Let’s be honest about something. Before Monday’s close, SNAP had been one of the uglier charts in social media all year. The stock opened 2026 near $8.13 and spent most of the first half grinding toward its 52-week low of $3.81. Then Q2 earnings dropped after the bell on August 3rd, and the stock jumped roughly 17% from its prior close. That is a big move. The question worth sitting with is whether the business actually earned it.

Here’s what the numbers say. Revenue came in at $1.599 billion, up 19% year-over-year and roughly $60 million ahead of the $1.53 billion consensus. The adjusted loss per share was $0.10, beating the $0.12 estimate. Those are clean beats, not squeakers. But the number that really got attention was adjusted EBITDA, which surged to approximately $250 million from just $41 million in the same period a year ago. That kind of year-over-year margin expansion is not something Snap has historically been able to show. It got attention for good reason.

What drove it? Mostly cost discipline meeting AI-powered ad tools at the right moment. Total adjusted costs grew just 4% year-over-year while revenue ran at 19%. Gross margin expanded 7 percentage points to 58%. AI-driven advertising products, including Smart Campaign Solutions, delivered a 56% increase in platform conversions. Dynamic Product Ads revenue grew 43%. Cost per purchase fell 18%. App purchase volume jumped 128%. These are advertiser ROI numbers that hold budgets in place regardless of what else is going on in the macro.

Worth pausing on the subscription line, because it has quietly become a real business. The “Other Revenue” category, which includes Snapchat+ as well as Memories Storage and the newer Lens+ tier, rose 85% year-over-year to $316 million in Q2. Annualize that and you are looking at over $1.2 billion in non-advertising revenue. A year ago, almost no one was modeling subscriptions as a meaningful Snap revenue driver. That story has changed faster than most investors have updated their models.

Free cash flow hit $121 million for the quarter, marking the eighth consecutive period of positive FCF. Operating cash flow was $176 million. These are not breakout numbers in isolation, but for a company that spent years burning cash without a clear path to profitability, the consistency of positive cash generation matters. Management is now targeting sustained positive net income starting in 2027.

Now the part that keeps the bull case complicated.

North America DAU fell 7% year-over-year to 92 million and was flat sequentially. Europe followed a similar trajectory. Global DAU reached 493 million, up 5% from a year earlier, but that growth is almost entirely coming from regions where Snap’s ad infrastructure is underdeveloped. The platform earns significantly less revenue per user in those markets. Global ARPU for the quarter came in at $3.25, above the $3.16 analyst estimate, but that number is being carried by North America monetization even as North American users leave. North America revenue was $943 million, up 15% year-over-year despite flat user count. The monetization improvement is real. The user trend is not solved.

Slight tangent, but it matters here: North America DAU growth was concentrated entirely in users 35 and older. That is not inherently bad, but it is a meaningful shift in the platform’s identity. Snapchat has always been built on its grip over younger users. If that core skews older and regulators start applying tighter age assurance rules across multiple markets, the product decisions available to Snap narrow considerably. CEO Evan Spiegel acknowledged on the earnings call that evolving rules around youth safety, privacy, and online safety requirements could affect product experiences or user growth over time.

The regulatory exposure is real and growing. Australia has moved toward tougher under-16 social media restrictions. Multiple U.S. jurisdictions have active litigation. Snap settled at least one social media addiction case ahead of trial in late July. These issues do not show up in the EBITDA line today, but they shape what the product can look like in 2027 and beyond.

What Q3 Guidance Actually Says

Management guided Q3 revenue to $1.70 billion to $1.74 billion, with the midpoint slightly above the $1.70 billion consensus. Adjusted EBITDA guidance is $300 million to $350 million, against a prior estimate of roughly $330 million. On the surface, that looks solid. But Spiegel specifically called out World Cup-related advertising spending as a contributor to Q2’s strength. That spending does not repeat in Q3. Management says Q3 revenue growth is expected to match Q2’s 19% pace despite the tougher comparison, which implies the underlying business is accelerating. That is either an encouraging sign or aggressive framing. Probably worth waiting to see the Q3 number before committing to one interpretation.

One more variable: the hardware story. Snap is planning a commercial launch for its augmented reality Specs glasses this fall, with a launch event scheduled for September 16 in Los Angeles. The glasses carry a $2,195 price point. Whether a $2,195 consumer AR device moves meaningful revenue in 2026 is genuinely debatable. What it does do is signal that Snap is trying to build a computing platform identity, not just maintain a messaging app. The AR space is crowded and expensive. But Snap is further along in the hardware development cycle than most people realize, and being early in consumer AR has some value if the category ever breaks through.

Here is where I land. The business is genuinely better than it was twelve months ago. The margin improvement is real, the subscription diversification is real, and the advertiser ROI metrics are the kind of numbers that attract budget. At the same time, the stock is still trading well below its 200-day moving average near $5.94, and the North America user problem has not been reversed, only stabilized. A relief rally off a strong quarter is warranted. Whether this is the start of a durable recovery or a bounce inside a longer downtrend probably gets answered when Q3 arrives in October, without a World Cup to help.

Both things can be true at once. Sometimes that is exactly where the trade lives.