Snapchat just announced that it's turning its ads into chatbots. Let that sit for a second.

The company revealed this week that it's rolling out conversational advertising powered by its AI infrastructure — meaning brands will be able to deploy ads inside Snapchat that users can literally talk to. Not click through. Not swipe past. Talk to. This is either the most interesting thing to happen to mobile advertising in years, or the fastest way to make an entire generation of users delete an app. Possibly both.

Introduction

Snapchat's conversational AI advertising is the story everyone in the digital marketing world is searching for right now, and for good reason. Snap Inc. has been in a slow-burn financial crisis for the better part of three years — revenue growth has stalled, advertiser confidence wobbled badly after Apple's iOS 14.5 privacy changes in 2021 gutted its targeting capabilities, and the stock has shed more than 80% of its value from its all-time high of around $83 in September 2021. The company desperately needs a new monetization story to tell Wall Street.

This is that story. Snap is betting that by embedding conversational AI directly into the ad unit itself, it can offer brands something no other platform currently does at scale: a two-way dialogue with a potential customer, inside the app, without ever leaving the feed. The pitch to advertisers is that instead of a static banner or a skippable video, their brand gets to respond to a user's questions, objections, and curiosity in real time.

Here's what we're going to break down: what Snap actually announced, why they're doing it now, what it means for users and advertisers, and whether this is a genuine product innovation or a repackaged hype cycle with a Snap-branded ghost on top.

Here's What's Actually Happening

Snap's conversational ads work by embedding a chat interface directly into an ad unit. A brand — say, a cosmetics company or a sneaker retailer — can deploy an AI-powered assistant that appears within the ad itself. A user sees the ad, can type a question like "does this come in wide fit?" or "what's the return policy?", and the AI responds instantly, without the user being redirected to a website or a customer service portal.

The technology runs on top of Snap's existing My AI infrastructure, which the company launched in February 2023 as a ChatGPT-style assistant built into the app. My AI accumulated over 200 million users in its first year, which gave Snap a large dataset of conversational patterns to train against — and, more importantly, a proof point that its user base would actually engage with in-app chat interfaces.

The announced rollout is initially targeting select brand partners in the U.S., with a broader expansion expected throughout 2025. Pricing hasn't been made fully public, but industry sources familiar with Snap's ad sales structure suggest these units will carry a significant premium over standard Snap ads, which currently average around $2.95 CPM (cost per thousand impressions) according to 2024 benchmarks from Influencer Marketing Hub.

Why Snap Is Doing This Now — And Why It Had To

Before Apple dropped the ATT (App Tracking Transparency) framework in April 2021, Snap was generating strong ad revenue growth — 116% year-over-year in Q2 2021 — largely because it could track user behavior across apps and serve highly targeted ads. Then Apple gave users the ability to opt out of cross-app tracking, and roughly 85% of iOS users did exactly that.

The result was brutal. Snap's Q3 2022 revenue came in at $1.13 billion, essentially flat year-over-year, and the stock dropped 25% in a single day after earnings. The company laid off 20% of its workforce — about 1,200 employees — in August 2022. It has been searching for a durable ad product ever since.

Conversational advertising is Snap's answer to a structural problem: if you can't track users across the web to build profiles, make the ad itself smart enough to gather intent signals in real time. (The company calls this approach "contextual engagement." What it actually does is ask users to voluntarily tell an ad what they're interested in, which is either brilliant or deeply optimistic, depending on your read of human behavior.)

CEO Evan Spiegel framed the launch in terms of "deepening brand relationships" during a recent investor briefing. The more honest framing: Snap needs advertisers to pay more per impression, and the only way to justify higher CPMs is to demonstrate measurably better outcomes than a static display ad can deliver.

What Advertisers Are Actually Getting

A New Kind of First-Party Data

This is the part that's genuinely interesting, and that the press release buries under marketing language. When a user types a question to a conversational ad, that interaction is a declared intent signal. The user isn't being inferred from browsing behavior — they're explicitly telling the brand what they want to know.

For advertisers, that's extremely valuable. A traditional ad impression tells you someone saw your ad. A conversational interaction tells you someone asked about sizing, or price, or a specific product feature. That's the difference between a page view and a sales conversation.

Brands like Taco Bell, Amazon, and various DTC skincare companies have already experimented with conversational commerce on platforms like WhatsApp and Instagram DMs. Snap is essentially building that capability natively into the ad unit, which removes friction and keeps the user inside the Snapchat environment.

The Measurement Problem Isn't Solved

Critics — and there are plenty of them in the performance marketing world — will point out that conversational engagement doesn't automatically translate to conversions. Someone asking an AI chatbot "what colors does this come in" isn't the same as someone clicking "add to cart."

That framing misses the point, but only partially. Snap has always struggled with lower-funnel attribution, and a conversational ad unit doesn't fix that. What it potentially does is improve upper- and mid-funnel engagement metrics — brand recall, consideration, time spent — which are the metrics that large brand advertisers (the ones with big budgets) actually care about.

The performance marketers running $500-a-day campaigns for direct-response e-commerce? They'll want to see click-through and purchase data before they move budget here. That data doesn't exist yet.

What This Means for Snapchat's 850 Million Users

Snap reports 850 million monthly active users as of its most recent earnings call, with daily active users at 422 million — a number that has grown steadily even as revenue has struggled. The core demographic skews young: roughly 75% of 13-to-34-year-olds in the U.S., UK, France, and Australia use Snapchat, according to the company's own 2024 data.

That demographic is also, famously, the one most allergic to advertising. Gen Z users have grown up with ad blockers, skip buttons, and a finely tuned radar for anything that feels like marketing. Asking them to chat with an ad is a significant behavioral ask.

Is this a problem? Depends on who you ask.

Snap will argue that opt-in conversational engagement is less intrusive than a forced video preroll — and they're not wrong. If a user chooses to ask an AI ad a question, the interaction is voluntary and potentially useful. That's a fundamentally different dynamic than an unskippable 15-second ad for a product you'll never buy.

The counterargument is that even well-intentioned AI ad interactions carry a manipulation risk. These systems are optimized to keep users engaged and move them toward a purchase decision. The line between "helpful brand assistant" and "persuasion engine" is thinner than Snap's marketing materials suggest.

How This Fits Into the Broader Conversational Commerce War

Snap isn't operating in a vacuum here. Meta has been quietly building conversational ad capabilities into WhatsApp Business for two years, and Instagram's DM-based commerce features have processed billions of dollars in transactions. Google launched its own conversational ad formats in Search in late 2023, allowing users to refine product queries in real time. Even Pinterest has tested AI-powered shopping assistants embedded in its ad units.

The race to conversational commerce is real, and Snap is entering it later than some of its competitors. But it has one meaningful structural advantage: the camera.

Snapchat's camera-first interface means conversational ads could eventually incorporate visual inputs — a user points their camera at a pair of shoes and asks the ad "do you have something like this in blue?" That's a use case none of Snap's competitors can replicate at the same level of native integration. Snap has been developing its AR (augmented reality) shopping tools since at least 2020, and the convergence of AR try-on, visual search, and conversational AI is where this product roadmap is clearly headed.

For more on how Snap's hardware and software bets are stacking up against the broader Apple ecosystem, check out our piece on Apple's New Hardware Chief Has a Lot to Prove — the competitive dynamics in the camera and AR space are directly relevant.

The My AI Precedent — And What It Tells Us

When Snap launched My AI in February 2023, the reaction was... mixed. Users complained that it appeared at the top of their chat feed without being asked. There were widely-reported incidents of the chatbot giving inappropriate responses to teenagers. Snap had to issue multiple updates and clarifications within the first few months.

But the usage numbers didn't collapse. My AI reached 200 million users. People kept using it. That resilience tells you something important about Snap's user base: they're more willing to experiment with AI features than the backlash headlines suggested.

The My AI rollout also gave Snap's engineering teams roughly two years of real-world data on how its users interact with conversational interfaces — what they ask, how they phrase things, where conversations drop off. That's not nothing. That's the training data that makes conversational ad units plausible rather than theoretical.

Should Advertisers Actually Try This?

If you're a brand marketer reading this trying to figure out whether to pilot Snap's conversational ads, here's the honest answer: it depends entirely on your goal and your audience.

• Brand awareness campaigns targeting 18-to-28-year-olds: Yes, worth testing. The engagement metrics will be genuinely interesting, and the CPM premium is justifiable if your audience skews Snap-heavy.
• Direct-response e-commerce trying to drive immediate purchases: Wait for six months of third-party performance data before committing significant budget. The attribution story isn't there yet.
• Product categories with high consideration cycles — cars, financial products, travel, consumer electronics: This format was basically designed for you. A user who will spend three weeks researching a purchase decision is exactly the kind of person who will ask an ad a question.

The one thing I'd push back on is the framing that this is a silver bullet for Snap's revenue problems. It isn't. It's a smart product bet that could meaningfully improve advertiser ROI for specific use cases — but it won't reverse three years of structural ad revenue challenges in a single product cycle.

The Bottom Line

Snap's conversational advertising launch is legitimately interesting — more interesting than most ad product announcements, which tend to be incremental feature updates dressed up in breathless language. The underlying idea, that an ad unit should be able to answer questions rather than just make assertions, is sound. The execution, building on top of My AI's existing infrastructure and Snap's camera-first UX, is coherent.

But the history of "revolutionary" ad formats is littered with products that made sense on a whiteboard and failed in the wild because users simply didn't want to interact with them the way the product team imagined. Remember when QR codes were going to transform advertising in 2012? Or when Facebook Canvas ads were going to replace mobile websites? The graveyard is full of clever ideas that underestimated user indifference.

Here's my actual take: Snap's conversational ads will work better than the skeptics expect and worse than the press release implies. The product will find a genuine audience among brand advertisers targeting young consumers in high-consideration categories, and it will generate enough positive case studies to sustain the format. What it won't do is solve Snap's core problem, which is that it needs to either grow its user base significantly beyond its current 422 million daily actives or find a way to extract meaningfully more revenue per user. Conversational ads can help with the latter. But at $2.95 average CPMs, the math still requires a lot of conversations. And not everyone wants to chat with an ad.