← back to rankingFDS · FactSet Research Systems Inc.
Financial - Data & Stock Exchanges · mkt cap $9.3B · calls: Q3 FY2026 vs Q2 FY2026
83.0 conviction · conf-adj 83
conf 6/10 🚀 reported partial
enthusiasm:27.0 · trend:8 · quantifies:5 · impact:0 · under_radar:14 · credibility:12 · business_impact:8 · disruption:0 · commitment:6 · confirmation:3
Enthusiasm latest 9 / prev 8 (rising)
FactSet's AI thesis is unusually well-evidenced for an information-services company: management pairs every strategic claim with discrete, verifiable metrics — MCP API volumes up 13x QoQ to 450+ paying/trial clients, AI SKUs directly contributing >10% of ASV growth (up from ~0% a year ago), and coding agents authoring 27% of committed code enabling a ~10% technology workforce reduction. The company is executing a two-sided AI flywheel in which internal productivity gains (50%+ cut in data-extraction touch time, 80% vectorization cost reduction, 25%+ manual curation savings) fund external AI product investment, while Google Cloud, Finster, Anthropic, and TIFIN.AI partnerships expand distribution and capability — with a measurable multiplier effect where MCP consumption demonstrably pulls through upsizing of core workstation and data-feed subscriptions. Credibility is reinforced by >95% ASV retention, 30% average contract term extension achieved without price compression, and the empirically observed shift toward enterprise agreements, though the emerging consumption-pricing layer introduces ASV forecasting uncertainty management has openly acknowledged.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $2.3B · net income $0.6B · net margin 25.7% · diluted EPS 15.55
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: 0.64% · next-FY EPS uplift: 3.5% · vs analysts: inline · priced in: low (model's call-read: medium; verdict above is the hard-data one used for ranking) · confidence: 6/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
AI SKUs >10% of ASV growth
revenue | >10% of ASV growth | consensus FY25->FY26 rev growth ~$147M as ASV-growth proxy; 10% direct AI rev = ~$14.7M / $2,321.7M = ~0.64%. EPS uses avg of current 25.7% net margin (X) and 40% software incr margin (Y); $14.7M x ~33% = ~$4.9M / $655.8M adj NI = ~0.74% EPS. Phasing: Q3 run-rate as FY26 floor; broader ASV halo not quantified. | 0.64 | 0.74 |
ASV growth 50% higher for AI clients
engagement · soft | 50% higher QoQ ASV growth | growth-rate differential disclosed; AI-client ASV share not isolable; correlation/halo, would double-count direct AI SKU line | | |
MCP 450+ engaged clients (vs 120+)
engagement · soft | 450+ vs 120+ | adoption breadth, no ACV/conversion disclosed; leading indicator | | |
MCP API call volume 13x QoQ
engagement · soft | 13x QoQ | usage velocity, no $/API-call or revenue linkage disclosed | | |
~90% of MCP deals show contract value uplift
engagement · soft | ~90% of deals | uplift frequency, no avg uplift % or deal count/base; feeds the >10% AI ASV line | | |
>20% of top 100 clients pay for MCP
engagement · soft | >20% | penetration rate, no top-100 ASV base disclosed | | |
~20% of MCP endpoint users net new
engagement · soft | ~20% net new | TAM-expansion/user-mix signal, no endpoint count or ARPU | | |
Top-10 bank doubled data subscriptions
revenue · soft | 2x, multiyear | single-client anecdote, client's FactSet spend not disclosed; inside the >10% AI ASV figure | | |
Top hedge fund grew 6x via MCP
revenue · soft | 6x | single-client anecdote, starting ASV not disclosed; inside direct AI ASV | | |
Cap Markets Intelligence: 30+ of top 100 banks in trial
engagement · soft | 30+ trials | pipeline count only, launched mid-Q3, no win-rate/ACV | | |
Top 50 clients: >90% use 4+ AI solutions; 48/50 on 3+
engagement · soft | >90% / 48 of 50 | adoption depth, no incremental ASV per solution disclosed | | |
100 bps in-year productivity savings (>50bps secured)
cost | 100 bps | 0.01 x $2,321.7M disclosed rev = $23.22M pre-tax; x0.79 (21% tax) = $18.34M / $655.8M adj NI = 2.8% EPS; topline ~0. Disclosed base -> HARD. | | 2.8 |
AI coding agents author 27% of committed code
productivity · soft | 27% | engineering leverage; dollar effect rolls into 100 bps target — sizing separately double-counts | | |
Coding token use 5x / AI code lines ~10x QoQ
productivity · soft | 5x / 10x | input-usage velocity, no cost base; rolls into 100 bps | | |
~10% technology workforce reduction
cost · soft | ~10% | tech FTE count & loaded comp not disclosed; savings roll into 100 bps aggregate — null to avoid double-count | | |
Data table extraction touch time -50%
productivity · soft | >50% | operator-time metric, no cost pool disclosed; feeds 100 bps | | |
Fundamentals team headcount -5%
cost · soft | 5% | team size/cost not disclosed; feeds 100 bps aggregate | | |
Banker onboarding +22% interactions, +5 NPS
engagement · soft | 22% / +5 NPS | capacity+satisfaction, no churn/ASV retention $ linkage | | |
25% of engineer capacity freed (AI-enabled teams)
productivity · soft | 25% | capacity redeployment, not realized cost; feeds 100 bps | | |
BAU upgrade/patching effort -90%
productivity · soft | >90% | activity reduction in subset of eng work, no absolute spend; feeds 100 bps | | |
New-product dev cycle month->1 day
productivity · soft | 30x | time-to-market, no $ base; feeds 100 bps | | |
Private-co classification 4x at flat cost
productivity · soft | 4x YoY | unit-economics improvement, no absolute cost base; feeds 100 bps | | |
Manual curation -25%+ across 4 AI tools
productivity · soft | 25%+ | curation effort, no cost base; feeds 100 bps | | |
Vectorization cost -80%
cost · soft | 80% | unit-cost metric, absolute base not disclosed; feeds 100 bps | | |
Assumptions: Earnings sized off consensus ADJUSTED NI $655.8M / adj EPS $17.08 (GAAP NI $597.0M ignored to avoid inflated %). Incremental net margin 40% on AI-SKU subscription revenue (above 25.7% net / ~32% op margin, reflecting high drop-through of incremental data/software ASV). Tax 21%. Phasing: full next-FY (FY2026) run-rate for both the >10% AI-ASV contribution and the 100 bps productivity target (mgmt says on track to deliver full 100 bps in H2 FY26). ASV-growth dollar base proxied by consensus FY25->FY26 revenue growth ($149.7M) since ASV growth $ not separately disclosed. All 20+ productivity/headcount claims are treated as components rolling into the single 100 bps aggregate and nulled individually to avoid double-counting.
Top line: Direct AI-SKU revenue is small but real: >10% of ASV growth from AI = ~$15.0M of the ~$149.7M annual ASV/revenue gain = +0.65% of the $2,321.7M base. Management flags a 'much bigger' indirect halo (AI clients growing ASV 50% faster than the rest of the book, 450+ MCP clients, API volume 13x QoQ, 90% of MCP deals lifting contract value), which is directionally powerful but not cleanly sizeable — it could push the broader AI-attributable uplift toward 1.5-2%, but that portion stays soft.
Bottom line: The clean cost anchor is the 100 bps in-year productivity target: 0.01 x $2,321.7M = $23.2M pre-tax, $18.3M after-tax (21%) = +2.8% on adjusted NI. Adding the 40%-margin drop-through on the $15.0M direct AI revenue (~$6.0M NI = +0.9%) gives a total adopter-side EPS uplift of ~3.7%. The dozen+ efficiency claims (27% of code AI-authored, ~10% tech-workforce cut, Fundamentals -5%, vectorization -80%) are the mechanism behind that 100 bps, not additive to it.
The quantified impact (~0.65% revenue, ~3.7% adj-EPS) fits inside what consensus already assumes: FY26 revenue growth of +6.46% comfortably contains a 0.65% AI-SKU direct contribution, and FY26 adj-EPS growth of only +3.9% is largely explained by the 2.8% productivity uplift plus operating leverage — i.e. analysts appear to have banked the guided 100 bps. The genuinely un-priced optionality is the flywheel/structural story (Google Gemini partnership, MCP 13x QoQ, 20% net-new users, roadmap to buy-side/wealth) which consensus partly reflects in the +10.3% FY27 EPS step-up. Net: near-term quantified math is roughly inline/slightly supportive, not a clear beat; the upside is unquantified optionality, so priced_in is medium rather than low.
MODEL CONSENSUS (impact)
partial
Core AI SKU + 100bps savings are the only HARD lines; all engagement/anecdote metrics stay soft. Aggregate EPS ~3.5% = 0.74 rev-flow + 2.8 cost.
Conflicts reconciled
- AI SKU rev_uplift_pct: X=0.62 vs Y=0.65 -> used 0.64 (avg of $144.3M/$149.7M growth proxies)
- AI SKU eps_uplift_pct: X=0.57 (25.7% margin) vs Y=0.9 (40% margin) -> used 0.74 (avg; both defensible)
- 100bps productivity line: X omitted vs Y HARD 2.8% -> kept Y, base is disclosed revenue
- ~90% MCP deals type: X=revenue vs Y=engagement -> engagement (uplift frequency, no $ base)
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | 0.65 | – |
| EPS uplift % | 3.7 | – |
| Priced in | medium | – |
| vs analysts | inline | – |
| Confidence | 6 | – |
| Top line | Direct AI-SKU revenue is small but real: >10% of ASV growth from AI = ~$15.0M of the ~$149.7M annual ASV/revenue gain = +0.65% of the $2,321.7M base. Management flags a 'much bigger' indirect halo (AI clients growing ASV 50% faster than the rest of the book, 450+ MCP clients, API volume 13x QoQ, 90% of MCP deals lifting contract value), which is directionally powerful but not cleanly sizeable — it could push the broader AI-attributable uplift toward 1.5-2%, but that portion stays soft. | – |
| Bottom line | The clean cost anchor is the 100 bps in-year productivity target: 0.01 x $2,321.7M = $23.2M pre-tax, $18.3M after-tax (21%) = +2.8% on adjusted NI. Adding the 40%-margin drop-through on the $15.0M direct AI revenue (~$6.0M NI = +0.9%) gives a total adopter-side EPS uplift of ~3.7%. The dozen+ efficiency claims (27% of code AI-authored, ~10% tech-workforce cut, Fundamentals -5%, vectorization -80%) are the mechanism behind that 100 bps, not additive to it. | – |
| Reasoning | The quantified impact (~0.65% revenue, ~3.7% adj-EPS) fits inside what consensus already assumes: FY26 revenue growth of +6.46% comfortably contains a 0.65% AI-SKU direct contribution, and FY26 adj-EPS growth of only +3.9% is largely explained by the 2.8% productivity uplift plus operating leverage — i.e. analysts appear to have banked the guided 100 bps. The genuinely un-priced optionality is the flywheel/structural story (Google Gemini partnership, MCP 13x QoQ, 20% net-new users, roadmap to buy-side/wealth) which consensus partly reflects in the +10.3% FY27 EPS step-up. Net: near-term quantified math is roughly inline/slightly supportive, not a clear beat; the upside is unquantified optionality, so priced_in is medium rather than low. | – |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
AI SKUs direct contribution to ASV growth: >10% of ASV growth, up from ~0% a year prior (Q3 FY2026, topline)
“Just in this quarter, over 10% of the ASV growth came directly from AI SKUs, and there was a much bigger impact in the broader ASV growth as well.”
ASV growth rate differential — AI solution clients vs. non-AI clients: 50% higher QoQ ASV growth for AI solution clients (Q3 FY2026, topline)
“quarter-over-quarter, overall ASV growth among clients using our AI solutions was 50% higher than for the rest of the book”
MCP active client engagement: 450+ clients (contracts + trials) vs. 120+ one quarter prior (Q3 FY2026 vs. Q2 FY2026, topline)
“Our MCP server has over 450 clients actively engaged under contracts and trials.”
MCP API call volume growth QoQ: 13x quarter-over-quarter (Q3 FY2026 vs. Q2 FY2026, topline)
“API call volume is experiencing rapid growth with Q3 volumes at 13x the level we experienced in Q2.”
MCP deal-level contract value improvement rate: ~90% of deals with MCP component show contract value uplift (Q3 FY2026, topline)
“MCP is a real accelerant — we see whenever there are deals with an MCP component, more often than not, in about 90% of cases, we've seen contract value improvements.”
MCP paid adoption among top 100 clients: >20% (Q3 FY2026, topline)
“Over 20% of our top 100 clients are using MCP on a paid basis.”
Net new user personas reached via MCP endpoints: ~20% of MCP endpoint users are net new (not previously FactSet users) (Q3 FY2026, topline)
“around 20% of the users of our endpoints in both MCP trials and paid implementations are net new users, whether in existing clients or new clients.”
Top-10-bank data subscription expansion driven by AI: 2x (doubled) on multiyear terms (Q3 FY2026, topline)
“One of the top 10 banks literally doubled their data subscriptions with us because of AI — multiyear contracts.”
Hedge fund ASV expansion via MCP: 6x growth (Q3 FY2026, topline)
“A top hedge fund grew 6x with us because of our MCP delivery.”
Capital Markets Intelligence banking pipeline: Active or pipeline trials at 30+ of top 100 banking clients (Q3 FY2026 (product launched mid-quarter), topline)
“We are seeing strong early engagement with active or pipeline trials at over 30 of our top 100 banking clients.”
AI solutions adoption — top 50 clients: >90% using 4+ AI solutions (Q3); 48 of 50 using 3+ (Q2) (Q3 and Q2 FY2026, topline)
“[Q3] Over 90% of our top 50 clients are now using 4 or more FactSet AI solutions. [Q2] Today, 48 of our top 50 clients are using at least three of our AI solutions.”
AI-authored share of committed code: 27% (Q3); ~20% / 'nearly one fifth' (Q2) (Q3 and Q2 FY2026, bottomline)
“[Q3] Coding agents now author 27% of committed code. [Q2] AI coding assistance now author nearly one fifth of our successful code commits.”
AI coding token use growth: 5x QoQ; AI-written code lines ~10x QoQ (Q3 FY2026, bottomline)
“coding-related token use grew 5x quarter-over-quarter, while committed lines of AI-written code grew almost 10x.”
Technology workforce reduction (AI-enabled): ~10% (Q3 FY2026, bottomline)
“we initiated a roughly 10% reduction in our technology workforce and freed up significant capacity to accelerate strategic product development.”
Data table extraction operator touch time reduction: >50% (Q3 FY2026, bottomline)
“we have reduced operator touch time for data table extraction by more than 50%.”
FactSet Fundamentals team headcount reduction: 5% (Q3 FY2026, bottomline)
“Within FactSet Fundamentals, we've consolidated multiple data pipelines into one, allowing us to redeploy significant capacity and reduce the size of this team by 5%.”
Banker digital onboarding — consultant live interaction uplift and NPS: 22% QoQ increase in live interactions; 5-point NPS gain (Q3 FY2026, both)
“approximately 4,000 bankers used our digital onboarding tools, and the capacity unlocked resulted in a 22% quarter-over-quarter increase in live user interactions by our consultants, helping drive a 5-point increase in Net Promoter Score among our junior banker population in Q3.”
Engineer capacity freed by AI coding tools: 25% of capacity in AI-enabled teams (Q2 FY2026, bottomline)
“AI coding assistance now author nearly one fifth of our successful code commits and free up a quarter of our engineers' capacity in those teams.”
BAU software upgrade/patching effort reduction: >90% (Q2 FY2026, bottomline)
“This includes over 90% reduction in efforts spent on business-as-usual activities like software upgrades and patching.”
New product development cycle compression (AI automation): Month-long cycle collapsed to 1 day (Q2 FY2026, bottomline)
“Some teams have radically reduced time to market for new product development by fully automating the delivery life cycle and collapsing a month-long cycle to one day.”
Private company classification capacity (Rubik's AI project): 4x YoY at flat cost (Q2 FY2026, bottomline)
“We have quadrupled classification capacity year over year while keeping costs flat, capturing scale economies in our business.”
Manual data curation reduction from AI tools in data operations: 25%+ average reduction across 4 distinct AI tools (Q2 FY2026, bottomline)
“This quarter, we have deployed four distinct AI tools across different parts of our data operations, generating 25%+ reduction in manual curation on average.”
Vectorization cost reduction: 80% (Q2 FY2026, bottomline)
“we have been able to reduce the cost of vectorizing client data by 80% while delivering faster and more accurate results”
Text-to-formula agent absorption of help-desk volume: Double-digit monthly growth in formula requests; CSR-handled volume declining (Q2 FY2026, bottomline)
“Our help desk experiences double-digit monthly growth in formula support requests. But the volumes handled by our client service representatives have now started to decline as the agent absorbs an increasingly large share of these inquiries each month.”
Full-year in-year productivity savings from AI/automation (vs. 100 bps target): >50 bps already secured at Q2 midpoint (H1 FY2026, bottomline)
“Of our planned 100 basis points in savings, we have already secured more than half and remain on track to deliver the full benefit in H2.”
MCP API call volume growth (intra-quarter, Q2 period): March at 3x February level (Q2 FY2026, topline)
“API call volume is steadily growing as well, with March volumes at three times the February level.”
PAST (realized)
- Q3 FY2026 — Sanoke Viswanathan: 'coding-related token use grew 5x quarter-over-quarter, while committed lines of AI-written code grew almost 10x. Coding agents now author 27% of committed code. With these efficiency gains, we initiated a roughly 10% reduction in our technology workforce.'
- Q3 FY2026 — Sanoke Viswanathan: 'we have reduced operator touch time for data table extraction by more than 50%. Within FactSet Fundamentals, we've consolidated multiple data pipelines into one, allowing us to redeploy significant capacity and reduce the size of this team by 5%.'
- Q3 FY2026 — Sanoke Viswanathan: 'approximately 4,000 bankers used our digital onboarding tools, and the capacity unlocked resulted in a 22% quarter-over-quarter increase in live user interactions by our consultants, helping drive a 5-point increase in Net Promoter Score among our junior banker population in Q3.'
- Q3 FY2026 — Sanoke Viswanathan: 'over 10% of the ASV growth came directly from AI SKUs, and there was a much bigger impact in the broader ASV growth as well.'
- Q3 FY2026 — Sanoke Viswanathan: 'One of the top 10 banks literally doubled their data subscriptions with us because of AI — multiyear contracts. A top hedge fund grew 6x with us because of our MCP delivery.'
- Q3 FY2026 — Sanoke Viswanathan: 'Q3 volumes at 13x the level we experienced in Q2.' (MCP API call volume)
- Q2 FY2026 — Sanoke Viswanathan: 'AI coding assistance now author nearly one fifth of our successful code commits and free up a quarter of our engineers' capacity in those teams. This includes over 90% reduction in efforts spent on business-as-usual activities like software upgrades and patching. Some teams have radically reduced time to market for new product development by fully automating the delivery life cycle and collapsing a month-long cycle to one day.'
- Q2 FY2026 — Sanoke Viswanathan: 'We have quadrupled classification capacity year over year while keeping costs flat... This quarter, we have deployed four distinct AI tools across different parts of our data operations, generating 25%+ reduction in manual curation on average.'
- Q2 FY2026 — Sanoke Viswanathan: 'The text-to-formula agent that we launched in October 2025 has fundamentally changed how we handle client inquiries. Our help desk experiences double-digit monthly growth in formula support requests. But the volumes handled by our client service representatives have now started to decline as the agent absorbs an increasingly large share of these inquiries each month.'
- Q2 FY2026 — Helen Shan: 'we have been able to reduce the cost of vectorizing client data by 80% while delivering faster and more accurate results.'
CURRENT (now)
- Q3 FY2026 — Sanoke Viswanathan: 'Over 90% of our top 50 clients are now using 4 or more FactSet AI solutions. And quarter-over-quarter, overall ASV growth among clients using our AI solutions was 50% higher than for the rest of the book.'
- Q3 FY2026 — Sanoke Viswanathan: 'Our MCP server has over 450 clients actively engaged under contracts and trials.'
- Q3 FY2026 — Sanoke Viswanathan: 'Over 20% of our top 100 clients are using MCP on a paid basis.'
- Q3 FY2026 — Sanoke Viswanathan: 'whenever there is AI consumption through MCP, it is actually leading to an upsizing of our existing products, whether workstations, APIs or other standard data feeds. There is a multiplier effect on the existing business.'
- Q3 FY2026 — Sanoke Viswanathan: 'We are rolling out a new AI-powered sales enablement platform targeted at improving the quality of our sales pitches, increasing deal velocity and improving win rates.'
- Q3 FY2026 — Sanoke Viswanathan: 'active or pipeline trials at over 30 of our top 100 banking clients.' (Capital Markets Intelligence)
- Q2 FY2026 — Sanoke Viswanathan: 'Today, 48 of our top 50 clients are using at least three of our AI solutions, with several more in trials.'
- Q2 FY2026 — Sanoke Viswanathan: 'our MCP solution... has been our fastest-growing solution in the market.'
- Q2 FY2026 — Sanoke Viswanathan: 'As clients move AI into production, they are pulling FactSet Research Systems Inc. deeper into their operations, not replacing us.'
FORWARD (guidance)
- Q3 FY2026 — Sanoke Viswanathan: 'We will roll out similar capabilities to our buy-side and wealth clients with our Institutional Research Intelligence and Adviser Intelligence product suites.'
- Q3 FY2026 — Sanoke Viswanathan: 'The partnership will focus on enhancing FactSet's Workstation with Google's enterprise search, Deep Research API, Grounding and other multimodal capabilities; bringing financial intelligence directly into Gemini Enterprise and expanding our MCP and agent sharing functionalities; and developing a new generation of agents using the Gemini Enterprise Agent platform.'
- Q3 FY2026 — Sanoke Viswanathan: 'We spoke about the AI flywheel effect a couple of quarters ago — it is still very early stages, but we are starting to see that in action.'
- Q3 FY2026 — Joshua Warren: 'As AI and agents reshape how information is sourced, synthesized and acted upon, FactSet is positioned for durable structural growth.'
- Q2 FY2026 — Sanoke Viswanathan: 'We expect this success to be replicated across our AI solutions in all layers of the stack.'
- Q2 FY2026 — Helen Shan: 'We expect these benefits to accelerate through next fiscal year and beyond.'
- Q2 FY2026 — Sanoke Viswanathan: 'I see tremendous scope for growth of the AI-based opportunities... our focus is going to be very much on improving earnings going into 2027 and beyond.'
- Q2 FY2026 — Sanoke Viswanathan (re: workforce AI displacement): 'our connected data, and the embedding that we have in these workflows just becomes exponentially more valuable in a world where agents are becoming the primary call on that data.'
TRACK RECORD — PROMISE vs DELIVERY
74/100 track record too-early 6 calls reviewed
FactSet rarely makes quantified, dated AI commitments—mostly reporting adoption, investment, and workflow metrics instead—and the one explicit FY2026 target (100 bps productivity) looks on track at midyear without being fully judgeable yet. Where numeric AI claims appear, later calls generally show real product rollout, client uptake, and monetization rather than quiet walk-backs.
100 bps of FY2026 productivity improvement from AI/automation initiatives — promised Q2 FY2026
too-early Management disclosed the annual target in Q2 and said first-half initiatives had already captured more than half; Q3 reported further AI-driven gains (27% AI-authored code, >50% data-ops touch-time cuts, expanded agent adoption) but did not confirm full-year delivery before fiscal year-end.
Replicate MCP Server traction across the broader AI solutions stack — promised Q2 FY2026
delivered MCP engaged clients rose from 120 to 450 and API volumes grew 13x quarter-over-quarter by Q3 FY2026, with Portfolio Analytics MCP launched and AI SKUs cited as >10% of quarterly ASV growth.
Launch six distinct AI offerings to automate complex tasks and support agentic workflows — promised Q4 FY2025
delivered FactSet reported all six offerings live in FY2025 and by Q1–Q3 FY2026 cited >45% sequential AI-product usage growth, 48→90%+ top-50 client adoption, and paid MCP use at 20%+ of top-100 clients.
Text-to-formula agent to deflect a large share of routine formula-support traffic — promised Q1 FY2026
partial Agent handled ~35% of daily formula questions at launch; by Q2–Q3 FY2026 human support volumes declined as adoption grew and consultants shifted to higher-value work with rising NPS.
10x faster third-party data ingestion without added headcount — promised Q1 FY2026
partial Q2–Q3 FY2026 cited quadrupling private-company classification capacity YoY at flat cost and >25% average manual-curation reductions from four deployed data-ops AI tools, but no explicit reconfirmation of the 10x ingest benchmark.
Share medium-term AI strategy and business plan at post–FY2026 Investor Day — promised Q2 FY2026
too-early By Q3 FY2026 management rebranded AI under FactSet Intelligence, detailed a three-layer stack, and reaffirmed an upcoming Investor Day, but the full medium-term plan had not yet been presented within this transcript set.
PRICED-IN (REFINED)
LOW (room left)Est. revisions falling · Fwd P/E 14.6 · EV/Sales 4.2x
AI claim maps to Americas Segment, EMEA Segment, Asia Pacific Segment
Estimate momentum is negative: buy ratings fell from 4 to 2 over six months, consensus is heavily Hold (20 of 28), and price targets stepped down (lastMonthAvg 237.33 < lastQuarterAvg 237.71 < lastYearAvg 274.86) even as the stock trades at 248.53. Forward consensus bakes only modest growth (FY25–27 revenue ~6.5% CAGR, EPS 17.08→19.57), not an AI-driven re-acceleration. Valuation is not stretched—14.6x forward P/E and 4.2x EV/Sales are reasonable for a mature financial-data vendor—so falling revisions on a non-rich multiple imply AI efficiency/monetization upside is not yet reflected in estimates or multiples; any AI revenue lift would most plausibly flow through the Americas, EMEA, and Asia Pacific segments.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
4Q2 FY20255Q3 FY20257Q4 FY20257Q1 FY20268Q2 FY20269Q3 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
AI shifted from sparse workflow mentions to a CEO-led FactSet Intelligence platform with adoption metrics, agent products, and measurable internal productivity gains.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
8/10 qualitative impact material near-term · mixed evidence
Where AI matters: ASV growth, MCP monetization, internal productivity
AI is already a measurable commercial and cost engine—>10% of Q3 ASV growth from AI SKUs, MCP driving 90% deal-level contract uplift and core subscription upsizing, plus ~100bps productivity savings from coding agents and data-ops automation—but direct revenue math (~0.6% of base) and much of the MCP flywheel remain early and partly anecdotal.
Caveats: Direct AI-SKU revenue is still a small slice of total revenue despite >10% of ASV growth contribution; Emerging consumption/MCP pricing adds ASV forecasting volatility and may trade stable seat ARPU for lower per-query economics over time; Rising token/compute costs are a new margin line item with uncertain long-run ROI; Platform partnerships (Google/Gemini) expand distribution but create long-term disintermediation risk if clients route through partner UIs instead of FactSet workstations
AI DISRUPTION / CANNIBALIZATION RISK tailwind · 3/10
LLMs can commoditize basic research synthesis and eventually pressure seat-linked workstation pricing, but FactSet's moat is proprietary connected data embedded in workflows; current evidence shows expansion (net-new MCP users, doubled bank subscriptions, 6x hedge-fund growth) rather than cannibalization, and the pivot to enterprise plus consumption/API layers offsets UI-seat risk.
OPTIONS / MARKET STRUCTURE
option liquidity: fair
proxy inputs — dollar-ADV $239M · beta 0.715506 · px $250.00
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.
CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions adding, management language 8/10 committed.
INSIDERS selling 4 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) adding as of 2026-03-31: 107 new / 176 closed positions; 377 increased / 255 reduced; institutional ownership +5.08pp; -75 net 13F holders
MGMT LANGUAGE 8/10 committed Heavy on shipped metrics and workforce actions; tempered by early-stage and positioning language.
commit “Coding agents now author 27% of committed code.”
commit “Q3 volumes at 13x the level we experienced in Q2.”
commit “we initiated a roughly 10% reduction in our technology workforce”
VERBATIM AI QUOTES
“Over 90% of our top 50 clients are now using 4 or more FactSet AI solutions. And quarter-over-quarter, overall ASV growth among clients using our AI solutions was 50% higher than for the rest of the book, early evidence that our AI adoption is helping drive retention and expansion opportunities.”
— Sanoke Viswanathan, Q3 FY2026
“In engineering, coding-related token use grew 5x quarter-over-quarter, while committed lines of AI-written code grew almost 10x. Coding agents now author 27% of committed code. With these efficiency gains, we initiated a roughly 10% reduction in our technology workforce and freed up significant capacity to accelerate strategic product development.”
— Sanoke Viswanathan, Q3 FY2026
“In data operations, where we have fully implemented new tools, we have reduced operator touch time for data table extraction by more than 50%. Within FactSet Fundamentals, we've consolidated multiple data pipelines into one, allowing us to redeploy significant capacity and reduce the size of this team by 5%.”
— Sanoke Viswanathan, Q3 FY2026
“In Q3, approximately 4,000 bankers used our digital onboarding tools, and the capacity unlocked resulted in a 22% quarter-over-quarter increase in live user interactions by our consultants, helping drive a 5-point increase in Net Promoter Score among our junior banker population in Q3.”
— Sanoke Viswanathan, Q3 FY2026
“Our MCP server has over 450 clients actively engaged under contracts and trials. API call volume is experiencing rapid growth with Q3 volumes at 13x the level we experienced in Q2.”
— Sanoke Viswanathan, Q3 FY2026
“Through the partnership we announced with Finster in March, we have launched Capital Markets Intelligence — senior bankers can now send an e-mail to an agent describing what they need and receive insights and artifacts directly built on current FactSet data, comparable company analysis and deal precedents, automatically generated and delivered back. What used to take hours or days now happens in minutes. We are seeing strong early engagement with active or pipeline trials at over 30 of our top 100 banking clients.”
— Sanoke Viswanathan, Q3 FY2026
“As AI reshapes financial institutions, FactSet is becoming mission-critical AI infrastructure.”
— Sanoke Viswanathan, Q3 FY2026
“Just in this quarter, over 10% of the ASV growth came directly from AI SKUs, and there was a much bigger impact in the broader ASV growth as well. One of the top 10 banks literally doubled their data subscriptions with us because of AI — multiyear contracts. A top hedge fund grew 6x with us because of our MCP delivery. Over 20% of our top 100 clients are using MCP on a paid basis.”
— Sanoke Viswanathan, Q3 FY2026
“MCP is a real accelerant — we see whenever there are deals with an MCP component, more often than not, in about 90% of cases, we've seen contract value improvements.”
— Sanoke Viswanathan, Q3 FY2026
“around 20% of the users of our endpoints in both MCP trials and paid implementations are net new users, whether in existing clients or new clients. These are new workflows and new workloads coming on thanks to our ability to deliver data to new AI workloads.”
— Sanoke Viswanathan, Q3 FY2026
“whenever there is AI consumption through MCP, it is actually leading to an upsizing of our existing products, whether workstations, APIs or other standard data feeds. There is a multiplier effect on the existing business.”
— Sanoke Viswanathan, Q3 FY2026
“As AI and agents reshape how information is sourced, synthesized and acted upon, FactSet is positioned for durable structural growth.”
— Joshua Warren, Q3 FY2026
“Tokens are interesting in that they were not a line item we really thought about in 2025 — all of the token spending is net new. But we treat tokens like any other resource. We have a series of operational controls around monitoring them — there's a whole regime around developer training, intelligent model routing, using the right tool for the right job, budgeting.”
— Joshua Warren, Q3 FY2026
“Today, 48 of our top 50 clients are using at least three of our AI solutions, with several more in trials.”
— Sanoke Viswanathan, Q2 FY2026
“AI coding assistance now author nearly one fifth of our successful code commits and free up a quarter of our engineers' capacity in those teams. This includes over 90% reduction in efforts spent on business-as-usual activities like software upgrades and patching. Some teams have radically reduced time to market for new product development by fully automating the delivery life cycle and collapsing a month-long cycle to one day.”
— Sanoke Viswanathan, Q2 FY2026
“We have quadrupled classification capacity year over year while keeping costs flat, capturing scale economies in our business. This quarter, we have deployed four distinct AI tools across different parts of our data operations, generating 25%+ reduction in manual curation on average.”
— Sanoke Viswanathan, Q2 FY2026
“The text-to-formula agent that we launched in October 2025 has fundamentally changed how we handle client inquiries. Our help desk experiences double-digit monthly growth in formula support requests. But the volumes handled by our client service representatives have now started to decline as the agent absorbs an increasingly large share of these inquiries each month.”
— Sanoke Viswanathan, Q2 FY2026
“Our MCP Server that is built on a robust ecosystem of content APIs was launched in December and already has over 120 clients actively engaged. API call volume is steadily growing as well, with March volumes at three times the February level. We expect this success to be replicated across our AI solutions in all layers of the stack.”
— Sanoke Viswanathan, Q2 FY2026
“As clients move AI into production, they are pulling FactSet Research Systems Inc. deeper into their operations, not replacing us.”
— Sanoke Viswanathan, Q2 FY2026
“we have been able to reduce the cost of vectorizing client data by 80% while delivering faster and more accurate results”
— Helen Shan, Q2 FY2026
“AI is playing a dual role, enhancing client value through new capabilities while driving productivity gains.”
— Helen Shan, Q2 FY2026
“Of our planned 100 basis points in savings, we have already secured more than half and remain on track to deliver the full benefit in H2.”
— Helen Shan, Q2 FY2026
ANALYST QUESTIONS ON AI
Q (Q3 FY2026, Faiza Alwy (Deutsche Bank)): I wanted to talk more about your AI monetization strategy. Thanks for the detail — 90% of your top 50 clients use 4-plus AI products and 50% faster ASV growth. Can you give context around the monetization? Is it a direct price for AI usage, access to more data sets? You also talked about clients consolidating their AI workflows with you. Are you agnostic as to whether they're using your MCPs or your specific tools inside FactSet workstation?
A: Sanoke Viswanathan: At the moment, we are maximizing our AI monetization with a lens of maximizing enterprise value — it's all leading to the growth acceleration in ASV, increased retention and expansion in existing clients. Just in this quarter, over 10% of the ASV growth came directly from AI SKUs, and there was a much bigger impact in the broader ASV growth as well. One of the top 10 banks literally doubled their data subscriptions with us because of AI — multiyear contracts. A top hedge fund grew 6x with us because of our MCP delivery. Over 20% of our top 100 clients are using MCP on a paid basis. [Long term:] We have a strong moat in our connected data and in our embedded workflows. We see this as a leapfrog moment for us on a stable subscription base... We are able to flex all of this into commercial agreements that are not just seat linked, but are true enterprise agreements with a stable, large subscription base and a flexible construct on top to capture upside from consumption in the future. The momentum is picking up.
Q (Q3 FY2026, Kelsey Zhu (Autonomous)): A lot of info services companies have talked about AI implementation driving accelerated data demand. Can you talk more about FactSet's strategy to monetize on this trend, both near term and long term? And how should we think about incremental revenue opportunities from MCP, especially around the expansion of new user personas?
A: Sanoke Viswanathan: The short-term monetization we see is an acceleration in our ASV growth. MCP is a real accelerant — we see whenever there are deals with an MCP component, more often than not, in about 90% of cases, we've seen contract value improvements. At the moment, it's playing out in the overall broad ASV acceleration. As we go along, we are going to see more and more discrete AI SKUs — more than 10% of the ASV this quarter came from that, up from virtually 0 last year. On new user personas — around 20% of the users of our endpoints in both MCP trials and paid implementations are net new users, whether in existing clients or new clients. These are new workflows and new workloads coming on thanks to our ability to deliver data to new AI workloads... whenever there is AI consumption through MCP, it is actually leading to an upsizing of our existing products, whether workstations, APIs or other standard data feeds. There is a multiplier effect on the existing business.
Q (Q3 FY2026, Curtis Nagle (Bank of America)): Maybe just talk about the impact of higher token costs on margin in the quarter. You mentioned an expectation to get higher returns on that spend, and the Google partnership might help. Can you unpack that a little more?
A: Joshua Warren: Tokens are interesting in that they were not a line item we really thought about in 2025 — all of the token spending is net new. But we treat tokens like any other resource. We have a series of operational controls around monitoring them — there's a whole regime around developer training, intelligent model routing, using the right tool for the right job, budgeting. Sanoke mentioned the ROI we're seeing on tokens. We are very pleased to be growing our investment in them.
Q (Q3 FY2026, Andrew Nicholas (William Blair)): I'm curious how you think about any differences in the opportunity, pace of adoption and right to win within wealth. The TIFIN.AI partnership is part of that answer, I imagine. Josh, given your background, do you have any additional insights on market positioning?
A: Sanoke Viswanathan: In wealth and the broader consumer finance market, there is an exciting opportunity both to power up the adviser experience, which is going through a transformation, and to directly serve the end customer's experience, especially in wealth management. One of the largest business problems for advisers is how to stay on top of customer portfolios, market events and all the analytics associated with them in order to deliver high-quality experiences and improve their coverage ratios. This is precisely where the TIFIN agents will help — we can marry that with our market data signals, internal data, internal research and deliver high-quality adviser experiences. Joshua Warren: Clients are in different stages of evolution... Across the board, whether it's wealth, sell side or buy side, what we're seeing is people are starting to recognize that as more work shifts from humans to agents, the quality of the output depends on the quality of the data going in. That creates a real opportunity for FactSet to play a significant role.
Q (Q2 FY2026, Kelsey Xu (Autonomous Research)): If you transition all of your workstation ASV into Data Solutions ASV and apply usage-based pricing on top of that, what would that look like? How important is it for FactSet that it continues to own the user interface product, especially for research analysts?
A: Sanoke Viswanathan: We are seeing user persona shifting from just the traditional users who continue on the workstation to also include technology teams, data science teams, and the broad enterprise user across our clients, which happen to be very large financial institutions for the most part. So we are seeing actually a real compounding of this at the moment... we are striking enterprise contracts with them that gives both them and us a lot of flexibility in how to continue to deliver value to them in the future... the core of all of this is our data, which is highly valuable in whatever context our clients consume it. Goran Skoko: the concept of utilizing our content or components of FactSet Research Systems Inc. is not new to us... In terms of enterprise level agreements... and consumption laid on top of that, which we think will more than compensate to any type of attrition on the workstation side going forward.
Q (Q2 FY2026, Scott Wurtzel (Wolfe Research)): Can you guys talk about the pace or degree of AI product adoption in the wealth channel? Is this a potential longer tail opportunity?
A: Sanoke Viswanathan: Wealth is a more heterogeneous environment for us amongst our end markets... We are seeing a gradual pickup in AI adoption in wealth. It is certainly behind what we have seen in the sell side, in the buy side, but we see it picking up, and I expect it to be an important growth driver as we go into the next few quarters. Goran Skoko: Some of the AI solutions that we see adoption involve are around prospecting. So we have some of our largest clients adopting our intelligent prospecting and monitoring solution that is something that drives new business for our clients. We are rolling out some of our AI solutions with two of our largest clients currently.
Q (Q2 FY2026, Craig Huber (Huber Research)): There is obviously a lot of concern as AI evolves here over time, that the white collar workforce takes pressure... if you have a 10% or 15% pullback in the number of white-collar workforce at the buy side, sell side, how vulnerable are you?
A: Sanoke Viswanathan: To imagine a scenario where you have that kind of headcount reduction in our end markets, we have to then appreciate that the agentic workflows have become such high grade and high quality that they are able to really displace the humans who do those jobs today... those agents are going to need very, very high-quality inputs in order to be able to execute the job... our data, our connected data, and the embedding that we have in these workflows just becomes exponentially more valuable in a world where agents are becoming the primary call on that data. So we believe that we are very well positioned to capture the upside in that kind of scenario. And it will all come down to the optimization of pricing between the seats that those humans may have adopted versus the data consumption that the agents will have in the future.
Q (Q2 FY2026, Andrew Nicholas (William Blair)): As you think about where FactSet sits in the ecosystem relative to newer competitors or even the model providers, maybe address what you think are some of the disadvantages you have from being a legacy provider? And to what extent are those disadvantages addressable?
A: Sanoke Viswanathan: We have a strong partnership with Anthropic. We are one of the prominent financial services connect on the cloud marketplace. And I must say it is our fastest-growing marketing channel... we make gains when clients connect through Claude into our datasets, consume more and more of our data, and our contracting is directly with our clients. And Anthropic does well when that happens because those agents that the clients may be deploying use more and more tokens. We have always been an open architecture company. And it is really coming into its own in this AI environment. Helen Shan: For broad solutions, many of the AI-native firms can be very good in their point solution. But for clients who want something that is integrated, and they do not want to have multiple solutions, we are really best positioned to be able to deliver that.