conf 2/10 partial
enthusiasm:18.0 · trend:-5 · quantifies:0 · impact:0 · under_radar:5 · credibility:0 · business_impact:8 · disruption:0 · commitment:0 · confirmation:0
Enthusiasm latest 6 / prev 7 (falling)
DraftKings frames AI/ML mainly as an enabler of sports pricing, risk/market-making, promo optimization, and internal productivity—not as a separate revenue line. Enthusiasm was stronger and more explicit in Q4 (CEO-led AI/ML narrative and promo-engine rollout) and thinner in Q1 (one quantified productivity claim; modeling cited for predictions market-making). The only hard number tied to AI language is 2–3× team output; broader margin/handle gains are discussed as data-driven optimization without consistent AI attribution, so credibility rests on modeling depth and early promo AI rather than disclosed AI P&L.
Grounded on actual base — revenue $6.1B · net income $0.0B · net margin 0.1% · diluted EPS -0.0081
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: % · vs analysts: unclear · priced in: medium (model's call-read: high; verdict above is the hard-data one used for ranking) · confidence: 2/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|---|---|---|---|
| Team productivity 2–3× output (some teams) productivity · soft | 2 to 3x last year's output (some teams) | Partial output multiplier, not a $ or company-wide %. No disclosed payroll/opex $ for affected teams nor their share of FY2025 revenue (6,054,525,000), so rev_uplift_pct=100×Δrev/6,054,525,000 cannot be formed. No $ saving stated; net margin ~0.06% and diluted EPS −0.0081 make any EPS-uplift% a denominator artifact (thin-margin/loss guardrail). | ||
| Hundreds of DS/MLE + trading desk (modeling scale) other · soft | hundreds (data scientists and ML engineers) | Capability/headcount scale only; 'hundreds' is unbounded, no compensation per FTE and no revenue/cost $ attached. Cannot map to a revenue line or compute a saving without inventing a $ base. | ||
| Promo-engine AI → efficiency and top-line other · soft | expected to improve efficiency and top-line results (no $/%) | Forward qualitative only; no incremental $ revenue, bps take-rate, or opex $ disclosed for FY2026 phasing. | ||
| Predictions leadership / scaling modeling & trading engagement · soft | scaling sports modeling, trading, market-making, combo products (no $/%) | Strategic intent without quantified bookings, revenue, or margin; cannot map to rev_uplift_pct=100×claim_$/6,054,525,000. | ||
| Top global sports market maker ambition revenue · soft | ambition to be a top global sports market maker (no $/%) | No disclosed incremental revenue/bookings or conversion schedule; market-maker economics not sized in the call. |
Assumptions: Next fiscal year = FY2026 (ending 2026-12-31) vs FY2025 actual base (2025-12-31). Any hypothetical revenue flow would default to 21% tax and current net margin; rejected as misleading given operating income −15,817,000 (operating margin −0.26%) and near-zero/negative GAAP earnings, and no revenue claim exists to apply it to. EPS sizing would use consensus adjusted EPS (FY2025 epsAvg −0.15612, NI −75,267,702), not the near-zero GAAP base, per the thin-margin/meaningless-EPS% guardrail. Productivity 2–3× cannot be phased to FY2026 revenue without a team-weight %. All claims are own-operations → adopter, B2C OSB/iGaming revenue line; no supplier-side AI-compute revenue exists.
Top line: No quantifiable AI revenue uplift: no $ bookings/revenue, bps, or phasing in any claim. Both quantified items (2–3× team output; hundreds of DS/MLEs) are productivity/capability statements, and the forward items (promo engine, predictions, market-making) are purely qualitative. Consensus already embeds FY2026 revenue 6,790,084,941 vs FY2025 actual 6,054,525,000 → Δ735,559,941 = +12.15%; AI is plausibly supportive of that path but is not separately sizeable.
Bottom line: No hard EPS uplift: productivity is subset-of-teams with no $ opex save; 'hundreds' headcount has no $. GAAP NI ~3.71M on 6.05B revenue (0.06% margin), operating income −15.8M, diluted EPS −0.0081 — a near-zero/depressed base where any modest save explodes as EPS% (artifact), so est_eps_uplift_pct is null. Consensus FY2026 net income ≈123.4M (epsAvg 0.24249 × ~495.9M shares) vs FY2025 consensus −75.3M; the AI-first/streamlined-teams narrative most likely shows up as that modeled margin swing, not as a separately attributable EPS lift.
[impact n/m (all claims soft/unanchored)] Aggregated hard AI uplift is null (all claims soft). Street FY2026 vs FY2025 actual: +12.15% revenue (735.6M), NI swing from consensus −75.3M to +123.4M (+198.7M); FY2027 NI extends to ~451.2M (cumulative FY2027 vs FY2025 actual revenue +25.74%). That profitability inflection and double-digit growth are exactly what 'streamlined, AI-first teams at 2–3× output' would drive, so the unquantified AI productivity story sits inside numbers analysts already carry. With zero hard dollar anchors from management, the math points to no demonstrable gap above consensus.
partial
Full agreement on every verdict and all-null pcts; only difference was claim count and which earnings base to narrate. Adopter, priced-in high, confidence 2.
| Field | Opus 4.8 | GPT-5.5 |
|---|---|---|
| Rev uplift % | – | – |
| EPS uplift % | – | – |
| Priced in | high | – |
| vs analysts | unclear | – |
| Confidence | 2 | – |
| Top line | No quantifiable AI revenue uplift. Both quantified claims (2-3x team output; hundreds of DS/MLEs) are productivity/capability statements with no dollar figure and no disclosed base, and the three forward-looking items (promo-engine efficiency, predictions leadership, market-maker ambition) are purely qualitative. AI is plausibly supportive of the consensus +12.2% FY26 revenue path (6,054.5M -> 6,790.1M), but none of it is separately sizeable from the inputs. | – |
| Bottom line | EPS-uplift % is not meaningful here: diluted EPS is negative (-0.0081), operating income is negative (-15.8M), and net margin is 0.06% — a near-zero/depressed earnings base where any saving would produce an artificially huge % from the tiny denominator. Per the loss-making/thin-margin guardrails, est_eps_uplift_pct is set to null. The 'AI-first execution / streamlined teams' productivity narrative most likely shows up as the margin swing consensus already models (net income -75.3M -> +123.4M FY26), not as an incremental, attributable EPS lift. | – |
| Reasoning | Consensus already embeds a sharp profitability inflection — revenue +12.2% (6,054.5M -> 6,790.1M) and net income from -75.3M to +123.4M FY26, EPS -0.156 -> +0.242, extending to +0.947 FY27. That operating leverage is precisely what 'streamlined, AI-first teams at 2-3x output' would drive, so the (unquantified) AI productivity story sits inside numbers analysts already carry. With zero hard dollar anchors from management, the math points to no demonstrable gap above consensus. | – |
Rows highlighted where the two models disagreed.
—/100 (no quantified promises) no-quantified-promises 6 calls reviewed
Across six calls (Q4 FY2024–Q1 FY2026), DKNG discussed AI/ML qualitatively (AI-first culture, trading agents, model-driven Predictions) but never stated a measurable AI target with both a number and a dated milestone; productivity and Predictions metrics cited later were results or investment guidance, not prior quantified AI pledges to score.
Est. revisions flat · Fwd P/E -162.1 · EV/Sales 2.0x
AI claim maps to UNITED STATES, Product and Service, Other
AI enthusiasm across 6 calls — trend ↗ rising
AI absent early, briefly noted mid-2025, then conviction-backed LTV, modeling, and productivity story.
8/10 qualitative impact material near-term · mixed evidence
Where AI matters: sportsbook pricing, risk/market-making, promo optimization
Hundreds of in-house DS/MLEs, live trading, and profitable in-house market-making tie ML directly to hold, liquidity, and promo ROI—the core sportsbook/Predictions economics—not a separate AI product. Only hard metric is 2–3× output on some teams with no $ P&L attribution, so upside is real but not separately quantifiable above consensus.
Caveats: No management-disclosed AI revenue, margin bps, or opex $—productivity and modeling claims stay soft vs Street inflection; Promo-engine AI described as just starting with no phased $/bps impact; GenAI may accelerate sharp-bettor/syndicate edge and AI-native prediction rivals, compressing modeling advantage
AI DISRUPTION / CANNIBALIZATION RISK tailwind · 3/10
AI does not replace regulated wagering or automate away the book; revenue still comes from handle, vig, and licensed distribution. Sharper bettor tooling and operator model parity can pressure hold over time, but DKNG’s scale data and in-house stack are more offensive moat than structural cannibalization of what they sell.
option liquidity: good
proxy inputs — dollar-ADV $330M · beta 1.666 · px $25.30
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.