Subscribe
Learn Library

One Virtual Influencer Made $2.5 Million in a Year

The article examines the booming virtual-influencer industry, where AI-generated personas like Lu do Magalu earn millions while delivering engagement rates three times higher than human creators. It also highlights the parallel rise of AI-synthetic fraud and tightening global regulation.

influencerai-marketingavatarevidence
2026-08-02Go Next Marketer9 min read

A while back, I was scrolling through Instagram when I stumbled on an account called Lu do Magalu.

82 million followers. One shoppable post every day. Brands lined up to work with her.

I tapped through a few posts. Genuinely polished stuff. But the more I watched, the more something seemed wrong—

This girl is not a real person.

She's a virtual persona created by Brazilian retailer Magazine Luiza. A digital human, drawn by a CGI artist.

In 2024 alone, this virtual influencer pulled in roughly $2.5 million from 74 sponsored posts. That works out to about $34,000 per post.

What does that look like? An average real-life creator earns about one-fortieth of her yearly total from an entire year of sponsored content.

Mind-blowing.

I had to know: what is actually going on here? How can someone who doesn't exist siphon off this much real marketing budget?

So I dug through a pile of data. Turns out, Lu do Magalu is not an isolated case. An entirely new marketing category has quietly ballooned into the hundreds of billions.

A Business That Barely Existed — and It's Already Worth $11.74 Billion

Let's start with a number.

In 2026, the global virtual-influencer market is worth $11.74 billion.

What does that mean? Here's some perspective: in 2023, it was just $4.6 billion. In three years, it more than doubled.

And it's not stopping there. Research firm The Business Research Company projects that by 2032, this market will hit $154.6 billion. A CAGR (compound annual growth rate) of 41.29%.

41% annualized growth. Name one traditional marketing channel that delivers numbers like that.

Virtual influencer market growth: $4.6B (2023) → $11.74B (2026) → $154.6B (2032), CAGR 41.29%

The broader influencer-marketing pie (including real human creators) saw $32.55 billion in global spending in 2025. In other words, virtual personas have already eaten a substantial slice.

Why is it growing so fast?

Plain and simple, it comes down to one thing: virtual personas are cheap, obedient, and they work 24/7.

What Does "Cheap" Even Mean?

Let me break down the math.

A top-tier human creator? One piece of content with them starts at $20,000 to $30,000. Plus exclusivity clauses, scheduling headaches, and the constant worry that they'll say something stupid in public and drag your brand down with them.

A virtual influencer?

Building an enterprise-grade virtual persona in 2025 costs an average of $1.4 million. Sounds like a lot, right? But that's a one-time investment. Once the character is built, she can post every day — 365 posts a year, no problem. Marginal cost is essentially zero.

Lu do Magalu's per-post rate is indeed around $34,000, on par with top-tier human creators. But she can post every single day. Ask a top human creator to pump out daily content and they'll burn out — and you'll be paying more for every extra post.

So here's how brands do the math: pay a chunk of development cost upfront, then it's basically a printing press running on autopilot.

In 2023, the average enterprise investment in virtual-persona development was $720,000. By 2025 it jumped to $1.4 million — a 94% increase in two years.

Why the jump? Because the brands that ran the numbers are doubling down.

What Does "Obedient" Mean?

This is where virtual personas become invaluable.

She will never get caught in a scandal. She will never have a public meltdown. She will never fire off a tweet at 2 a.m. the night before your biggest sale that sends your PR team into crisis mode.

Good luck negotiating that with a human creator.

Here's another data point: brand adoption of virtual influencers jumped from 60% to 73%. Note — that's a 2026 figure, a 13-point increase in a single year.

Which industries are going hardest?

  • Beauty and personal care: 89% of brands are using them.
  • Fashion: roughly 78%.
  • Gaming: 76%.
  • Consumer electronics: 64%.

Why is beauty so far ahead? Because beauty is a visual-driven category, and a virtual persona's face can be tuned to any state you want — never cakey foundation, never an oily shine, always in perfect lighting. No human model can deliver that level of consistency.

And who's the most conservative?

Financial services: 22%. Healthcare and pharma: 9%.

That makes sense. When it comes to managing money or your health, consumers want "real." Show them an algorithm-drawn persona telling them to buy a mutual fund, and they won't buy it — literally.

Three Numbers That Settle Whether Virtual Influencers Actually Work

Market size alone doesn't tell the whole story. When brands spend money, the ultimate question is: does it actually work?

I picked three data points that speak loudest.

One: engagement rate.

HypeAuditor's panel data shows that in 2026, the average engagement rate for virtual influencers is 5.67%. For human creators over the same period: 1.89%.

Virtual personas clock 3 times the engagement of real people.

Why? Think about it — an impossibly perfect persona putting out polished content every day. Fans are chasing an idealized character. That kind of devotion is naturally intense.

But here's the interesting part: in categories that really live or die on "authenticity," real humans can flip the script and outperform virtual personas. Parenting, personal finance — anything where the audience needs to feel "I'm really living this, I'm really using this" — human-led content can beat AI personas by 2.7x.

Two: conversion rate.

About 58% of U.S. consumers follow at least one virtual influencer. That sounds like a lot.

But who actually opens their wallet? Among Gen Z (roughly, those born after 1997), about 35% say they've bought something because a virtual influencer recommended it.

One in three. For a brand-new marketing format, that's genuinely impressive.

Three: brand-side satisfaction.

Among brand managers who've used virtual influencers, roughly half rated the experience "very positive."

Even more telling is where CMOs (Chief Marketing Officers) are putting their money. By 2026, CMOs are projected to allocate up to 30% of their influencer-marketing budgets to virtual personas.

And the ones going hardest — what's their payoff? Brands that invest more than 25% of their budget in virtual influencers see ROI 41% higher than those investing under 10%.

When data like that drops, everyone follows. Of course they do.

And Now, the Bad News

Everything has a flip side. Virtual influencers blew up — and scammers blew up right alongside them.

In 2026, estimated global losses from influencer-marketing-related fraud hit $4.8 billion.

Of that, $2.1 billion came from "AI-synthetic fraud" — using AI-generated fake faces, fake identities, and fake content to swindle people.

Notice the shift: AI-synthetic fraud has now overtaken traditional bot-driven engagement fraud as the biggest category.

Same tech, two outcomes — upside (5.67% engagement, $11.74B market) vs downside ($4.8B fraud losses, $2.1B AI-synthetic, $23.7B deepfake)

Why? Because today, using AI tools to fabricate a fake influencer and pump out fake content costs less than $50.

Fifty dollars to launch a scam, leveraged into a fraud economy worth hundreds of millions.

Sumsub's data adds an even wilder layer: deepfake-driven losses tied to influencer-marketing scams are projected to reach $23.7 billion globally in 2026 — a figure that spans investment scams, romance scams, brand-impersonation schemes, and more.

About 74% of deepfake fraud content is produced using tools that cost under $50.

Here's the awkward part.

The marketing world is pouring money into building virtual influencers, while the fraud economy uses the same technology, cheaper, to scam consumers. It's the exact same tech. The only difference is what the person behind it wants to do.

A Deeply Uncomfortable Truth

There's one data point here that's especially worth chewing on.

Among marketing teams, 36.67% use AI for creator discovery. 21.11% use AI for content generation.

But those using AI for anti-fraud? Only 7.22%.

In other words, brands enthusiastically use AI to find influencers and write copy — but almost never use it to check whether those influencers are fake.

Why?

I thought about it, and here's my read: finding influencers is spending money. Spending-money tasks that get AI-efficiency boosts are visible to the boss and trackable against KPIs. But catching fraud is saving money. Saving-money tasks don't show up on anyone's performance metrics, and procurement still has to run through manual audits.

It's a real trade-off. It's just that this particular trade-off carries a hefty price tag — $4.8 billion a year, sitting right there.

Regulation Has Caught Up

This time, the Americans moved fast.

On October 21, 2024, the U.S. FTC (Federal Trade Commission) finalized its rule banning fake and AI-generated consumer reviews.

What does that mean? AI-generated celebrity endorsements, fabricated user reviews, undisclosed AI-influencer recommendations — all of it is now illegal. Each violation carries fines up to $51,000.

The EU's AI Act followed suit, mandating disclosure at the member-state level. The UK's FCA (Financial Conduct Authority) cracked down on financial influencers too.

The impact was immediate. In 2026, the FTC and UK FCA jointly investigated 2,340 creators. Influencer-promoted investment scams surged 47% year-over-year.

Consumers lost $1.9 billion to influencer-recommended investment scams.

$1.9 billion.

So, How Should We Read All This?

By now, you've probably noticed the pattern. Virtual influencers are a double-edged sword — sharp enough to be terrifying.

The upside: engagement rates 3x higher than human creators, marginal cost approaching zero, maximum brand control. It's one of the biggest disruptions to hit marketing in the past decade.

The downside: the fraud economy wields the same technology to siphon off hundreds of billions every year. Consumer trust in "is this recommendation real or fake?" is being burned through at terrifying speed. And the regulatory hammer has already fallen.

My take: virtual influencers aren't going away. They'll become standard equipment for top-tier brands. But the next few years won't be won on "can you build a virtual persona." They'll be won on "can you do compliance."

The brands that wrote third-party anti-fraud audits into their contracts early (this share jumped from 19% to 54% in 2026), the ones that honestly disclose "this is an AI persona," the ones that treat virtual personas as long-term assets rather than one-off stunts — they'll win.

And the players who just want to ride the traffic wave, who treat AI as a cheap tool, who half-ass disclosures and audits — they'll most likely take their own brand down with them the moment regulation tightens.


This is the full picture of the virtual-influencer landscape as I see it, at the midpoint of 2026.

It's no longer a "future trend." It's already happened. Lu do Magalu's $2.5 million year is all the proof you need.

But like anything new — whether it's worth doing doesn't depend on how hot it is. It depends on whether you can handle what comes with it.

And here's my wish for you: while you embrace the new technology, may you also hold the line that's hard to hold.