Snap Stock Analysis: A 92.6% Crash, Shrinking at Home and a $2,195 Bet on Glasses

Snapchat is not a failed product. Nearly a billion people open it every month, it invented the Stories format that every social network copied, and its revenue grew 18.9% in the second quarter of 2026. Snap the company is another matter. On 16 October 2026 its shares closed at $6.13, 92.6% below their 2021 peak and 63.9% below the $17 price at which they were sold to the public in 2017. Snap has not had a single profitable year as a public company.

This article follows the argument of the video: the crash and where the valuation stands today, the hype that preceded it, what is genuinely going well, the financial system that keeps Snap in the red, the users it is losing and the ones it is gaining, the infrastructure it rents, its strategic moves and its AI efforts, and finally Specs, a glasses bet with an uncomfortable precedent. Market figures are anchored to the 16 October 2026 close; financial figures run to the second quarter of 2026 unless another date is given.

Snap's Epic Crash

Snap's highest close was $83.11 on 24 September 2021. Its lowest was $3.93 on 27 March 2026. On 16 October 2026 it closed at $6.13: 92.6% below the peak and 63.9% below the $17 IPO price. Over the same period, from Snap's first trading day to 16 October 2026, the S&P 500 rose 221.6%, from 2,382 to 7,659.^[1]

At the 16 October close Snap was worth about $10.37 billion, using the latest reported share count. At the peak it was worth $133.8 billion, 13 times more. Two days stand out. On 22 October 2021, after Apple's App Tracking Transparency change weakened Snap's ad targeting, the stock fell 26.6% in one session. On 24 May 2022 a guidance warning took it down 43.1%, the largest one-day fall in its history.^[1][2]

What the valuation says now

Snap has no trailing price-to-earnings ratio because it has no trailing earnings. On analysts' consensus forecast for 2027 earnings per share, $0.134 from 37 analysts, it trades at about 45.7 times earnings. That figure is our own calculation and not comparable with the forward P/E Yahoo publishes for other companies, which uses a different earnings field. On Yahoo's own forward basis Snap is 7.86, 14% below Pinterest (9.1) and 51% below Reddit (16.1). The stock trades at 1.63 times trailing sales.^[1]

The balance sheet has turned. At the end of 2017 Snap held $2.04 billion more in cash than it owed. On 30 June 2026 its debt of $3.54 billion was 33% larger than its $2.66 billion of cash and securities, leaving net debt of $875 million. All three agencies rate it below investment grade: S&P BB-, Moody's B1 and Fitch BB. Meta, by comparison, is rated AA-, six steps above the lowest investment grade.^[2][16]

The Hype and the Crash

Snapchat launched in 2011 out of Stanford, first as Picaboo and, from September 2011, as Snapchat. It popularised disappearing messages and created the Stories format: posts that vanish after 24 hours, play in order and carry no public likes. In 2013 the Wall Street Journal reported that Facebook had offered $3 billion in cash and that Snap had said no. Both companies declined to comment at the time; Spiegel later confirmed there had been "a real conversation" but said $3 billion is "not technically the number".^[7][9] Snap is worth about 3.5 times that reported offer today.

Instagram copied Stories in 2016, and the format spread to WhatsApp and most other social apps. Spiegel takes the copying almost as a compliment, but he also drew a lesson from it: "15 years ago we essentially learned that software is not a moat, which is something that everyone is discovering today with AI."^[9]

Snap went public on 2 March 2017 at $17 a share and closed its first day at $24.48, up 44% from the IPO price and worth $28.9 billion. Revenue grew between 43% and 64% a year from 2018 to 2021, from $825 million in 2017 to $4,117 million in 2021. At its Investor Day in February 2021 Snap told investors to expect "multiple years of 50% plus revenue growth". The stock peaked seven months later, at the height of the zero-interest-rate era that inflated many growth stocks.^[8]

The promise held for exactly one year. Growth fell to 11.8% in 2022, from $4,117 million to $4,602 million, and to 0.1% in 2023. Had Snap kept growing 50% a year from 2020, its 2025 revenue would have been about $19.0 billion. It was $5.93 billion.^[2]

The Good: Current Growth and Health

Snap's latest numbers are its best in years. Revenue in the second quarter of 2026 rose 18.9% to $1.60 billion, from $1.35 billion a year earlier. That beat Pinterest (18.2%), though it trailed Meta (28.0%) and Reddit (61.1%). The growth came from an easy comparison and a new revenue line: 57% of the increase came from Other revenue, which is mostly subscriptions, while advertising grew only 9%, from about $1.17 billion to $1.28 billion.^[3][4]

Annual revenue reached $5.93 billion in 2025, 7.2 times its 2017 level. Daily users rose 5.1% to 493 million, from 469 million a year earlier and about three times the 166 million of the first quarter of 2017. Monthly users reached 971 million, 3% short of the 1 billion goal Snap set at its 2023 Investor Day.^[2][3]

Free cash flow has been positive every year since 2021. Profitability, measured before stock pay and other adjustments, has improved sharply: adjusted EBITDA was $250 million in the second quarter, about six times the $41 million of a year earlier. Gross margin rose to 58.2%. Even so, the comparison with peers shows how far Snap still has to go.^[3]

Products

Snapchat opens to the camera, and that camera is still where Snap is strongest. Snapchatters use Lenses, its augmented reality filters, about 9 billion times a day on average, and 75% of users engage with AR daily. By Q4 2025 Snap reported more than 700 million people engaging with generative AI Lenses more than 17 billion times, 40% more people and 2.8 times as many uses as it reported a quarter earlier.^[3]

• Spotlight, a TikTok-style short-video feed, had over 550 million monthly users in Q2 2025.
• Lens Games reported 255 million monthly players in Q1 2026.
• Snap Map, My AI and Memories round out the app; My AI uses Google's Gemini.

Subscriptions: the real new business

Snapchat+ launched in June 2022 at $3.99 a month, weeks after ad growth collapsed. Lens+ followed in 2025 for AI Lenses, alongside a Platinum tier and paid Memories Storage plans. Other revenue, led by these subscriptions, reached $316 million in Q2 2026, up 85% from about $171 million, and 19.8% of total revenue. Spiegel says Snapchat+ hit 25 million subscribers and "more than a billion revenue run rate". Fewer than 3% of monthly users pay.^[3][9]

Behind the camera sits a large ecosystem: more than 450,000 Lens developers have built more than 5 million Lenses. Spotlight paid creators over $250 million in 2021. Stories creators have received an ad-revenue share since 2022, and Spotlight creators joined that programme in 2025. Snap's new targets are sustained positive net income from 2027 and a flat diluted share count in 2027.^[3][17]

The Bad: Hovering on Losses

Snap has lost money in every full year since its IPO: $11.08 billion from 2017 to 2025. Over the same years it paid $10.50 billion in stock-based compensation, 95% of the losses. Snap issues new shares to pay its staff and then buys stock back to limit the dilution of existing shareholders.^[2]

Stock pay was $1,017 million in 2025. Free cash flow was 57% smaller. Stock pay costs no cash, so it is added back in the cash flow statement, which is how Snap can report a loss and positive free cash flow in the same year. Per employee it came to about $193,000 in 2025, 25% less than Meta's $259,000 and 16% more than Pinterest's $167,000. What sets Snap apart is not the amount per head but what it does to the results: at Meta, stock pay is about 10% of revenue at a highly profitable company; at Snap it is 17% and the reason for every annual loss.^[2][4]

The falling-price spiral

Stock pay is set in dollars, so a falling share price means more new shares for the same pay. For illustration: at $6.13, the last twelve months of stock pay, about $1.03 billion, equals roughly 168 million new shares, about 10% of all shares. At the $83.11 peak the same pay needed 14 times fewer shares.

Buybacks are the counterweight. Snap has spent $2.85 billion on them since 2022, and in 2025 its buybacks of $751 million were 72% more than its free cash flow. They still did not stop the dilution: the share count is up 4.1% since February 2022, from 1,624.3 million to 1,691.2 million, and 43.4% since April 2017.^[2]

Depending on expensive creditors

Between 2019 and 2022 Snap borrowed cheaply through convertible notes paying 0% to 0.75% interest, because investors could swap them for shares if the stock rose; the 2027 notes, for example, converted at $89.25 a share. With the stock near $6 that option became worthless. In 2025 Snap issued $1.5 billion and $550 million of senior notes at 6.875%, due in 2033 and 2034, and used the money to buy back most of the old convertibles. Interest on its notes rose from $41.9 million in the first half of 2025 to $72.8 million in the first half of 2026. In June 2026 S&P upgraded Snap to BB-, citing among other reasons its cost cuts.^[2][16]

Layoffs, and a hiring plan that flipped

Staff grew 3.5 times from 1,859 at the end of 2016 to a peak of 6,446 in mid-2022, then fell 26.7% to 4,723 in June 2026. There were four major rounds of cuts, and the charges since 2022 total $430 million, about half of it severance.^[2][11]

The 2026 round came as a reversal. In February the CFO guided to about 7% headcount growth for 2026, from 5,261 at the end of 2025. On 31 March the activist fund Irenic Capital published a letter demanding cuts and a spin-off or shutdown of Specs. On 15 April Snap announced cuts affecting about 1,000 people, including 16% of full-time staff, and closed more than 300 open roles; the stock rose 7.9% that day, from $5.60 to $6.04.^[11][12]

Earlier savings did not last. The 2022 restructuring cut annualised costs by more than its $500 million target, but by the fourth quarter of 2025 adjusted operating costs were back within 1% of their pre-cut level, on 55% more revenue.^[11]

Dropping the World and Being Dropped at Home

For a social network, growth is usually everything. Snap's problem is where the growth happens. In Q2 2026, 61% of its daily users were in Rest of World, outside North America and Europe, and they brought in 19% of revenue. North America had 19% of users and 59% of revenue.^[3]

Snap's average infrastructure cost is about $0.82 per daily user per quarter, our estimate from its own disclosures. That is 82% of what an average Rest of World user brings in. Snap does not disclose cost by region, but the comparison explains why it has started treating these users differently.

The shrinking home market

North American daily users fell 9%, from 101 million in 2023 to 92 million. Losing 9 million users at North America's average revenue per user is about 6% of quarterly revenue. Revenue there still grew 15% year on year in Q2 2026: users fell 6% while revenue per user rose 23%. Snap is squeezing more revenue out of a shrinking base.^[3]

Losing teenagers

Piper Sandler's teen survey shows the shift. Snapchat was teens' favourite social app from 2016 to 2021, peaking at 47% in Fall 2017. TikTok passed it in Spring 2022 and Instagram in Spring 2024, and by 2025 Snapchat was at 14%, 70% lower. Pew's measure of US teens who use Snapchat at all fell from 60% in 2023 to 55% in 2025, 8% lower.^[5][6]

Pulling back from the rest of the world

In Q4 2025 Snap "substantially" cut its paid user acquisition, and daily users fell 0.6%, from 477 million to 474 million, in one quarter. It also lowered the cost of serving users in low-revenue countries, accepting less engagement there. Rest of World users still grew 12% in a year, from 271 million to 303 million, but Snap no longer pays to grow them.^[3]

Regulation adds pressure. Australia's under-16 ban removed about 400,000 Snapchat accounts. The European Commission has proposed a KIDS Act with no social media under 13, and more than 20 countries have bans or plans. Snap's audience is older than its image: about 80% of users were over 18 in Q2 2024, US growth is "led by people aged 35 and older", and India alone had 250 million monthly users in July 2025, 27% of the total at the time.^[3]

The Infrastructure

Snap owns no data centres. It has run on Google Cloud since 2011 and signed its Amazon Web Services agreement in 2016. At the IPO it committed $2.0 billion to Google and $1.0 billion to AWS, raised to $1.1 billion in 2018. In June 2026 it had $2.6 billion of non-cancellable commitments, covering cloud, leases and content deals.^[2]

Our estimate of Snap's server cost rose from $818 million in 2022 to $1,584 million in 2025, up 94% while daily users grew 26%, from 375 million to 474 million. It now equals about 27% of revenue. Machine learning for ad targeting and content ranking, AI Lenses and My AI drove the increase. My AI uses Google's Gemini; Snap's own camera AI models can run on users' phones to save server cost.^[2]

The cost explains the margin gap only in part. In 2025 Snap's gross margin was 55.0%, against 80.1% at Pinterest, 82.0% at Meta and 91.2% at Reddit. Pinterest also rents its computing from AWS and still keeps 80%, so renting alone does not explain the gap. Snap's extra costs come from heavy camera, video and AI workloads and from revenue shared with creators. Spiegel himself wrote it plainly in September 2025.^[10][19]

The 10-K adds a reminder of the dependence: Snap has had outages at its cloud providers and expects more.

Strategic Directions

Snap spent about $1.94 billion on acquisitions from 2017 to 2025, measured by purchase price. The biggest single bet was glasses technology.^[2][15]

• Smaller deals: Voisey (music app, shut 2022), StreetCred (location data), Pixel8earth (3D maps), NextMind (brain-computer interface), Saturn (student calendar app) and Illumix (AR mapping for Specs).
• Before the IPO: Looksery (face filters, the base of Lenses), Bitstrips (cartoon avatars) and Vergence Labs (smart glasses).

The focus now is profitability. Free cash flow per share is Snap's stated primary financial objective. In advertising it has rebuilt around performance ads, which advertisers pay for per install or purchase and which reached 75% of ad revenue in Q1 2025. Small and medium businesses are now more than 30% of ad revenue, while brand advertising was flat or down in most quarters from Q4 2022 to Q2 2025. New formats include Sponsored Snaps in the chat inbox and Promoted Places on the map.^[3]

Beyond ads and subscriptions, Spiegel's idea is a multi-app strategy: build new apps cheaply with AI and push them through Snapchat's reach, with the Saturn calendar as an example. Snap discloses no revenue for these apps yet.^[9]

Innovating Astray

Snap uses AI across the organisation for efficiency, and it reports striking numbers. These are company claims, not independent measurements.^[11][3]

It also builds its own models. In 2025 Snap introduced AI video Lenses powered by an in-house generative video model, and its Imagine Lens, where users type a prompt to create or edit a snap, was used nearly 2 billion times in about three months. The goal is cost, not frontier performance: small models that run on users' phones avoid paying for every use on a server. Revenue from consumer AI is mostly not disclosed; the main paid route is the Lens+ tier.^[3]

Specs: A Magic Leap Reprise

Snap has tried glasses before. Spectacles 1 sold for $129 to $130 from late 2016; Snap said it sold more than 220,000. It still took a $39.9 million write-off in Q3 2017, and The Information reported "hundreds of thousands" of unsold units. Spectacles 2 cost $150 and Spectacles 3 $380. Spectacles 4 and 5 were developer kits, with Spectacles 5 offered at $99 a month.^[20][2]

Specs, unveiled on 16 June 2026 at $2,195, is a different product: a standalone computer in glasses, with two Qualcomm chips, Snap's own operating system and waveguide display, and control by hands and voice. Snap names games, learning, utility and AI assistance as uses. A $2,395 version comes with a charging case, and shipping is expected "this fall" in the US, UK and France. The stock fell 9.6% on the day of the unveiling, from $5.71 to $5.16.^[13][1]

Snap says it has spent "more than $3 billion" on AR over 11 years. Both activist funds estimate Specs costs about $500 million a year; Snap does not disclose it. Blue Duck Capital wants Snap to sell 20% of Specs at a $5 billion valuation to fund it from outside.^[12]

The Magic Leap comparison

On the Q2 2026 call Snap did not disclose preorder numbers and did not directly answer whether a spin-off of Specs is off the table.^[14] Spiegel compares the first buyers to the people who bought the first Macintosh in 1984 and expects a mass market only "in the 2030 time frame".^[9] He also ties the bet directly to the core business.

Nobody can make him change course. The founders held 99.5% of the votes at the end of 2025, so outside shareholders cannot outvote them; Snap's public Class A shares carry no vote at all.^[2]

Outlook

The picture is pessimistic. Snap looks like a zombie social media app: a successful product inside a business that has never made a profit, kept alive by cash flow that depends on paying staff in shares.

The stock itself is a gamble. At $6.13 it is 56% above its March 2026 low of $3.93, and it stays volatile, but so many things can go wrong that its direction is hard to call. Snapchat was highly relevant in 2017. Since then TikTok, Instagram and X have developed so much, and private messengers such as Signal and Telegram offer disappearing messages of their own, that it is hard to see where Snapchat fits.

A thin moat

Snap's real strength is AR: the know-how, the app and the technology behind Lenses, plus its own image and video models built just for them. That lead may not last. If general image and video models learn to do the same out of the box, knowing AR very well stops being a protected niche. Distribution, the other moat Snap talks about, is real but sits in the wrong places: most of the growth is in Rest of World, where users bring in little, and the North American base that pays is shrinking.

Squeezing instead of building

The likely path is that Snap will not innovate heavily and will try to get as much as it can out of the users it has, through subscriptions and new apps pushed through Snapchat. That makes use of the reach Snap is good at. History suggests it rarely ends well: push monetisation too far and an app stops being enjoyable, and people leave. What Snap should do is invest in new things, and that is exactly what it struggles to do, because investors punish it whenever spending goes up.

Specs as the swing factor

Specs could still surprise. If they ship this fall and sell well, the stock could reach a new peak. Only two things would move it that far: clearly strong sales, or Specs capturing the moment the way AI, then AI agents and AI coworkers, captured investors. Neither is visible yet. At over $2,000, glasses need use cases that replace part of a computer, say 40% of the time on Specs and 60% on a laptop, and the young users Snapchat is built for live on their phones more than on computers. Our view is that Specs is more likely to flop, and that Snap would do better not to sink more money into it.

What would change the picture

Snap needs growth from new use cases and a broader audience, not only from people looking for filters or for another TikTok. That does not have to mean business or professional users. TikTok and Facebook are neutral apps, without flowers or cute dog ears, and they become what each user makes of them. Snapchat comes out of the box heavily tilted towards young people, the audience it is now losing. Broader appeal to other age groups would do more for Snap than another hardware bet.

Wall Street is cautious rather than hostile: 29 of 42 analysts rate the stock hold, with a mean target of $7.44, 21.3% above the 16 October close. The stock can be volatile and could rise if Snap shows something great, especially in AI or if Specs do well, but the risk is very high. Snap reports its third-quarter results on 4 November 2026.^[1]