---
title: "RDDT (Trans): Focus on feed improvements"
type: "News"
locale: "en"
url: "https://longbridge.com/en/news/294431032.md"
datetime: "2026-07-31T00:22:09.000Z"
locales:
  - [zh-CN](https://longbridge.com/zh-CN/news/294431032.md)
  - [en](https://longbridge.com/en/news/294431032.md)
  - [zh-HK](https://longbridge.com/zh-HK/news/294431032.md)
---

# RDDT (Trans): Focus on feed improvements

**Below is Dolphin Research’s transcript of** $ Reddit.US **FY26 Q2 earnings call**

**I. Key takeaways**

1\. **Shareholder returns**: Buybacks became a more meaningful capital tool this quarter, with approx. $235 mn repurchased (~1.5 mn shares) at an Avg. price of ~$157. Remaining authorization is ~$760 mn. Capital allocation priorities are: invest in the biz., opportunistic M&A, and buybacks, while maintaining high profitability. Cash, cash equivalents and marketable securities were $2.8 bn at quarter-end, up over $700 mn YoY and slightly higher QoQ, as operating cash inflows were largely offset by repurchases.

2\. **Q3 guidance**: Revenue of $860–870 mn (+47% to +49% YoY; ~48% at midpoint). Adj. EBITDA of $385–395 mn (+63% to +67% YoY), implying a 45% Adj. EBITDA margin at midpoint. The outlook incorporates current investment plans.

3\. **SBC guidance cut**: Full-year SBC as a % of revenue lowered from high-teens to low-to-mid teens. YTD SBC is under 13% of revenue, down over 1,000 bps vs. 1H last year. Q3 SBC and related taxes are expected at $140–155 mn, driven by vesting of grants issued at lower prices and re-grants at higher prices. Full-year dilution is expected near the low end of the 1%–3% target range (excl. further offset from buybacks and any incremental from M&A).

4\. **Key financials**: Q2 revenue of $805 mn (+61% YoY), the eighth straight quarter with 60%+ growth. Adj. EBITDA margin was 43%, marking the fifth consecutive quarter in which revenue growth plus Adj. EBITDA margin exceeded 100%. GPM was 91.3% (+50 bps YoY). Adj. total costs and expenses were $462 mn (+39% YoY; broadly in line with the 41% Avg. growth over the past four quarters), including COGS of $70 mn (+53% YoY) and Adj. Opex of $392 mn, or 49% of revenue (57% a year ago). Net income was $253 mn; basic EPS $1.31 and diluted EPS $1.25 (vs. $0.48 and $0.45 in the prior-year period). Operating cash flow more than doubled YoY to $262 mn, with LTM OCF exceeding $1 bn for the first time. Capex was only $1 mn (<0.2% of revenue). Fully diluted share count was 207 mn, up 0.3% QoQ and 0.2% YoY on a net basis; QoQ net dilution was 0.3%, with gross dilution of 1% excl. buyback effects. LTM revenue per employee topped $1 mn for the first time, a key IPO benchmark. Headcount increased by ~100 in Q2.

5\. **Disclosure change**: Q2 is the last quarter with separate metrics for logged-in vs. logged-out users. Starting Q3 FY26, the company will continue to disclose total US and Intl DAUs and WAUs, but will no longer break out logged-in/logged-out DAUs.

**II. Earnings call details**

**2.1 Prepared remarks**

1\. **Platform positioning and user strategy**

a. The core thesis: as the internet grows more automated and synthetic content proliferates, human-to-human dialogue becomes more valuable. AI compresses the internet into summaries, while Reddit offers deep discussion, vigorous debate and first-hand experience — people do not want a Reddit summary, they want Reddit itself. Information is abundant as never before; what is scarce is context, personal viewpoints and first-hand accounts, so decisions are shifting from a single authority to real people with no commercial stake, across shopping as well as health, careers and travel.

b. The top priority remains growing DAUs by improving the new-user experience to drive return frequency. The company already has a large WAU base, and the focus is converting WAUs to DAUs. Long-term targets are unchanged: 1 bn global DAUs and 100 mn US DAUs.

c. Scale today: WAUs exceed 500 mn with DAUs over 130 mn. Nearly 200 mn users visit weekly in the US, and both US and Intl WAUs grew QoQ. Reddit now hosts 26 bn posts and comments, up by 1 bn since Q1.

d. User mix improved this quarter, with new App-user retention (relative basis) up 50% YoY, a product quality KPI management values highly. Though off a small base, the direction is clear. Product work plus consumer marketing offset search referral headwinds.

e. Search-driven traffic was volatile intra-quarter and less stable later in the period, with external visibility and referrals still low. Management stressed they are not building for drive-by traffic, but for daily destination use. Direct users are worth multiples of search/viral traffic.

2\. **Consumer product progress**

a. Feed: updated recommendation models increased content contribution and conversation frequency. Management acknowledged feed tech lags the industry frontier, leaving ample room to improve: models are much smaller, update on a cadence of days vs. hours/minutes, only ~10% of user signals are modeled, and candidate pools are restricted to the past week, excluding large volumes of evergreen content like parenting advice and book/film reviews. There are order-of-magnitude opportunities across candidate selection, model size, refresh speed and signal breadth, with work planned over the next year.

b. Video: the video-in-comments feature now drives over 10% of video posts, expanding the top of funnel. Next steps are fixing video presentation in the home feed — the feed → theater mode → comments page → back-to-feed loop — and exploring a listenable Reddit experience (background audio for posts), akin to off-platform podcasts that read Reddit content. Tests are planned later this year.

c. Search and Answers: steady progress; the search bar is now global in-app, with users and query volumes up this quarter. For most queries, Reddit is becoming the best place to search Reddit content.

d. Web-to-App conversion: the company continues to funnel high-frequency web users to mobile App, who already tend to be Reddit users with good retention and stronger in-app behavior. This remains a focus.

e. Content ecosystem and post success rate: new post and comment composers are live on mobile and web, with more accurate community recommendations. Posts per DAU and comments per DAU (logged-in) are up globally and in the US, and active communities are rising. A key pain point is new-user posts being removed due to account age; LLM-assisted moderation is allowing high-quality newcomers through while better blocking spam. Management views Reddit’s creator ecosystem as distinct: social platforms serve the top 1% of professional creators, while Reddit serves everyone else.

f. Interactive games on the developer platform are helping build repeat usage and daily habits. Engagement benefits have been observed.

3\. **Ads: automation and Reddit Max**

a. Reddit Max entered private testing earlier this year, showing strong adoption and results among mid- and lower-funnel advertisers. Advertisers using Max rose 60%+ QoQ, with Max revenue up 150%+ over the same period.

b. In Q2, Max added an App-install objective and launched tailored creatives, using AI to identify relevant communities and audiences and generate creative variants to boost relevance and performance. These enhancements are expanding effectiveness.

c. Case study: Lenovo saw purchase value 40% higher vs. standard campaigns using Max. The plan is to scale Max to a broader advertiser base, make it the primary on-ramp for SMBs, and add features that simplify buying.

4\. **Ads: full-funnel products and performance**

a. Upper funnel: video views are the fastest-growing upper-funnel goal, with revenue more than doubling YoY. The 6-sec Engaged Video Views objective, rolled out broadly in Jan, drove a 130%+ uplift in 6-sec completion and a 70%+ uplift in full views in testing, while automated bidding cut cost per 6-sec view by 80%. Interactive formats and AMA ads continue to see investment.

b. Lower funnel: shopping and App-install objectives are delivering measurable growth. Max for App ads is in beta, reducing Avg. cost per conversion by 15% and increasing results by 28% in tests; online pharmacy Doc Morris cut CPI by 20% and beat benchmark CPA by 70%+ with Max, then increased budget.

c. Shopping: along the DPA roadmap, an alpha of shopping listing ads has launched — a multi-advertiser carousel that matches product catalogs to high-intent shopping conversations on Reddit (+40% YoY). This extends performance capability from single-advertiser catalog ads to a broader module.

d. Conversions and clicks: conversions more than doubled YoY in Q2, and clicks rose 30%+ YoY. Mid- and lower-funnel performance objectives were key contributors, with DPA and App-install revenue more than doubling YoY.

5\. **Ads: customer mix and ecosystem**

a. Mix: ad revenue grew 64% YoY in Q2, driven by both price and impressions. US revenue rose 56% YoY; Intl was up 84%, and Rest of World grew 80%+. ARPU increased 36% YoY to $6.18.

b. Clients and channels: active advertisers rose 70%+ YoY. Scaled channels serving mid-market and SMBs doubled YoY. 11 of 15 verticals grew 50%+ YoY, led by retail and travel.

c. Infrastructure: Ads API revenue grew 40%+ QoQ, with over half from clients using credit lines. Shopify integration fully launched in May; 500+ accounts have connected with healthy activation and Avg. revenue per advertiser. Next, focus is on Shopify UX, merchant acquisition and boosting DPA adoption.

d. Measurement: dual attribution (beta) launched in Q2, combining first-party signals with existing MMP data to improve ML, optimization and advertiser outcomes, letting App-install clients see Reddit’s role beyond last-click. Third-party studies: Circana found CPG ROAS on Reddit is 1.5x other social platforms on Avg., with higher user LTV in food, beverage and personal care; TransUnion identified Reddit as the most efficient paid social channel for EMEA retail, with Avg. ROAS 7x and 13% of impact spilling over to other channels.

e. Brand and market: the new brand campaign ‘People Are the Best’ launched. At Cannes, the team met with 1,000+ clients and partners, holding 500+ meetings that covered nearly all major accounts.

6\. **Marketing spend**

a. Q2 paid marketing results were mixed: brand signals were positive, but UA still needs stronger retention to improve ROI. Total marketing spend (paid + brand) rose slightly as a % of revenue QoQ, and the mix will keep evolving, potentially with more brand investment.

b. Management views marketing as a key lever toward the ‘North Star’ DAU goals (100 mn US, 1 bn global), alongside the $1 mn revenue per employee metric. However, stepping up UA depends on continued retention gains.

**2.2 Q&A**

**Q: What drove the strong ad performance this quarter?**

A: It was another strong quarter with notable consistency. First, our narrative is resonating broadly with advertisers — Reddit’s influence in the purchase journey is clear, and advertisers increasingly recognize that their customers value advice from real people, which is what Reddit hosts, so they understand the importance of our environment. We continue to deliver competitive performance, building strategic momentum. Our strategy is to make every impression more valuable. When conversions and clicks both rise, advertisers get more outcomes at the same or better prices. Our work on signals, ML and optimization is steadily enhancing competitiveness. Tool adoption is also inflecting, lowering onboarding friction and contributing to ~70% YoY growth in advertisers. Once onboarded, better efficiency retains them and supports budget expansion. With higher credit-line usage and reduced cross-funnel friction, it is easier to keep campaigns ‘always on’. Finally, our business is diversifying across many dimensions, which makes it resilient: 11 of 15 verticals grew 50%+ YoY, active advertisers rose 70%+, scaled channels for mid-market and SMBs doubled, and Rest of World grew 80%+. A more diverse footprint translates into greater resilience and flexibility.

**Q: On user growth in Q2 and early Q3, and attracting high-quality users, what should investors take away?**

A: Our targets remain 1 bn global DAUs and 100 mn US DAUs. Reddit is a destination, with nearly 200 mn weekly US visitors, and we grow primarily through product work that drives direct daily use; here we are making progress. We added DAUs via product, but this was offset by a decline in search referrals. On product, new App-user retention improved — up 50% YoY on a relative basis — one of our top priorities. Reach is expanding, WAUs broke 500 mn, and both US and Intl WAUs grew QoQ. We are making it easier for users to find the right communities and post; Reddit has 26 bn posts and comments, up 1 bn since Q1, active communities are growing, and we are converting more web users to higher-value App and direct users, while video-in-comments expands the top of funnel. We do face headwinds in search referrals, especially later in the quarter, with low visibility, but we are focused on what we can control — product-driven direct users. This will be transformational and will take time, and it is how we get to 1 bn/100 mn DAUs.

**Q: With WAUs growing faster than DAUs, which products will convert WAUs to DAUs? What are the feed priorities?**

A: Raising DAUs is about increasing frequency by making Reddit stickier, and the feed is central — near-, mid- and long-term feed upgrades will lift frequency, and the runway is large. Frankly, our feed tech lags the frontier, which gives us catch-up headroom, and the talent we have brought in over the last quarter is encouraging for building the next-gen feed. Incremental improvements matter too. Last quarter we ran hundreds of experiments, many moving toward GA, each chipping away at retention and quality to drive growth — from new-user onboarding to posting and overall App performance — and we are starting to feel momentum. Conversion is also key — moving our large web top-of-funnel into mobile App. We are leaning in to drive more downloads. Many of these users are already Reddit users, show better retention, and perform better in-App, so we see progress across multiple tracks with significant room left.

**Q: There has been a lot of reporting on data licensing. With renewals approaching early next year, any strategic updates?**

A: No change in strategy, and we are progressing. Reddit content is in demand and is not commoditized — people seek it by name, and many come directly. Our partnerships are multi-layered: Reddit is used for training, for post-training to teach models how to ‘speak’, for grounding, and as a search index. So the value goes far beyond raw training data. These deals are not binary. They must make sense for Reddit, but value can flow back in many ways, and our goal is to maximize that value to Reddit.

**Q: Shares are down sharply, and investors see a ‘user problem’, especially in the US where DAUs dipped slightly despite heavy spend. As search shifts to AI, logged-out traffic may remain pressured. What underpins confidence in reaching 100 mn US logged-in and 1 bn global?**

A: We are confident because our product work is delivering. Reddit is about community and conversation, which are universal. In the US we have content for everyone; the question is discovery, and we are making progress there. Significantly improving new-user App retention in a single quarter is a point of pride, and retention gains drive growth and are within our control. Yes, external search is volatile, especially logged-out web, but that is not where our business sits. We will monetize that traffic as best we can, which is why we push web-to-App conversion whose efficiency is improving. Over the long run, we remain confident in hitting milestones; it takes time and more product work, but direct-user growth is in our control, and we are seeing progress.

**Q: Is there a scenario where you do not renew data licenses with Google and OpenAI next year?**

A: The outcome range is wide, and we will examine every aspect to maximize value for Reddit. With Google, the relationship goes beyond licensing — it predates the formal agreement, spans the ‘ten blue links’ era to Reddit’s role in AI Overviews, of which licensing is only one piece. Reddit content use-cases continue to expand at Google and elsewhere. This is not binary. Training, AI Overviews and future formats are distinct touchpoints, each with its own decision, and we will maximize Reddit’s value across them. It is not only about traffic or clicks; we want to bring users into the Reddit ecosystem and flywheel of communities and conversations through these partnerships.

**Q: Beyond Google and OpenAI, are there more buyers for data?**

A: Yes. Over the past year, more parties have shown interest in Reddit data across different use-cases and scales. We see this as a growing market with many opportunities.

**Q: On late-quarter search headwinds — were they routine or outsized, and what drove them? (Transcript lacked the search-specific answer; recorded response focused on feed tech, video in feed and search monetization.)**

A: On feed, the priority is improving the underlying tech, which does the heavy lifting — surfacing posts from subscribed subreddits and driving subreddit discovery. A key near-term effort is presenting video in the home feed. As video grows on Reddit, we have thought about this along two dimensions: first, making video work in a mixed-media feed with text and images, ensuring smooth transitions feed → theater mode → comments → back to feed; second, building a ‘listenable Reddit’ where users can listen to posts in the background. Off-platform podcasts that read Reddit content are already emerging, and we see this as an attractive, almost different format to test later this year. On Answers and search, progress is steady. The search bar is now global in-app, with users and queries up this quarter. For many queries, Reddit is already the best place to search Reddit content, and we are closing gaps. For questions like ‘what should I watch?’, ‘what do people think?’, or ‘what should I buy?’, Reddit offers some of the internet’s best answers, and we are beginning to explore ads in this surface. From an ads standpoint, search ties closely to shopping. We believe richer media modules can enhance search pages by pulling in product visuals from our existing catalogs when users search or discuss specific items, and early tests started in electronics and are expanding. This should enrich the core search experience and lift engagement, and we agree agentic features are now the best way to search Reddit. Second is capture after engagement at the product level. Shopping listing ads can contain multiple retailers, product types and brands within one module, fitting the search page well and representing, in our view, the right ad format there. We will keep tracking page formation and user adoption; the opportunity is clear, and the infrastructure and most capabilities are in place.

**Q: What specifically drove the 50% YoY improvement in new App-user retention?**

A: The 50% relative uplift in new App-user retention is a tough metric to move and one of our most important KPIs, arguably the best measure of product quality — whether users come back the following week after opening Reddit. Several factors contributed. The biggest is conversion: turning core or frequent web users into App users, who are higher quality in-App, and we have many of them. Second is in-App improvements: a much better new-user onboarding sequence that connects users to their ‘home’ communities more effectively. We also executed numerous ‘in the trenches’ fixes like login, account recovery and preserving session state after re-install, which collectively moved the number.

**Q: Where will you focus to upgrade the feed models?**

A: Broadly, our models are far smaller than frontier systems, so we can scale up model size. Our refresh cadence is slower — in days vs. hours or minutes — and we only model ~10% of user behavior today, which can rise materially. Candidate pools are limited to the last week, making the feed essentially real-time while Reddit has a vast evergreen corpus — parenting advice, book or film reviews — that remains valuable for long periods but is not shown in feed. We can upgrade almost every dimension — candidate selection, model size, speed and signal coverage — with potential order-of-magnitude gains in some areas. We will execute over the next year, and we expect each step to work given how low our starting point is.

**Q: Can you break down SMB progress (Shopify integration, Max), and how should we think about 2H?**

A: SMB is growing well, yet there are still thousands of SMBs globally not on our platform. First, we continue to acquire in this channel, and Shopify clearly helps — we can automatically sync catalogs and integrate, and will do joint marketing to accelerate acquisition with greater intensity. Second, SMB onboarding to Max is not yet optimal. We are still doing foundational work — defining the right SMB entry point to remove friction and fit their needs so they can activate quickly and benefit from Max. This remains ahead of us but is important for reaching more clients, smoothing acquisition and delivering fast wins. SMBs also benefit from horizontal improvements — better efficiency across objectives helps everyone. Momentum is good, though more tailored product work is needed for this segment.

**Q: Other 1 bn-user platforms have creator ecosystems. Does Reddit need a different content approach to reach that scale, or is current content sufficient?**

A: We saw broad progress in the creator ecosystem: posts per DAU and comments per DAU (logged-in) rose globally and in the US, and active communities increased. This was driven by new post and comment composers on mobile and web, easier posting flows, and better routing to subreddits where users are more likely to succeed, plus more accurate community recommendations. We are also tackling one of the largest opportunities — raising post success rates by making it easier and more consistent for new users to post. A current challenge is new-user posts being removed because accounts are not ‘old’ enough, an outdated policy. With LLM-assisted moderation, we can admit high-potential users while keeping spammers and bad actors out. Video-in-comments is new, already contributing 10%+ of video posts with solid engagement. Our creator ecosystem will differ from other platforms: social media serves the top 1% of creators; Reddit serves everyone else. Everyone has questions, needs, advice or something interesting to share even if not a pro creator — that is Reddit’s purpose. Making posting easier and raising success for everyone is key, and we made good progress this quarter.

**Q: Given today’s ARPU, marketing ROI seems compelling. Why not spend more? How is retention on acquired users, and how can it improve?**

A: Marketing is critical to our goals, and we want to keep investing in UA. Our model is straightforward and similar to peers. The unlock is better retention. As noted, we improved this metric by 50% this quarter, showing product work is paying off, but more is needed — we still trail peers, so we keep pushing. On marketing, the ‘People Are the Best’ brand campaign stirred positive coverage. The more we balance UA and brand, the better the outcomes. We are committed to invest across cycles and may lean more into brand. Ultimately, better retention is what unlocks more UA, period. Our margin structure is attractive, so incremental users are valuable, and high ARPU helps — we just need more progress on retention to accelerate.

**Q: As the company scales, where are the biggest operating leverage opportunities over the next few years?**

A: Our incremental margins have been in the high-50% to low-60% range for much of the past few years, which is robust. But we do not manage to an incremental margin target; we assess ROI on each investment. We have been fortunate on flow-through — incremental revenue, especially above GP, can flow at 95%–99% — so the question becomes which investments to make. We are investing where it makes sense. We will not hold back just to keep incremental margins in the high-50%s — that is not how we operate. Our investments are already paying off: heavy spend on the ads tech stack shows up in monetization, and investment in sales correlates with eight straight quarters of 60%+ growth; these are the most visible. We are also investing in talent, bringing in ML expertise and upgrading the growth team in a substantive way. We do this because it is right. Margins should keep compounding, with a long-term target of 50%. Most leverage will come from incremental revenue scale, and we are not targeting any specific large cost bucket to cut.

**Q: You emphasize high-quality (App, logged-in) users. Does that mean the value of a high-quality user is multiples of referral users, allowing mix shift to offset a ~300k QoQ decline in US DAUs? What related metrics will be disclosed going forward?**

A: That is exactly right. Direct/App users are worth multiples of search or viral traffic. We are not building Reddit for drive-by traffic; we are building for direct use, which is the core of our business. You can see this in ads performance — despite volatility in search (which we have seen many times), revenue growth remains strong and steady. The reason is our business is grounded in repeat direct users and growth of that base. Through conversion and retention, we made good progress this quarter, and this work is within our control regardless of broader search dynamics.

<End of text\>

**Risk disclosure and statements:**[**Dolphin Research Disclaimer and General Disclosures**](https://support.longbridge.global/topics/misc/dolphin-disclaimer)

### Related Stocks

- [RDDT.US](https://longbridge.com/en/quote/RDDT.US.md)

## Related News & Research

- ["Reddit's Growth Story Is Just Starting," Says KeyBanc Analyst Ahead of Q2 Results](https://longbridge.com/en/news/294239735.md)
- [Will Reddit’s Expanded AI Data Licensing and Q2 Outlook Shift the Core Reddit (RDDT) Narrative](https://longbridge.com/en/news/294274118.md)
- [Reddit Stock Is Rising Ahead of Thursday Earnings: What Investors Need to Know](https://longbridge.com/en/news/294229178.md)
- [Reddit Reports Second Quarter 2026 Results | RDDT Stock News](https://longbridge.com/en/news/294409036.md)
- [Cumberland Partners Ltd Purchases Shares of 6,000 Reddit Inc. $RDDT](https://longbridge.com/en/news/293839852.md)