, deepfake-style, or the animated ones that are clearly not human? Lucas: That's the thing — the line is blurring. This one was photorealistic. I had to look twice. And it got me thinking: brands are really leaning into AI avatars as presenters, spokespeople, even customer service reps on social. So I dug into the numbers, and it's not just a gimmick anymore. Luna: Okay, I need a concrete example. Because I keep hearing 'AI influencer' and I picture like, that one CGI model from years ago. What's actually happening now? Lucas: So there's a beauty brand — I won't name them because I'm not a hundred percent on their current status — but last quarter they rolled out an AI avatar as their primary TikTok presenter. And they reported a thirty percent lift in engagement compared to their human-hosted videos. Thirty percent, Luna. That's not a rounding error. Luna: Thirty percent is huge. But wait — how much of that is just novelty? People clicking because it's weird, not because they're actually convinced. Lucas: That's exactly the question their social team had to answer. So they ran a split test: same script, same editing, one version with the AI avatar, one with a human creator. The AI version outperformed on watch time and shares. But — and here's the kicker — the human version crushed it in the comments. People were like 'I love that she's real' and tagging friends. The AI version got a lot of views but almost no conversation. Luna: So it's not a straight win. It's engagement versus community. That's a trade-off a lot of brands might not think about. Lucas: Right. And that's the nuance I want to dig into today. Because AI avatars are becoming a standard tool in the social media toolbox, but they're not a replacement for human creators. They're more like a supplement, and the smart brands are figuring out where they fit. Luna: Let's start with the 'why.' Why would a brand choose an AI avatar over a human? I mean, cost is obvious — you don't pay a salary or deal with scheduling. But is that the main driver? Lucas: Cost is part of it, sure. But the bigger driver is control. With an AI avatar, you control every word, every expression, every pixel. There's no off-script moment, no scandal risk, no 'I accidentally posted something offensive on my personal account.' And you can produce content at scale — one avatar can pump out a hundred videos a day without getting tired. Luna: But doesn't that make it... sterile? I feel like the whole point of social media is that it's messy and human. That's what builds trust. Lucas: Trust is exactly the battleground. And this is where it gets interesting. Because there's a growing segment of consumers — especially Gen Z — who actually don't care who or what is presenting, as long as the content is useful or entertaining. They're so used to filters and CGI in movies that an AI avatar isn't automatically a red flag. Luna: Really? I feel like I'd be able to tell. Maybe not every time, but... is there data on how often people can spot the difference? Lucas: There have been a few studies this year, and the results are all over the place. One survey from a marketing research firm found that about forty percent of respondents couldn't tell an AI avatar from a human in a side-by-side test. But another study — more academic — found that people are better at spotting them when they're paying attention, especially in longer videos. So it's not black and white. Luna: So if people can't always tell, is there an ethical obligation for brands to disclose? I feel like that's the swampy area. Lucas: It's not just ethical, it's becoming regulatory. The FTC has been drafting guidelines on ai generated content in advertising, and I think by early next year we'll see clear rules. Some platforms are already ahead of it — TikTok requires ai generated content to be labeled, and Instagram has a similar policy. But the enforcement is spotty. Luna: And what happens when a brand doesn't disclose and gets caught? I'm thinking of that time a few years ago when a company used a CGI influencer and people felt duped. Did it actually hurt them? Lucas: That's the cautionary tale. That brand faced a pretty serious backlash — not just in the comments, but in the press. They had to issue a public apology and pull the campaign. And their follow-up engagement dropped for months. So the risk is real. But the brands doing it well are being upfront. They're labeling the content, and some are even making the AI avatar a character with a backstory. Luna: A backstory? Like, they're giving the avatar a personality? That's kind of fascinating. Can you give me an example of that working? Lucas: There's a beverage brand — I'm thinking of one that launched an AI avatar named 'Ava' — she's supposed to be a 'digital native' who loves hiking and iced coffee. She has her own Instagram stories, she responds to comments, and she even did a 'takeover' of the brand's TikTok for a day. It sounds silly, but it actually humanizes the avatar. People engage with her because she's a character, not just a salesperson. Luna: Okay, so they're building a narrative. But that requires a lot of content. How are brands actually producing this stuff? Is it like, they film a person once and then deepfake them, or is it fully computer-generated? Lucas: Both, actually. The high-end approach is using a real person's likeness — they license the face and voice, and then the AI animates it. That's what a lot of beauty brands are doing with their founders. But the fully synthetic avatars — where the face is generated from scratch — that's becoming more accessible too. There are platforms now where you type in a script, pick an avatar, and it generates a video in minutes. Luna: And what does that cost? I've heard it's not as expensive as people think. Lucas: The price has dropped dramatically. You can get a basic avatar video for as little as a few hundred dollars a month, depending on the platform and usage. The custom, high-fidelity avatars that look like a real person — those are more in the tens of thousands to set up. But compared to paying a human creator for a full campaign, it can be a fraction of the cost. Luna: So it's not just a big brand play. Small businesses could actually use this too. Like, a local restaurant could have an AI avatar do their daily specials video. That's kind of wild. Lucas: Exactly. And that's where I think the real growth is coming. But there's a catch: the avatar needs to fit the brand's voice. If you're a fun, irreverent brand and your avatar sounds like a robot reading a teleprompter, it's going to flop. The successful implementations are the ones where the avatar's personality is carefully crafted to match the brand. Luna: Let's talk about platform differences. We've talked TikTok and Instagram, but is there any use for AI avatars on LinkedIn? I feel like that's a different vibe. Lucas: LinkedIn is actually a really interesting case. I've seen a few B2B companies use AI avatars for thought leadership videos — like a 'fake CEO' who shares insights on industry trends. And it seems to work, because the content is more informative than personality-driven. People are on LinkedIn to learn, not to connect with a face per se. So an avatar that delivers solid advice can be just as effective as a human. Luna: But isn't that a little bit... deceptive? If you're presenting an AI as a thought leader, and people think it's a real person, that feels like a trust breach. Lucas: That's a valid concern, and it's exactly why disclosure matters. The brands doing it ethically have 'ai generated' in the caption or a disclaimer in the video. But I've also seen some that don't, and that's risky. If a follower discovers the deception, the brand's credibility takes a hit. Luna: So what's the future? Are we heading to a world where half the influencers on my feed are digital? Or is this a bubble that's going to burst? Lucas: I don't think it's a bubble, but I do think we'll see a correction. The novelty will wear off, and then it comes down to performance. If AI avatars can deliver consistent ROI — and for a lot of brands, they are — they're here to stay. But the human element isn't going away. The real magic is in the blend: AI for scale and consistency, humans for authenticity and community. Luna: That makes sense. And I guess the brands that win are the ones that are transparent about it. Like, 'hey, this is our AI avatar, but our customer service team is real people.' Lucas: Exactly. And that's actually a great way to think about it. The avatar is like a new tool in the social media kit, not a replacement for the human connection. The brands that treat it that way are the ones seeing the best results. Luna: I think this is one of those episodes where we're going to get a lot of comments from listeners who have tried this or have strong opinions. I mean, it's a polarizing topic. Lucas: Definitely. And I'd love to hear from our listeners — if you've experimented with AI avatars, what's been your experience? Drop a comment or tag us. Luna: Before we wrap up, I just want to say — if this conversation has sparked something you might actually try, or if you've been enjoying these deep dives into social media trends, the best way to make sure they keep coming is to support the show. You can buy us a coffee at buy me a coffee dot com slash fexingo. Lucas: Yeah, that really helps us keep the podcast ad-free and independent. It takes two minutes, and it goes directly to production costs. No pressure — we're just grateful you're listening. Luna: And speaking of trying things — Lucas, you mentioned you were thinking about testing an AI avatar for one of your side projects. You still planning that? Lucas: I am, actually. I'm curious to see how it performs against my own videos. But I'm also a little wary — I don't want to become a robot spokesperson. I think I'll stick to the human side for now, at least until the uncanny valley gets a little less eerie. Luna: Ha, fair enough. But it's worth keeping an eye on — I feel like in a year, this might be as normal as using a green screen. Lucas: No doubt. And with that, we'll leave it there. Thanks for tuning in, everyone. Until next time.