RDDT (Trans): After Search, Feed Is the LT Monetization Driver
Dolphin Research's transcript of Reddit 1Q26 earnings call is below. For our take on the print, see 'Reddit: Enjoying the Upside of an Acceleration in Monetization'.
I. Key Takeaways
1. Q2 guide: revenue of $715mn–$725mn (+43%–45% YoY; ~44% at the midpoint); Adj. EBITDA of $285mn–$295mn (+71%–77% YoY), implying a ~40% margin at the midpoint.Total adjusted costs ~$430mn (+29% YoY), decelerating vs. prior quarters as the company starts to lap S&M investments from Q2 2025.
2. Margin expansion: Q1 Adj. EBITDA margin reached 40%, up ~1,100bps YoY.GAAP net income was $204mn with diluted EPS of $1.01, over 7x YoY (vs. $0.13 a year ago). GPM topped 90% for the 7th straight quarter (Q1 at 91.5%, +97bps YoY).
3. Record FCF: Q1 FCF margin was 47%, with operating cash flow above $300mn.CapEx was just $1mn (~0.2% of revenue), reinforcing the asset-light model.
4. Buybacks and dilution control: $1.0bn repurchase authorization (approved in Feb) with ~$995mn remaining; ~35k shares repurchased in Q1.Fully diluted shares at 206.4mn, +0.1% QoQ, +0.2% YoY, indicating minimal dilution.
5. KPI disclosure changes: Starting Q3 2026, Reddit will stop breaking out logged-in vs. logged-out DAUq, and will disclose only US/Intl DAUq and WAUq.
II. Call Details
2.1 Management Remarks
1) User growth and engagement
1. Nearly 500mn weekly users globally, with ~200mn in the US.US DAU around 50mn; the target is 100mn US DAU.
2. Search WAUq up ~30% YoY, with search a key DAU driver in recent quarters.Machine translation now covers 30 languages, with ongoing cost reductions to scale.
3. Q1 product progress: certain onboarding experiments moved to 100% rollout, feed improvements contributed, and new-user retention improved.H2 priorities: expand the top of funnel, improve new-user retention, and speed up iOS/Android across the board.
4. Bot verification: migrating existing 'good bots' (moderator-built community assistants) to the developer platform for transparent labeling, while strengthening defenses against unauthorized bots.
5. Passkeys rollout underway: integrating Face ID, Touch ID and other passwordless logins to improve security and reduce friction, boosting logged-in users.This is also a lightweight, privacy-friendly human verification method.
2) Ads biz.
1. Q1 ad revenue of $625mn (+74% YoY), marking the 7th straight quarter of 60%+ growth.Lower-funnel conversion-led revenue grew triple digits YoY, with performance ads now over 60% of total and healthy vertical mix; active advertisers +75% YoY. Q1 QoQ advertiser adds were led by Retail/CPG, Tech, and Media & Ent.
2. Reddit Max (automated ads): entered beta in Q1, onboarded by converting existing advertisers; thousands already using it.On Avg., Max lowered CPA by 17% and lifted conversions by 25%; ~50% of Max users adopted AI-driven creative. Example: home brand Cozy saw ROAS +35% and CPA -28%, while saving 2–3 hours per week; next steps are onboarding new advertisers directly onto Max and integrating Max into APIs for partners.
3. Brand Auto Bidding: now available to all advertisers, with tests showing ~16% Avg. pricing efficiency gains post-adoption.
4. DPA (Dynamic Product Ads): Avg. ROAS up over 90% YoY; Liquid I.V.'s DPA outperformed other conversion campaigns by ~40%.New Collection Ads and Reddit-native coverage formats (e.g., Redditor's Top Pick) improved shopping conversions.
5. Third-party ecosystem: announced Shopify integration at Shoptalk in Mar to enable SMB lower-funnel ads on Reddit.Measurement partner Fospha found Reddit lowered purchase costs by 34% while improving ROAS, helping ad spend scale 2.5x YoY; Reddit Pro is open to all publishers (used by Fortune Media, SF Chronicle, Dallas Morning News, Arsenal FC, etc.).
3) AI and data licensing
1. Reddit is the most-cited platform in AI outputs today, and 'Reddit' has been a top-10 term on Google Search for years.Each month, communities add content roughly equivalent to an entire Wikipedia.
2. As AI usage grows, demand for authentic human perspectives rises — Reddit content is the 'oil of the modern internet' and a foundational resource for next-gen AI.Active data partnerships with Google and OpenAI are mutually dependent; beyond commercial value, they provide tech support for machine translation, LLM-driven onboarding, and safety tools.
4) Content licensing
1. Q1 content licensing and other revenue was $39mn, +15% YoY.
2.2 Q&A
Q: What drives higher frequency and faster growth on the user side? And where are we on DPA adoption?
A (CEO Steve Huffman, on user growth): In Q1 we saw encouraging progress.A few onboarding experiments performed well and were rolled out to 100%, and the feed also contributed.
Looking ahead, the biggest lever may be performance — the speed of iOS and Android — where there is a lot to do.Onboarding also has plenty of room. Longer term, the feed will be the largest driver, which is why we are hiring ML talent aggressively. Search is progressing well too, with Search WAUq up ~30% YoY. Overall, the focus remains what we have always said — make the core product work better; we spent the past year upgrading team, process and tech, and our foundation is stronger than ever.
A (COO Jen Wong, on DPA): We launched DPA a year ago; shopping is a complex category and others started earlier, so our models and onboarding can still improve a lot.That said, we are pleased with ROAS gains, and partnerships with Shopify and WooCommerce are helping us bring more SMBs into DPA. It is still early, but the opportunity is exciting — thousands of existing advertisers have yet to adopt DPA. Categories like travel and auto are also on the roadmap, so it's a long runway.
Q: What are ML hires working on? What top-of-funnel tactics have worked best?
A (CEO Steve Huffman, on ML): ML touches everything — collecting more signals, weighting them better, and deploying models faster; we can speed up design-to-production substantially.It mirrors our ads journey: we are confident in the playbook because it has worked elsewhere, and we are hiring talent with at-scale (billions-of-users) experience. Our goal is to grow US DAU from ~50mn to 100mn, and we already have ~200mn weekly US users, so investing in the feed will lift frequency and retention to get us there.
A (COO Jen Wong, on top of funnel): Our largest existing top of funnel is search, and converting that traffic into core Reddit use cases (community browsing + feed) is a key opportunity.Second, we are adding new top of funnel through marketing, tailored by region. In mature markets like the US, we run targeted brand and acquisition campaigns against interest cohorts (e.g., parenting, NFL fans) to make clear that Reddit has content for them.
In Intl markets, we are building brand foundations — many know Reddit via search, but not our differentiated positioning as 'the most human place on the internet'.Reddit has done little marketing historically, so this remains fertile ground.
Q: Updates on bot verification, login and identity? What are the key learnings and next steps for Reddit Max?
A (CEO Steve Huffman, on bots and login): There are three intertwined layers here.The easiest is bot verification: our 'good bots' — mostly moderator-built community tools — are migrating to the developer platform for transparent labeling on Reddit, while we tighten controls on unauthorized bots; this supports human verification and platform defense.
On identity and login, Passkeys are key — Face ID, Touch ID, physical keys — a system where users must 'do something' (look at phone or touch a device) to log in.It is both more secure and more convenient, helping increase logged-in users; it is also the lightest, most privacy-friendly human verification approach. All three tracks are progressing; we shipped some in Q1 and more is coming this quarter.
A (COO Jen Wong, on Reddit Max): I am very pleased with Max adoption. Clients are eager to migrate and are happy with the out-of-the-box CPA improvements.Max also lets us roll out new performance features quickly — many clients auto-adopt one or two features they did not use before and see immediate uplift, shortening our feature launch cycles.
Clients also love the insights from Max — beyond performance gains, it shows what the system is uniquely optimizing on Reddit, like which communities best match your creative.Clients do not see a 'black box' — they are actually learning from it, and we will keep investing here. We have begun onboarding new advertisers directly into Max; it launched in Jan, this will be a multi-year adoption path, but progress and feedback are encouraging.
Q: Where does DAU rank in priorities? Can AI agent traffic that uses Reddit data count toward DAU via licensing?
A: Contrary to assumptions, DAU is the company's top priority — revenues are performing well, and DAU is both our mission (communities for everyone) and the fuel for the biz.We are focused on US DAU, aiming to grow from 50mn to 100mn; the opportunity is clear, with 200mn US weekly users on Reddit. We need to lift frequency — e.g., from once a week to daily. Recent quarters emphasized onboarding with progress in new-user retention; feed is the next big driver, search keeps contributing, machine translation helps, and performance is a major lever. Bottom line: DAU is Reddit's top priority, especially US DAU.
On AI and third-party agents, this is our ecosystem.We have important partnerships with Google and OpenAI, mutually beneficial, and we keep exploring other top-of-funnel opportunities to make our products complementary; no new updates to share on those relationships today.
Q: Is there a timetable for 100mn US DAU? If weekly users grow faster than daily, does that imply weakening engagement? Is AI data priced too low? Any exclusivity plans?
A (CEO Steve Huffman): When I returned ~10 years ago, we had ~12mn DAU; we have grown 10x to over 120mn since.Our goals now are 1bn global DAU and 100mn US DAU; I do not have a timetable, but we will persist.
The strategy is the same: build the best version of Reddit.We made major upgrades to team, process and tech over the past year; we can reach our prior best, but that is not enough, so we keep raising the bar. We want it asap, but it will take steady product improvements.
On AI data pricing, it is a fair question — join our partner discussions.The world sees the value of Reddit data; current and potential partners see it, too. Ultimately, without real intelligence there is no artificial intelligence — and that real intelligence comes from Reddit.
One trend: the more the internet is cleansed, summarized and attention-optimized by AI, the more people crave authentic human information — AI needs it, and users need it.Our business is real human connections and conversations. No comment on exclusivity.
Q: Where have talent and infra upgrades focused over the past year, and when will they show up in KPIs? Any risk to UX from higher ad load?
A (CEO Steve Huffman, on org upgrades): We have been upgrading end-to-end — people, process, tech — and are mid-journey.We made important leadership changes in product and engineering and continue to hire experienced talent; there is more to do, perhaps always. But we are acting — onboarding, feed and search have changed and are starting to drive growth, with retention improving.
I want faster cycles — code to production speed, experiment measurement precision, decision velocity all can improve.I see this as one integrated program to take Reddit to the next level. We made progress, more levels to go, and we aim to see immediate improvements in results — some are visible this quarter.
A (COO Jen Wong, on ad load): Reddit's overall ad load remains well below peers, especially in feed-to-feed comparisons where the gap is large; and in fast-growing surfaces like search, we do not even serve ads yet.So on an absolute basis, I am comfortable with the ad experience. We will keep testing.
As ML lifts relevance and as active advertisers scale (+75% YoY), even a modest increase in density can be supported by diversity while maintaining performance.Our core strategy is not to raise ad load, but to grow users (a 10x opportunity) and make each impression more valuable — more competition, stronger optimization, and higher clicks/conversions/installs so marketers get more outcomes per impression, not just more impressions. Naturally, impressions will grow with users, but we focus on value per impression.
Q: Are monetization trends converging between logged-in and logged-out users? Any macro/geopolitical impact on the ad market?
A (COO Jen Wong): On logged-in vs. logged-out, we have studied this extensively.We think in terms of 'impression value' — in our two primary surfaces (feed and conversation page), value per impression is actually quite similar. Logged-in ARPU is higher only because they spend more time and see more impressions.
We do not have a monetization disadvantage by user type — for logged-out users we have strong contextual signals, while for logged-in we have additional data like subscribed communities.On the ad market, geopolitics brings volatility; we have seen that before. No acute issues in any vertical.
Partners tell us some are planning on shorter cycles — month-by-month with lower visibility; but commitments and overall budgets have not materially changed, just shorter windows.We stay close to clients and use Reddit insights to help them navigate, which is proving helpful. The overall ad market looks fairly stable, just with more near-term planning and lower visibility.
Q: Beyond the daily/weekly active ratio, what engagement metrics do you track? Where is Reddit Max adoption today?
A (CEO Steve Huffman, on engagement): We focus on several core metrics.
First, new-user retention: whether users return on day 7, day 30, day 90 — the core measure of product quality and stickiness.Second, frequency: how many days per week users visit.
On a weekly histogram, the two tallest bars are typically 1 day and 7 days, consistent with Reddit intuition — once we retain you, you come daily; many users, however, bounce after arriving from search or after a brief trial.Converting more '1-day users' into '7-day users' is a huge opportunity. We are also beginning to track 'sessions per day', a more granular feed-quality and retention measure, though new to us. The most important, long-term metric remains new-user retention; while not disclosed, third-party data provides proxies. The answers are the same: quality, performance, relevance.
A (COO Jen Wong, on Max adoption): Reddit Max is the sales team's top push this year.Thousands of advertisers already use it, with many more to convert. We are onboarding new advertisers directly into Max and integrating Max into the API to reach partners who transact via API.
Launched in Jan, it is still early and a multi-year adoption arc — cf. Google's PMax and Meta's Advantage+ which have been years in the making.Current adoption velocity and client feedback are very encouraging.
Q: How has user engagement changed in Intl markets after machine translation? How much does personalization drive logged-in monetization?
A (CEO Steve Huffman, on Intl): Each market is different — that is a key learning.Machine translation is a good starting point to build a content base, but over time what matters most is attracting more local, locally-created and locally-consumed communities; results vary a lot by market.
We concluded we must focus on 'community success' — how easy it is to create and grow communities on Reddit; that is not just an Intl issue, it is true in the US as well.Making it easier to create and scale subreddits is a major product priority with ample runway everywhere.
On logged-in monetization, as Jen noted, ARPU is higher for logged-in users because we know them better and can personalize, which drives retention and time spent.More users on the app, more users logged in, and faster personalization all drive higher engagement and monetization. All roads lead to the same strategy: help users find relevant content so they return more often.
Q: Any updates on how gen-AI partners use Reddit data? Impact from Google algorithm changes?
A (CEO Steve Huffman): Reddit has been and remains the most cited source in AI across platforms, and 'Reddit' has sat in Google's top-10 search terms for years.AI and internet users love Reddit content because human validation and endorsement of AI outputs matter.
You can get a surface-level answer from AI, but many queries need context and diverse perspectives — different parts of an answer matter differently depending on the user.Even a simple question like 'what movie to watch tonight' depends on your preferences and age; Reddit excels at such answers. The more the internet is cleansed and summarized by AI, the more people crave Reddit's authentic human perspective.
On Google algorithm changes, this is par for the course and always hedged — sometimes it helps, sometimes it hurts, but rarely impacts our traffic in a sustained way.We did see some changes this quarter, but nothing further to highlight.
Q: Trends in content contribution rates? Beyond money, what matters when choosing new data-licensing partners?
A (CEO Steve Huffman): On the content ecosystem, our product strategy rests on three pillars discussed here — onboarding, performance/retention, and community ecosystem; contribution sits in the third.Communities attract users, users create content — a flywheel; user content attracts more users — another flywheel; Reddit has a big opportunity here. We track post success rate — the share of posts that persist (e.g., not removed by mods).
'Post guidance' is an LLM that helps users understand subreddit rules; we have made significant progress.We are improving posting UX and lifting account-age and Karma restrictions for new users — many communities historically restrict new posters, making new-user acquisition harder. We deploy better AI spam/abuse detection to protect communities from malicious new users while welcoming high-quality newcomers. Perhaps next quarter we will spend more time on communities and contribution.
On data-licensing partners, beyond money there are many dimensions — citation and mindshare matter, as do broader relationships.These firms have data centers and foundation models, so partnerships are multi-layered; machine translation, LLM-driven onboarding and safety tools are among the benefits we receive, making these relationships more than just commercial contracts.
Q: What is the impact of AnswerPlus/Answers search on session length? Which Intl markets have started monetizing?
A (CEO Steve Huffman, on search): Search is performing very well — search DAU, search WAU, and queries are all up materially YoY.Search is a key retention driver and a DAU contributor; the search team is executing strongly.
Reddit Answers is more deeply integrated, with more agentic behaviors like comparing two options (e.g., 'should I watch movie A or B').We are integrating product catalogs so that when you ask 'what are the best headphones', we can return product links directly. Search is now a core Reddit use case and contributes to nearly all the metrics we care about.
A (COO Jen Wong, on Intl monetization): We have direct sales in the US, Canada, UK, continental Europe, and Australia (with parts of APAC covered via our Singapore office), serving both enterprise and SMB/mid-market.Elsewhere, we work through channel partners, enabling cross-border campaigns. We reassess coverage regularly, guided by user growth by market and aligning monetization coverage accordingly.
<End>
Risk Disclosure and Disclaimer:Dolphin Research Disclaimer and General Disclosure
