← back to rankingADP · Automatic Data Processing, Inc.
Staffing & Employment Services · mkt cap $92.4B · calls: Q3 FY2026 vs Q2 FY2026
53.0 conviction · conf-adj 53
conf 3/10 partial
enthusiasm:27.0 · trend:8 · quantifies:0 · impact:0 · under_radar:5 · credibility:5 · business_impact:8 · disruption:-6 · commitment:6 · confirmation:0
Enthusiasm latest 9 / prev 7 (rising)
ADP's AI thesis shifted from product capability in Q2 to a company-wide business and margin story in Q3, with management framing AI as making HCM more complex and therefore more valuable for ADP. Credibility improved because Q3 included concrete productivity metrics across payroll agents, HR search, Lyric workflows, service operations, and internal compliance work. The quantified impacts are mainly bottomline/productivity so far, with topline claims still indirect through retention, bookings, pricing, and Lyric pipeline commentary.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $20.6B · net income $4.1B · net margin 19.8% · diluted EPS 9.98
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: % · vs analysts: inline · priced in: medium (model's call-read: high; verdict above is the hard-data one used for ranking) · confidence: 3/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
ADP Assist payroll agents save ~30 min per payroll
productivity · soft | 30 min/payroll | Time saving accrues largely to the CLIENT, not ADP's P&L. No payroll-run count, agent labor cost, or affected expense base disclosed -> 30min x unknown payrolls x unknown $/hr is unsizable; after-tax saving = unknown * 79% / $4.0797B NI = null. Productivity item, topline ~0. | 0 | |
Smart Actions search cut clicks/time ~80% for common HR actions
productivity · soft | ~80% | 80% fewer clicks/time on common HR actions — a UX productivity metric with no affected ADP cost or monetization base disclosed -> saving_$ uncomputable; null. Topline ~0. | 0 | |
Lyric cut recruiting steps from 23 to 8
engagement · soft | 23 -> 8 steps (-65.2%) | (23-8)/23 = 65.2% step reduction, but single-customer workflow anecdote with no ADP revenue line, pricing, seats, or adoption base -> revenue uplift = unknown / $20.5609B = null. Illustrative, not aggregatable. | | |
Lyric enabled 71% leaner payroll operations model
engagement · soft | 71% leaner | Single-client benefit (org replacing ~a dozen systems), client-side leanness not ADP cost/revenue; no incremental Lyric revenue, retention, or affected base disclosed -> null. Mild unsized positive for retention/upsell. | | |
GenAI service platform on 20% of service pop., >40% by end FY26
productivity · soft | 20% -> >40% | Deployment penetration of an internal service platform; implies future service-labor productivity but no labor $ pool or per-agent savings disclosed -> saving_$ uncomputable; null. Topline ~0. | 0 | |
8% YoY reduction in client contacts for 900k+ small-business clients
cost · soft | -8% contacts YoY | Hardest operational figure, but anchored to contact VOLUME not a disclosed service-cost line. Illustratively, if SB support cost were ~$1B, 8% = $80M pretax -> $63M after-tax -> ~1.5% EPS — but that base is NOT disclosed, so kept null. Topline ~0. | 0 | |
India year-end: AI cut core volumes 35% and labor 35%
cost · soft | -35% volume / -35% labor | Real productivity %, scoped to ONE process (India year-end compliance). No headcount or labor $ for that process disclosed -> cannot convert to after-tax savings; null. Topline ~0. | 0 | |
Continued acceleration of margin expansion from AI in FY27
other · soft | qualitative ('acceleration') | Explicitly forward/qualitative ('still early in our planning'). No bps or $ target -> incremental NI uncomputable against $4.0797B NI; null. | | |
Assumptions: Current revenue = $20.5609B; current net income = $4.0797B; net margin = 19.84%; tax rate 21% (after-tax cost savings use 79%). Incremental net margin defaults to company net margin 19.84%, but moot: no claim provides a dollar revenue target or dollar cost-saving base. Phasing: claims are in-period operational metrics, not multi-year revenue ramps, so no phasing split. Segment mapping: SB contact-reduction -> service/support opex (base undisclosed); India item -> compliance-ops labor (base undisclosed); Lyric/Assist -> client-side value, unsized. EPS sensitivity anchor: $100M after-tax = +2.45% of $4,079.7M NI. All claims adopter-side; none supplier-side AI infrastructure revenue.
Top line: Effectively zero direct topline on a $20.56B base. Every AI claim is a productivity/cost or client-workflow improvement, not incremental ADP revenue or bookings. Strongest product metrics (Lyric steps -65.2%, 71% leaner payroll ops) are customer examples without price, adoption, or revenue-base disclosure; management quantified no AI revenue dollars, so est_rev_uplift_pct = 0%.
Bottom line: All impact is bottom-line/margin, but none is dollar-quantified. ADP gives operating metrics — 30 min/payroll, ~80% less Smart Actions time, 8% fewer SB contacts, 35% less India year-end labor — without the payroll counts, contact costs, or labor pools needed to translate into after-tax savings, so each eps_uplift is null rather than invented. Direction is clearly positive, but unsized: even a $100M after-tax block is only +2.45% EPS vs $4.08B NI, so realistic aggregate AI EPS lift is plausibly low-single-digit % and embedded in guided margin expansion.
[impact n/m (all claims soft/unanchored)] Consensus already bakes in ADP's steady margin expansion: EPS 9.14 (FY24) -> 9.98 (FY25) = +9.4% on revenue +6.8%, i.e. net margin widening ~40-50bps/yr — exactly the 'continued margin expansion from AI productivity' management describes. FY25 consensus EPS (9.97936) essentially equals actual (9.98) and consensus NI ($4.0786B) ~ actual ($4.0797B), so analysts already track the productivity story. With no disclosed AI revenue and no sizable disclosed savings dollars, there is no math pointing clearly above consensus — adopter-side aggregate rev +0.0%, EPS indeterminate/low-single-digit and already in the margin glide path.
MODEL CONSENSUS (impact)
partial
Both agree: all adopter-side, zero hard revenue, EPS uplift uncomputable/null, priced_in high. Reconciled verdict to inline and confidence to 3.
Conflicts reconciled
- vs_analyst_expectations: X=unclear vs Y=inline -> used inline because Y grounds it in the consensus margin-expansion math (EPS +9.4%, ~40-50bps/yr) that AI productivity feeds
- confidence: X=4 vs Y=3 -> used 3 (more conservative) since estimate rests entirely on soft, undollar-anchored claims
- supplier_rev_uplift_pct: X=0 vs Y=null -> used 0 because no supplier-side AI revenue exists
- rev_uplift_pct(productivity/cost): X=0 vs Y=null -> used 0 per method (cost/productivity topline ~0); kept null for Lyric client anecdotes
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | 0.0 | 0 |
| EPS uplift % | – | – |
| Priced in | high | high |
| vs analysts | inline | unclear |
| Confidence | 3 | 4 |
| Top line | Effectively zero direct topline. Every AI claim is a productivity/cost or client-workflow improvement, not incremental ADP revenue or bookings. The strategic argument ('AI makes HCM more important', 'widen the gap vs competitors') supports retention and Lyric upsell, but management quantified no AI revenue dollars, so est_rev_uplift_pct = 0.0% on a $20.56B base. | No hard next-FY revenue uplift is calculable. The strongest product/engagement metrics are Lyric steps down 65.2% and payroll operations 71% leaner, but both are customer examples without price, adoption, or revenue-base disclosure. |
| Bottom line | All impact is bottom-line/margin, but none is dollar-quantified. The cleanest figures — 8% fewer SB client contacts and 35% less India year-end labor — lack the cost bases needed to translate into after-tax savings, so each eps_uplift is null rather than invented. Direction is clearly positive (service-labor productivity feeding the stated margin-expansion acceleration), but unsized: even a sizable $100M after-tax block is only +2.45% EPS against $4.08B NI, so realistic aggregate AI EPS lift is plausibly low-single-digit % and embedded in guided margin expansion. | The AI claims are clearly bottom-line oriented, but ADP gives operating metrics rather than dollars: 30 minutes per payroll, 80% less Smart Actions time/clicks, 8% fewer small-business contacts, and 35% labor reduction in one compliance process. Without payroll counts, contact costs, or affected labor pools, EPS uplift cannot be responsibly computed. |
| Reasoning | Consensus trajectory already bakes in ADP's steady margin expansion: EPS 9.14 (FY24) -> 9.98 (FY25) = +9.4%, on revenue 19.17B -> 20.48B = +6.8%, i.e. net margin widening ~40-50bps/yr — exactly the 'continued margin expansion from AI productivity' management describes. The FY25 consensus EPS (9.97936) essentially equals actual (9.98), so analysts are tracking management's productivity story with no quantified AI upside left to capture. With no disclosed AI revenue and no sizable disclosed savings dollars, there is no math pointing clearly ABOVE consensus — the gap is ~0. Adopter-side aggregate: rev +0.0%, EPS indeterminate/low-single-digit and already in the margin glide path. | Consensus already embeds FY2025 actual-like EPS of $9.97936 versus actual diluted EPS of $9.98 and net income of $4.0786B versus actual $4.0797B. Since the quantified AI claims produce $0 hard revenue uplift and no computable EPS uplift, there is no measurable upside above consensus from the disclosed figures alone. |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
ADP Assist payroll agent time savings: 30 minutes per payroll (since January launch, Q3 FY2026, bottomline)
“Our ADP assist payroll agents have saved an average of 30 minutes per payroll.”
Smart Actions search click/time reduction: around 80% (Q3 FY2026, bottomline)
“Our Smart Actions search has reduced clicks and time spent by around 80% for common HR actions.”
Lyric recruiting process step reduction: 23 down to just 8 (Q3 FY2026 customer example, bottomline)
“One senior HR leader at a supply chain firm shared that the AI tools within Lyric have significantly reduced the number of steps in the recruiting process from 23 down to just 8 by providing advanced candidate in size.”
Lyric payroll operations efficiency: 71% leaner payroll operations model (Q3 FY2026 customer example, bottomline)
“Another client, a global holding company used [ Lyric ] to replace more than a dozen disparate systems, which enabled a 71% leaner payroll operations model and that's just the beginning.”
GenAI service platform deployment: 20% of total service population; over 40% expected (as of March; by end of fiscal 2026, bottomline)
“As of March, 20% of the total service population was on the own platform, and we expect to reach over 40% by the end of fiscal '26.”
AI-powered small business contact reduction: 8% year-over-year reduction in client contacts (fiscal Q3 FY2026, bottomline)
“As an example, our continued investment in our RAM platform, along with the AI-powered tools that were deployed to support our more than 900,000 small business clients have enabled an 8% year-over-year reduction in client contacts in fiscal Q3, our busiest quarter of the year.”
AI compliance-process volume and labor reduction: 35% reduction in core volumes; 35% reduction in labor (India year-end at March 31, FY2026, bottomline)
“We actually deployed AI this year for the first time, reduced the core volumes by 35% in the year-end process, also reduced the labor by 35%.”
AI-related margin outlook: continuing this acceleration when it comes to margin expansion (fiscal 2027 planning, bottomline)
“While it is still early in our planning process for fiscal '27, we remain very focused on continuing this acceleration when it comes to margin expansion as we realize further productivity benefits from our AI transformation.”
PAST (realized)
- Our ADP assist payroll agents have saved an average of 30 minutes per payroll.
- Our Smart Actions search has reduced clicks and time spent by around 80% for common HR actions.
- One senior HR leader at a supply chain firm shared that the AI tools within Lyric have significantly reduced the number of steps in the recruiting process from 23 down to just 8 by providing advanced candidate in size.
- Another client, a global holding company used [ Lyric ] to replace more than a dozen disparate systems, which enabled a 71% leaner payroll operations model and that's just the beginning.
- We actually deployed AI this year for the first time, reduced the core volumes by 35% in the year-end process, also reduced the labor by 35%.
CURRENT (now)
- AI remains central to our technology strategy, and we are moving full speed ahead to leverage it in attracting, serving, and retaining our clients.
- The margin expansion we achieved reflects disciplined investment. We are funding our AI transformation across our products, internal tools and service delivery while continuing to deliver on our financial commitments.
- The investments that we are making in AI, in service tools and in product innovation are yielding meaningful productivity improvements in our business, allowing us to reduce our cost to serve, while at the same time, enhancing our clients' experience.
- As of March, 20% of the total service population was on the own platform, and we expect to reach over 40% by the end of fiscal '26.
FORWARD (guidance)
- AI makes HCM more important, and we believe it unlocks tremendous value and opportunity for our industry that plays out in 2 ways.
- This advantage will continue to compound for our data and AI capabilities over time and will further widen the gap between ADP and our competitors.
- While it is still early in our planning process for fiscal '27, we remain very focused on continuing this acceleration when it comes to margin expansion as we realize further productivity benefits from our AI transformation.
- I would still say we are in the relatively early innings in terms of taking dividends, if you like, from these investments and really seeing sort of the lift we expect to get from this over the coming years.
TRACK RECORD — PROMISE vs DELIVERY
55/100 track record too-early 6 calls reviewed
ADP rarely issues quantified, dated forward AI targets, instead reporting achieved AI productivity metrics (e.g. ~30 min/payroll saved, ~80% fewer clicks, 8% fewer client contacts) after the fact. The one clear quantified forward AI commitment — 40% of the service population on the GenAI 'zone' platform by end of FY26 — was made on the latest call and is not yet judgeable, so the track record is genuine but too-early to score.
Scale 'the zone' GenAI service platform from 20% of total service population (as of March) to over 40% by the end of fiscal 2026 — promised Q3 FY2026
too-early Target set on the most recent call; fiscal 2026 has not closed, so on-track but not yet verifiable in this set.
Expand ADP Assist / role-based agents from simple to autonomous agents managing workflows end-to-end (FY26 roadmap) — promised Q4 FY2025
too-early Persona-based payroll, HR, tax and analytics agents were launched through FY26, but the 'autonomous, start-to-finish' milestone was aspirational with no hard number/date, so not strictly scoreable.
Drive AI-led productivity into reported results (payroll agents saving time, reducing client contacts) — promised Q1 FY2026
too-early By Q3 FY2026 ADP reported ~30 min saved per payroll, ~80% fewer clicks on common HR actions, and an 8% YoY drop in client contacts — but these were reported outcomes rather than prior quantified promises.
PRICED-IN (REFINED)
MEDIUMEst. revisions flat · Fwd P/E 28.3 · EV/Sales 4.3x
AI claim maps to HCM, HRO, Professional Employee Organization Services Segment
Analyst ratings show only slight improvement since early 2026, while price targets are mixed: last-month average is above the last-quarter average but still below the last-year average. Forward revenue and EPS growth are solid but not explosive, suggesting consensus is not rapidly repricing an AI acceleration thesis. Valuation is rich for a mature employment-services company at 28.3x forward EPS and 4.3x EV/Sales, so even with flat revisions, some AI upside appears embedded. That combination points to medium priced-in risk rather than low or high.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
3Q2 FY20254Q3 FY20258Q4 FY20259Q1 FY20269Q2 FY202610Q3 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
AI moved from passing recognition to a central strategy with agents, usage metrics, client outcomes, internal productivity, and data-driven differentiation.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
8/10 qualitative impact material medium-term · mixed evidence
Where AI matters: HCM product automation and service cost-to-serve
ADP has real deployed AI across payroll agents, HR search, Lyric workflows, service operations and compliance, with concrete time/contact/labor reductions that touch core HCM delivery. The case is still mostly margin, retention and product-quality driven rather than proven incremental revenue, so it is material but not yet transformational.
Caveats: No disclosed AI revenue, bookings, pricing uplift or EPS dollars; Customer workflow savings may accrue more to clients than ADP; AI could commoditize basic HR administration and pressure service-heavy offerings; Execution risk in scaling internal service tools from 20% to over 40% of service population
AI DISRUPTION / CANNIBALIZATION RISK two-sided · 4/10
AI can reduce the administrative HR/payroll labor and workflow complexity that supports parts of ADP's HRO/PEO/service value proposition, creating some pricing and labor-arbitrage pressure. The core payroll/compliance system-of-record model is more durable because accuracy, regulatory logic, client data, trust and integration matter more than generic automation.
OPTIONS / MARKET STRUCTURE
option liquidity: good
proxy inputs — dollar-ADV $778M · beta 0.841 · px $231.18
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 trimming, management language 8/10 committed.
INSIDERS selling 15 open-market sell(s) vs 1 buy(s) — net distribution
INSTITUTIONS (13F) trimming as of 2026-03-31: 192 new / 330 closed positions; 932 increased / 1013 reduced; institutional ownership +1.44pp; -137 net 13F holders
MGMT LANGUAGE 8/10 committed Management uses firm ownership, launches, quantified productivity gains and rollout targets, with some longer-term strategic positioning language.
commit “In January, we launched ADP Assist agents that apply advanced intelligence to real workforce challenges for us payroll and HR.”
commit “Our ADP assist payroll agents have saved an average of 30 minutes per payroll.”
commit “As of March, 20% of the total service population was on the own platform, and we expect to reach over 40% by the end of fiscal '26.”
VERBATIM AI QUOTES
“For us, success means leading the way in a trusted service-driven and AI-powered HCM and setting the industry standard for accuracy, compliance and partnership around the world.”
— Maria Black, Q3 FY2026
“AI makes HCM more important, and we believe it unlocks tremendous value and opportunity for our industry that plays out in 2 ways.”
— Maria Black, Q3 FY2026
“What differentiates ADP's approach is that our AI is built in the very core of how we orchestrate, govern and execute HR and pay processes, grounded in regulatory logic, operational data and decades of expertise.”
— Maria Black, Q3 FY2026
“Our ADP assist payroll agents have saved an average of 30 minutes per payroll.”
— Maria Black, Q3 FY2026
“Our Smart Actions search has reduced clicks and time spent by around 80% for common HR actions.”
— Maria Black, Q3 FY2026
“AI is only as good as the data it's built on, and ADP has the industry's strongest workhorse data foundation built over nearly 77 years.”
— Maria Black, Q3 FY2026
“The margin expansion we achieved reflects disciplined investment. We are funding our AI transformation across our products, internal tools and service delivery while continuing to deliver on our financial commitments.”
— Peter Hadley, Q3 FY2026
“The investments that we are making in AI, in service tools and in product innovation are yielding meaningful productivity improvements in our business, allowing us to reduce our cost to serve, while at the same time, enhancing our clients' experience.”
— Peter Hadley, Q3 FY2026
“AI remains central to our technology strategy, and we are moving full speed ahead to leverage it in attracting, serving, and retaining our clients.”
— Maria Black, Q2 FY2026
“We continue to scale the usage and capabilities of our client-facing AI, including the launch of new ADP Assist payroll, HR, analytics, and tax agents that apply advanced intelligence to real workforce challenges.”
— Maria Black, Q2 FY2026
“Unlike generic AI solutions, ADP's approach combines proprietary workforce insights with advanced automation to solve real workforce challenges while maintaining the security, governance, and compliance standards companies trust.”
— Maria Black, Q2 FY2026
“What it has done to your point is it has enabled our sellers to be both more efficient and, I think, also more effective.”
— Peter Hadley, Q2 FY2026
ANALYST QUESTIONS ON AI
Q (Q3 FY2026, Mark Marcon): One, you mentioned how AI is taking you more efficient. And I couldn't help but notice that the R&D or the program costs were relatively flattish despite the nice increase in terms of revenue. And I'm wondering if you can talk a little bit about some of the efficiencies that you're gaining across the board from AI and particularly in terms of new product development and the tools that you might be employing there, both in terms of reduced expenses, but also speeding up the development process.
A: We've certainly pivoted more of our spending in R&D towards AI initiatives, be it on the product side to benefit our clients as well as on the efficiency side. So there's a range of different things.
Q (Q3 FY2026, Dan Dolev): I wanted to ask about, I know there was a question about AI and R&D, but more about -- I think your competitor mentioned that there was some difficulty selling software modules. I just want to see from your perspective how this looks?
A: Again, the way we see it is we see the future of work as one that is AI infused and AI really powering workforce, but that doesn't take away the need to actually manage this year orchestration of paying people and keeping them compliant.
Q (Q3 FY2026, Scott Wurtzel): The commentary, I think, on ADP assist. It was great to hear, but I think more broadly now that you've had some of these products and AI features in the market for some time now. What is sort of the overall feedback that you've been hearing from clients regarding these products? And is there anything potentially more on the AI front from a product perspective that clients are looking for?
A: I would tell you, Scott, that the feedback is incredible. I cited a couple of examples. I probably could have gone on for 30 more off the top of my head.
Q (Q3 FY2026, Jason Kupferberg): So obviously, still a lot of debate in the market about how AI could impact seat-based revenue models. I think ADP has said in the past, a 1% change in pace per control impacts ES revenue by about 25 bps.
A: In saying that, we feel like there's -- it's a value-based price approach that we've always taken. So again, the value we confer is not necessarily linear with the number of employees the client has.
Q (Q3 FY2026, Daniel Jester): Maybe a 2-parter on ADP Assist. So your first one is, I don't know if you shared this in the prepared remarks. Have you made any comments about sort of uptake repeat usage, engagement levels with the customers that have access to it.
A: Once they get started, they get, call it, hooked on continuing to process improve and engage with these tools.
Q (Q3 FY2026, Daniel Jester): As you roll more of these out, how do you view out sharing some of the value from the time savings that these agents are providing?
A: So we're not really looking at this as a discrete usage type of fee at a piece by piece level. We really look at it as core in the fabric of how we operate and how we deploy our products to our clients.
Q (Q2 FY2026, Scott Wurtzel): Hate to ask a question on AI impacts on hiring, but just in the context of even over the last, you know, 24, 48 hours, seeing some incremental announcements from enterprises around layoffs and citing AI. I am just wondering if you have any updated views on that topic.
A: We are not actually seeing in those industry verticals. So, you know, things like financial services, things like professional services, tech, and so on. You know, we are actually seeing reasonably healthy growth.
Q (Q2 FY2026, Kartik Mehta): And, Peter, just a question on AI. I know you talked a little bit about AI and maybe the impact of employment for your clients. I am wondering for Automatic Data Processing, Inc., I think you have implemented AI, I think you have had success on the sales side.
A: What it has done to your point is it has enabled our sellers to be both more efficient and, I think, also more effective.