RDDT: A billion-dollar guide meets a shrinking search funnel

The 60-second read

Five things to know

1. The revenue engine still works. Q3 revenue grew 53% to $895M and Q4 guidance reaches $1.00–1.03B.
2. The search warning intensified. Management says continuing referral declines could produce a low-double-digit sequential DAUq decline; this is conditional, not formal DAU guidance.
3. The valuable core offers a counterargument. Logged-in and app audiences grew sequentially and represent over 80% of monetization, with overlap between groups.
4. Compensation changes owner economics. $362.6M of FCF becomes a $231.3M analytical proxy after deducting $131.3M of SBC.
5. A cheaper price is not a complete thesis. HOLD, $110 fair value, very high uncertainty; $77 provides the stated 30% margin of safety absent deterioration.

Three scenarios, one probability-weighted number

Analyst scenarios, not company guidance; 2027 owner-cash capitalization at current cash and fully diluted shares.

Signposts: what would change our mind

Analyst thresholds, not management promises. Reassess definitions, acquisition costs and cash-funded buybacks before assigning a grade.

A lower price, a harder question

Reddit closed at $91.73 on October 30, down 29.22% from the previous close in the saved FMP daily-price series. That is a completed-session comparison, not an estimate of the initial after-hours response. A separate quote snapshot showed $91.67 and a slightly different volume; this packet consistently uses the daily-history close. The distinction matters when a dramatic move becomes the headline: the valuation should have one reproducible price rather than whichever print makes the argument look strongest.

The fall changed the investment question. At the earlier valuation, investors needed to defend rapid audience expansion and strong monetization simultaneously. At this price, the business can deliver a useful owner return with considerably less growth. That does not mean the traffic problem has become harmless. Search is both an acquisition channel and a way existing users return. Losing it may remove lower-value visits first, then impair the flow of tomorrow’s valuable app users. The stock is cheaper because the range of plausible outcomes has widened.

Our call is HOLD, conviction three out of five, with a $110 fair-value anchor and very high uncertainty. We use a 30% margin of safety below that rounded anchor, or $77, before considering a price-led upgrade with the operating evidence unchanged. A roughly 20% gap from price to fair value is insufficient by itself when a traffic reset can reduce both future cash generation and the multiple investors pay for it. A business-led upgrade could arrive earlier if direct engagement and monetization hold through the next quarter.

The previous Q2 episode focused on a strong print followed by a stock decline. This update asks a different question: can Reddit replace lost referrals with a durable logged-in habit without buying that growth at an uneconomic cost? The next results should test the answer. Price reactions shown below are descriptive, not causal proof that one disclosure explains every trade.

A strong quarter does not settle Q4

Revenue reached $895.0 million, up 53.0% from $584.9 million. Operating income was $261.6 million and net income was $283.3 million. The company generated $363.8 million of operating cash flow and $362.6 million of free cash flow after $1.2 million of capital expenditure. Those are substantial profits and cash receipts, not merely an adjusted earnings story. The quarter deserves credit even though the market focused on the outlook for traffic.

The growth was also economically productive. Gross margin was 91.5%, while adjusted EBITDA rose to $402.3 million. GAAP operating margin was 29.2%; adjusted EBITDA margin was 44.9%. Those measures answer different questions. The second excludes stock compensation, related taxes and depreciation, while the first recognizes their cost. Calling either margin simply “profit margin” would obscure the expense investors still bear. Net margin was 31.7% and benefited from interest income and a low effective tax rate.

Management’s prior Q3 revenue guide was $860–870 million and adjusted EBITDA guidance was $385–395 million. The delivered results exceeded both upper bounds. The new Q4 revenue guide is $1.000–1.030 billion, with adjusted EBITDA of $480–510 million. The midpoint implies a 48.8% adjusted EBITDA margin and full-year revenue of $3.378 billion when added to reported nine-month revenue. These are arithmetic implications of guidance, not additional management promises.

The apparent contradiction is the research opportunity: financial guidance is strong while management warns that active uniques could contract if search trends persist. Holiday advertising demand, higher yield, improved targeting and a more valuable audience mix can support revenue even as low-value visits disappear. But guidance is not evidence that this mechanism will work indefinitely. We would judge the next report on revenue, direct engagement and marketing efficiency together, rather than allowing an advertising beat to answer a separate retention question.

Audience quality versus audience size

Reddit reports one consolidated operating segment. U.S. and international revenue are geographic disclosures, not separately reported profit centers. We do not invent regional operating margins. U.S. revenue was $718.9 million and international revenue $176.1 million, up approximately 50% and 67% respectively. Advertising generated $847 million; other revenue was $48 million. The latter includes data licensing and other products, so it should not be relabeled as a pure artificial-intelligence licensing segment.

Global daily active uniques averaged 130.3 million, up 12% year over year but unchanged from Q2 at the reported precision. U.S. DAUq fell to 50.7 million from 51.6 million a year earlier and 53.2 million in Q2. International DAUq reached 79.6 million. These are unique identifiers under the company’s methodology, not a census of individual people. Weekly active uniques were 515.2 million, but dividing daily by weekly uniques is only a frequency proxy; it is not a measured cohort-retention rate.

Global ARPU rose to $6.87 from $5.04. U.S. ARPU was $13.82 and international ARPU $2.43. The geographic ARPU series allocates advertising according to where impressions are delivered, while the financial revenue geography uses customer billing location. Consequently, dividing the reported U.S. financial revenue by U.S. DAUq does not reproduce reported U.S. ARPU. We preserve the company’s definitions instead of “correcting” the numbers into a false reconciliation.

The strongest counterargument to the bear case is mix. Management says logged-in and app users represent over 80% of monetization and grew sequentially. Logged-in DAUq were 53.5 million globally, including 23.4 million in the U.S. That is evidence of resilience in a valuable audience, but the two groups overlap; their percentages cannot be added to estimate a separate population. The bullish thesis requires the protected core to keep growing while lost referrals stop starving the funnel that replenishes it.

Count compensation and tax honestly

The adjusted-to-GAAP bridge is straightforward and material. Start with $402.265 million of adjusted EBITDA. Subtract $136.419 million of stock-based compensation and related taxes and $4.260 million of depreciation and amortization to reach $261.586 million of operating income. Add $27.463 million of net other income, mainly interest, to reach $289.049 million of pretax income. Subtract $5.785 million of tax to reach $283.264 million of net income. The bridge demonstrates why a 44.9% adjusted margin is not the amount available to common owners after all economic costs.

Cash-flow stock compensation was $131.306 million, a narrower measure than the $136.419 million EBITDA add-back including related taxes. We use the narrower figure when calculating free cash flow less SBC, and explicitly identify that measure as an analytical proxy. It is neither GAAP nor Reddit’s disclosed free cash flow. The historical proxy still includes interest cash receipts and is distinct from the forecast operating owner cash that excludes treasury returns. It helps quantify the economic burden of issuing equity without pretending the accounting expense and a cash repurchase are interchangeable transactions.

The quarter’s effective tax rate was about 2.0%. Applying an illustrative 23% normalized rate to pretax income produces about $1.10 of diluted EPS versus reported $1.40. This is a sensitivity, not a forecast of next quarter’s tax provision. The current 10-Q says the company had about $1.7 billion of federal net operating loss carryforwards at December 2025, and discusses valuation allowances, tax-deductible stock compensation and utilization limits. Positive book retained earnings do not establish that tax losses are exhausted. Book accumulated deficit and tax NOLs are different records.

We also keep share denominators separate. Q3 diluted weighted-average shares were 201.860 million for EPS, while the company’s point-in-time fully diluted count was 206.3 million. Our valuation uses the latter. Applying the smaller EPS denominator to the same enterprise valuation would overstate the per-share result. A one-quarter decrease in fully diluted shares is welcome, but the reduction came with repurchase spending and does not mean compensation ceased to be a recurring cost.

Cash conversion passes; durability is unproven

The cash result is strong enough that a bearish analysis should acknowledge it directly. Operating cash flow exceeded net income despite an $80.0 million receivables cash outflow during Q3. Accounts receivable rose to $730.1 million, consistent with a larger advertising business and collection timing. Our ending-receivables proxy is about 75 days using 92 quarter days. It is a rough diagnostic, not the company’s reported days-sales-outstanding calculation, and can move with billing seasonality and geography.

Capital expenditure was exceptionally small because Reddit relies on third-party infrastructure. Asset-light does not mean infrastructure-free: hosting expenses flow through the income statement and contractual commitments still matter. The low capex line makes operating cash conversion attractive, but it should not be read as proof that all future growth can occur without incremental spending. Product work, moderation, safety, marketing and cloud capacity are recurring requirements of the platform.

Free cash flow less SBC was $231.3 million, compared with company-defined free cash flow of $362.6 million. The gap matters for valuation because shareholders ultimately fund employee compensation through dilution, cash buybacks, or both. We do not deduct the same compensation twice: the forward owner-cash margins used in our model are already after recurring compensation and normalized tax, and the model holds the fully diluted share count flat. That flat-share assumption requires the margin to cover the economic cost of keeping dilution controlled.

The nine-check scorecard below is an evidence organizer, not a mechanistic buy signal. Minority leakage and inventory are not material analytical issues for this single-segment digital platform. Tax normalization and traffic-dependent guidance deserve more attention than a generic checklist score. The company exceeded the prior revenue and adjusted EBITDA ranges, but one successful quarter is not a guarantee that the new revenue guide survives a faster search deterioration. We want the quality assessment to distinguish cash already collected from the cost of sustaining that cash in future years.

The balance sheet buys time

Cash, equivalents and marketable securities totaled $3.033 billion at September 30. There was no funded debt drawn, although lease obligations and ordinary operating liabilities remain. On 206.3 million fully diluted shares, the cash-and-securities balance is $14.70 per share. That is a meaningful cushion, not a liquidation floor: a platform can spend cash defending distribution and product relevance, and equity value still depends primarily on the operating franchise.

At the $91.73 reference close, fully diluted equity value is $18.924 billion and enterprise value after cash is $15.890 billion. This differs from the provider’s roughly $17.637 billion quoted market capitalization, which uses its own ordinary share-count basis. We do not mix the provider market cap with a more conservative fully diluted per-share denominator. The enterprise-value bridge is built consistently from the share count actually used to divide our modeled equity value.

Reddit spent about $100 million repurchasing roughly 670 thousand shares in Q3 at an average $149.81. The latest reference price is roughly 39% below that average. This is not a realized loss on an asset held for resale, but it is relevant capital-allocation evidence: the same dollars would retire more shares at today’s price. Fully diluted shares declined 0.3% quarter over quarter; excluding repurchases the company says the count was approximately flat. That is better than uncontrolled dilution, but maintaining it is not free.

The balance sheet gives management room to test app conversion and paid acquisition without immediate financing stress. It does not establish that paid acquisition produces attractive lifetime value. We would prefer measured retention, engagement and acquisition economics to broad claims that acquired users behave like organic users. Similar initial retention can be encouraging while still leaving differences in acquisition cost, ad exposure and lifetime monetization. The cash cushion earns a higher survival probability in the bear case; it does not eliminate the bear case.

Three routes, one explicit judgment

We triangulate rather than presenting one precise spreadsheet output as truth. Route one is a five-year owner-cash DCF. Revenue assumptions for 2027–2031 are $3.700, $4.255, $4.808, $5.337 and $5.817 billion. Owner-cash margins are 24%, 26%, 28%, 29% and 30%, after recurring stock compensation and normalized tax, excluding treasury interest and investment returns. Cash is added separately once. Discounting those cash flows at 12%, with 3% perpetual growth, then adding current net cash and dividing by 206.3 million shares gives $91.97. Terminal value remains important, so the result is highly sensitive to durable margins and the required return.

Route two capitalizes 2027 owner cash of $888 million at 24 times, adds $3.033 billion of net cash, and divides by 206.3 million shares: $118.01. The owner-cash forecast is $3.7 billion revenue times a 24% margin. This route assumes the business retains a premium to mature advertising assets because of its unique content and growth runway. It is not a market-derived peer median. Reducing the multiple is a direct way to price weakened distribution control.

Route three is an enterprise-sales cross-check: $3.7 billion of revenue times 5.5, plus net cash, divided by the same share count gives $113.35. Sales multiples ignore how much growth costs, so this route receives no privileged status. It is useful only because the assumed multiple is compatible with a meaningful owner-cash margin. A revenue franchise that needs permanently higher acquisition spending would deserve a lower multiple even if reported sales stayed strong.

The three values average $107.78 with equal weights. We round that to a $110 communication anchor and use the $92–118 route band, while explicitly retaining the much wider scenario outcomes. Equal weights are a judgment that avoids letting the optimistic capitalization route dominate the more cautious DCF. The 30% required margin of safety gives $77. The current price is near our DCF value, so a HOLD can be appropriate despite headline upside to the rounded blended anchor.

At the current fully diluted EV, a 24-times owner-cash multiple implies about $662 million of sustainable annual owner cash. At a 24% margin that corresponds to approximately $2.76 billion of revenue. This is not a reverse DCF and does not assert the market literally forecasts that revenue. It is a transparent capitalization-equivalent stress test. The price already allows for weakness; the unanswered question is whether it allows enough for a persistent acquisition-channel reset.

Fresh target cuts beat a stale average

The provider’s aggregate target consensus was $192.90, with a $120 low, $265 high and $197.50 median at retrieval. The ratings snapshot showed one strong buy, seventeen buys and twelve holds. Those aggregate figures mix analyst update dates and should not be described as a unanimous post-earnings judgment. A dramatic reset makes this distinction especially important: a high average can remain on screen long after the freshest analysts have reduced their targets.

The date-stamped target news is more useful for this print. On October 30, FMP carried Bernstein at $120, Cantor Fitzgerald at $120, Guggenheim at $170 and Needham at $150. We report the firm, date and target without inventing a rating where the supplied item does not establish one. These are provider records of analyst actions, not our verification of the full brokerage reports. The four-target subset is not the whole Street and should not replace the aggregate with an equally misleading small-sample average.

The annual revenue-consensus snapshot was approximately $3.388 billion for 2026 and $4.368 billion for 2027. The 2026 figure is close to the $3.378 billion implied by management’s Q4 revenue midpoint. Our $3.7 billion 2027 assumption is deliberately below the provider consensus because it prices a slower transition from search-supplied discovery to direct engagement. The model still grows revenue; it does not assume the franchise collapses simply because one distribution source deteriorates.

Provider EBITDA, EBIT and EPS estimates may follow adjusted conventions that are not established by the endpoint. We therefore compare revenue directly and do not present a false GAAP-versus-consensus EPS beat. The useful disagreement with Wall Street is a measurable forecast difference: can monetization and direct habit support a higher revenue path without sacrificing owner-cash margins? Our lower anchor reflects skepticism about that combination, rather than an attempt to match whichever brokerage target appears most recent.

From referral engine to daily habit

Management’s response is more substantive than simply denying the referral decline. The shareholder letter describes improved feeds, app performance, video experiences, community tools and efforts to turn mobile-web visitors into app users. It says only about 13% of eligible mobile-web users currently see the app-conversion prompt. That leaves room to expand prompts, but it does not establish how many visitors will convert or whether repeated prompts will reduce engagement among people who prefer the web.

The proposed logged-in experience is the most consequential product change. Tests in selected non-U.S. markets begin in Q4, with results informing a broader version. Logging in can improve personalization, ad targeting, data-access control and regulatory compliance. It can also introduce friction before a casual visitor finds value. We treat this as an experiment with a plausible economic upside and a plausible audience cost, not as a guaranteed fix for Google-related traffic pressure.

Management also reports that newly acquired app users retain similarly to organically acquired app users in early observations. That is encouraging evidence, but acquisition cost is missing from that comparison. Sales and marketing expense was $209.4 million, up sharply from $128.7 million a year earlier. Some of that supports advertiser sales and other activities, so attributing the full increase to paid user acquisition would be wrong. The disclosure nonetheless makes efficiency a necessary question as management increases its investment.

Governance remains concentrated. Class A common carries one vote, Class B ten votes and Class C generally no votes. The current 10-Q estimates that Steve Huffman could vote approximately 75% of combined voting power through his holdings and the Advance voting arrangement. Outside holders cannot assume they can force a different product strategy. We incorporate that constraint through uncertainty and a required margin of safety, rather than adding an unexplained second discount after valuation.

1. What share of newly acquired app cohorts returns and monetizes after ninety days?
2. How does lifetime ad contribution compare with paid acquisition cost?
3. What fraction of lost search visitors would otherwise become logged-in users?
4. What are the test guardrails for login friction, total engagement and ad yield?
5. How will management report audience mix consistently through the product transition?

A conditional warning with real consequences

The most important new evidence is in the shareholder letter, not the earnings headline. Management says the search decline accelerated late in Q3 and continued into October, with average weekly referrals falling by more than one million. It warns that if the pace continues, Q4 total DAUq could fall by a low-double-digit percentage from the current roughly 130 million base. This is a conditional management risk statement. It is not formal numerical DAU guidance, and our model does not silently convert it into a guaranteed outcome.

For an explicit stress illustration, a 12% sequential DAUq decline would leave about 114.7 million uniques. Producing the $1.015 billion revenue-guide midpoint on that denominator would require global ARPU of about $8.85, approximately 28.8% above Q3. Twelve percent is our chosen interpretation for sensitivity testing, not management’s precise forecast. Holiday seasonality and removal of lower-value visits can raise ARPU mechanically, so the calculation is not proof the revenue guide is impossible. It shows how much weight the revenue outlook places on mix and monetization.

The bull case is credible: Reddit has distinctive communities, a large corpus of human discussion and improving advertising tools; the high-value core could remain healthy while less valuable search traffic declines. The bear case is also credible: search brings both new users and returning users, so the damage can spread beyond logged-out visits and require more marketing to replace. The relevant long-run test is owner cash per fully diluted share, not an isolated headline ARPU record.

Other material risks include advertising cyclicality, brand safety, dependence on volunteer moderators and user contributions, privacy regulation, AI-driven substitution and concentrated voting control. Data licensing can monetize the corpus while AI products simultaneously reduce click-through traffic. These effects can coexist; revenue from a licensing customer does not guarantee that the customer preserves Reddit’s distribution. The balance sheet reduces financing risk, but it does little to prevent audience behavior from changing faster than product development.

Define what would change the call

The next scheduled operating checkpoint is the Q4 results release, expected in the normal early-2027 reporting cycle; an exact announcement date is not established in the saved evidence. We use February 2027 as a review window, not as a promised earnings date. The revenue guide provides a clear first test: at least $1.000 billion would meet its lower bound, while adjusted EBITDA of at least $480 million would meet the accompanying lower bound. Hitting those measures would validate near-term monetization, but would not by itself settle the traffic thesis.

We would also look for the valuable audience to hold. Global logged-in DAUq of at least 53.5 million and U.S. logged-in DAUq of at least 23.4 million would preserve the Q3 baseline. Management should explain any definition change, disclosure removal or overlap with app metrics before a comparison is graded. A blended global DAU headline can improve because lower-monetizing international audiences grow even while the U.S. core weakens, so geographic and engagement context are necessary.

The economic checkpoint is after-compensation cash generation. We would prefer free cash flow less SBC to remain positive and fully diluted shares at or below roughly 207 million, with an explanation of the cash needed to achieve that count. These are analyst thresholds, not management commitments. A share decline funded by aggressive repurchases at an excessive price is different from stable dilution supported by growing owner cash.

The price checkpoint is independent. At $77, the existing $110 anchor would offer the stated 30% margin of safety, provided the research facts had not deteriorated. A lower price caused by permanently lower cash generation can reduce fair value as well, so the threshold is not an automatic order. Conversely, stronger evidence of direct habit could justify paying more before every traffic indicator turns positive. The discipline is to update both the forecast and the required return, rather than defending a prior target.

Sources, definitions and reproducible data

All dollar financial-statement values in the tables below are in millions unless a per-share label says otherwise. The underlying dataset retains the source URL for each quarter, SEC XBRL accession traces for extracted GAAP values and computed-field labels for margins, cash flow and valuation. Nine quarters run from Q3 2024 through Q3 2026; three fiscal years cover 2023–2025. Annual figures come from the audited 10-K financial statements, with quarter-four cash-flow values calculated as full year less the first nine months where necessary.

The 2024 stock-compensation spike reflects recognition associated with the IPO and should not be extrapolated as a recurring annual cash burden. Equally, excluding that historical spike does not justify eliminating current stock compensation from owner economics. The model’s forward margin is after recurring compensation and normalized tax, while the initial fully diluted share denominator includes outstanding awards under management’s disclosed measure. No prospective share-price performance is presented as a reported result.

Reddit is a U.S. issuer listed on the New York Stock Exchange, with Class A common as the covered security. It is not an ADS and there is no ADS conversion ratio in the model. The business is a single reportable segment. Geographic revenue and audience statistics provide useful operating detail, but geographic profit is not separately disclosed. Inventory days and minority-profit leakage are not meaningful drivers on the presented statements; we retain those limitations instead of manufacturing a complete-looking segment table.

Sources are the October 29, 2026 earnings release and shareholder letter furnished with the 8-K, the September 2026 10-Q, prior quarterly releases, the 2025 and 2024 10-Ks and saved FMP market endpoints. The 10-Q reports effective disclosure controls and no material change in internal control during the quarter. Generic future control-risk language is not evidence of a current material weakness. The source review separately verified the continuing common security and exact fiscal period.

For terminology, adjusted EBITDA is management’s non-GAAP operating measure; free cash flow is operating cash flow less purchases of property and equipment; owner cash is our analytical estimate after compensation and normalized tax; enterprise value is modeled equity value less cash and securities here; DAUq and WAUq are daily and weekly unique identifiers; ARPU is revenue per average daily active unique using the company’s geographic allocation. Probabilities, multiples, discount rates and margin-of-safety thresholds are analyst assumptions. They make the call testable, not certain.

• Q3 2026 earnings release, Exhibit 99.1
• Q3 2026 shareholder letter, Exhibit 99.2
• September 2026 Form 10-Q
• 2025 Form 10-K
• 2024 Form 10-K
• FMP daily price history, quote, profile, target news, target consensus, ratings and annual estimates; retrieved November 1–2, 2026. Saved endpoint responses and hashes accompany the dataset.