Bullish Pre-Earnings DD on RDDT: High Growth, High Margins, and AI Data Licensing

The ticker discussed by the poster is Reddit Inc. (RDDT). The core thesis is that, ahead of earnings, investors should pay attention to the combination of “high growth + high margins + AI data licensing,” while the market has over-priced concerns that Google AI Overviews will siphon off search traffic. The poster cites Q4 2025 figures showing Reddit quarterly revenue of about $725 million, year-over-year growth of about 70%, adjusted EBITDA margin of about 45%, operating cash flow of about $267 million, cash and marketable securities of about $2.5 billion, and authorization of a $1 billion buyback. These figures are broadly consistent with the financial charts in the images, which also show DAUq rising steadily from Q1 2024 to Q4 2025, an accelerating revenue curve, ARPU—especially U.S. ARPU—moving sharply higher, gross margin staying above 90%, and rapid expansion in adjusted EBITDA and operating cash flow. The poster adds revenue growth and EBITDA margin to get roughly 115%, using a “Rule of 40” lens to argue that Reddit’s operating efficiency is far above the traditional healthy SaaS threshold, while drawing cross-company comparisons with NVDA and META. The second evidence chain is the content cost structure: Reddit does not directly pay users for content, moderators work for free, and the main costs are described as servers plus trust and safety, which supports high gross margins and rapid expansion in EBITDA margin. The third evidence chain is AI licensing. The poster argues that data licensing revenue from Google at about $60 million per year and OpenAI at about $70 million per year already represents about 10% of revenue, and that if pricing later shifts to a dynamic model based on the number of times Reddit content is cited in AI answers, this revenue line could be significantly re-rated. On the bear side, the poster acknowledges that U.S. DAUq growth has fallen from previously higher levels to 9% in Q4 2025, and that the market is worried Reddit is too dependent on Google search as an entry point, while AI Overviews directly answering questions could weaken the traffic funnel. However, the poster argues that ARPU matters more than user count: global ARPU rose from about $2.94 to about $5.98 over two years, U.S. ARPU is about $10.79, and international ARPU is about $2.31, suggesting that even if U.S. users stagnate in the short term, there is still room for international monetization. On valuation, the poster argues that the trailing P/E of 59x is affected by post-IPO stock-based compensation and should not be overemphasized; the forward P/E of 38x and EV/Adj. EBITDA of 32x are not expensive after adjusting for growth compared with META’s 22x EV/EBITDA. The rough scenario assumes 2027 revenue of $4 billion and a 50% EBITDA margin, or $2 billion of EBITDA. Applying META’s 22x multiple yields an EV of about $44 billion, implying about 60% upside versus the then-current roughly $27 billion market cap/enterprise value. The position screenshot shows the poster holding 1,837 shares of RDDT:US, with average cost displayed as 205.3502 in CAD terms, current price as 161.90 in USD terms, market value of about CAD 405,013, and unrealized gains of about CAD 27,785, or +7.37%; in the comments, OP explains that the green gain comes from CAD conversion. The most important weaknesses are that the valuation scenario depends heavily on Reddit still sustaining high growth and a 50% EBITDA margin in 2027; AI licensing dynamic pricing and pay-per-citation models lack public certainty; international ARPU improvement does not necessarily mean it can approach U.S. levels; and Google/AI search traffic risk, the advertising cycle, and erosion of content quality and ad value from bots and AI slop are all key variables that the poster handles relatively lightly.

The comment section is lively overall and has a relatively large sample, but it is heavily WSB-style, with lots of jokes, sarcasm, and position-driven venting, so it cannot simply be treated as a serious consensus. Supporters mainly agree with the uniqueness of Reddit’s data: one of the highest-upvoted comments argues that Reddit differs from social platforms like Facebook, X, and Snapchat because it has a large amount of in-depth back-and-forth discussion around purchase decisions, product experiences, and interest-based communities. This kind of “real human discussion” is more valuable for AI training, search, and ad targeting. Others emphasize subreddit self-organization, free moderator management, vertical interest communities, and potential marketplace functionality as Reddit’s long-term moat. Some users cite their own usage frequency as a simple bullish reason, while others say they will keep buying dips or holding large stock positions and remain bullish long term. The bear arguments focus on three areas. First, the earnings trade itself is high-risk: many users joke that “once DD appears, it drops” or “PUTS it is,” and point out that RDDT may fall after earnings even if results are good, reflecting that the market may have already priced in good news or that valuation is too high. Second, bots, AI-generated content, and “Dead Internet Theory” are highly upvoted rebuttals. Multiple comments argue that Reddit’s value as AI data could be diluted by AI slop, and that if AI companies later discover the training data increasingly resembles model output itself, their willingness to pay could decline. Third, platform governance and moderator power are treated by some users as long-term risks. Some say they sold after being banned from large subreddits, arguing that lack of appeals and inconsistent rules harm user experience. On AI licensing, comments are split: some believe Google/OpenAI are currently paying too little and that prices will rise at renewal, while others point out that these agreements may not be updated until 2027, so they may not show up immediately in this earnings report. OP’s own replies mainly include saying it is too early to celebrate, arguing that RDDT will return to 250, admitting that not selling at 250 earlier was “stupid,” and acknowledging that those who sold at the highs made the right move. OP also explains the currency difference in the position screenshot as CAD conversion, and after earnings comments that the metrics look like a beat, with the stock rising after hours, though it remains to be seen whether the gains can hold. The highest-upvoted TLDR has typical WSB sarcasm, summarizing the bull case as “selling shitposts to AI companies while mods work for free.” Its real meaning is both to acknowledge the leverage in the business model and to mock its absurdity. Overall, the comment section shows strong intuitive agreement on the value of Reddit’s data and community, but clear disagreement remains over bot contamination, the sustainability of AI licensing, valuation, and post-earnings stock reaction.