← All companies

ZSZscaler, Inc.

Industry Cyber software Updated 25 Sep 2026

The idea

Organic net new ARR accelerated 7% → 10% → 17% through FY26 while the FY27 guide embeds ~4%; if Security for AI and Agentic SOC land, 17% is a floor and the stock re-rates from ~7.7x EV/FY27 revenue.

Events to watch

  • Investor Day 2026-10-06 (FY28-30 targets, non-seat ACV mix)
  • Q1 FY27 results ~2026-11-24
  • net new ARR and NRR (kill below 112%)
How we got the target

The dossier's illustrative FY29 scenario EVs (bear $25bn / base $45bn / bull $68bn) plus $1.77bn net cash, divided by ~180m diluted shares. The dossier itself calls these 'illustrative, not a model'. Time frame: FY2029 (Jul 2029) — beyond two years.

Target: worked out from expected earnings. Prices: market data get_latest_stock_price, close of 2026-09-22 (fetched 2026-09-23)

What would prove it wrong

6 warning signs
  1. Organic net new ARR growth below 10% in any two consecutive quarters

  2. NRR below 112%, or seats per customer falling without non-seat ACV offsetting it

  3. Q1 FY27 capex guide above low teens % of revenue, or FCF margin guide below 23%

  4. Another senior sales departure before 1H FY27 ends

  5. Investor Day FY28–30 targets implying growth ≤15% with no non-seat acceleration

  6. Macro override: 10-year held above 5% with a second Fed hike signalled

How far along

10 of 10 steps done

Next step

Investor Day 2026-10-06 — read FY28–30 targets and non-seat ACV mix against the 17%

Needs your decision

product/expert conversation, thesis conviction

Failed checks

none. S0 liability test scores 0; S7 earnings-power tenet 4/10 (not cheap on the

Who runs it

The Front Office

Step by step

  1. S0How it does as AI gets cheapDone

    3/4

    lock-in 1, unscrapeable telemetry 1, liability 0, bottleneck 1; falsifier: hyperscaler/Microsoft agent-inventory bundling; gate met (owner checks dropped 2026-09-21)

  2. S1Where the idea came fromDone

    owner request, no thesis; no prior owner notes in Gmail or Drive.

  3. S2Quick deal-breaker checkDone

    25% growth, 80% GM, net cash, FCF $779m; flags: SBC $822m ≈ FCF, share count +3.5%/yr, GAAP loss since IPO · capital allocation (2026-09-25, dossier-2026-09-25.md): neutral. FY2022-26 operating cash flow $3,666.3m; capex $752.7m and capitalised software $257.8m leave free cash flow $2,655.7m, of which $1,334.5m went on acquisitions and nothing at all to shareholders — no buyback since FY2019, no dividend ever.

    Full notes

    Diluted shares 140.9m to 160.2m (+13.7%, 3.3%/yr), funded entirely by issuance plus 3.8m shares handed over to settle the 2025 converts' premium. ROIC is not meaningful and this line will not invent one: GAAP operating income was negative all five years (-327.4 to -133.3 $m) and the cumulative GAAP net loss is $755.0m, so the incremental return is a narrowing loss. What can be measured: revenue per diluted share $7.74 to $20.92 (x2.70) and FCF per diluted share $1.64 to $4.86 (x2.96) — the dilution did buy growth. Goodwill $78.5m to $1,218.1m, no impairment in five years. Largest deal: Red Canary, $675m, exited FY26 at $141m ARR, and management expects no net new ARR contribution from it in FY2027 (Q4 call [33]) with "elevated churn" in FY26 [35]. Bonus metrics (DEF 14A 2025-11-21): revenue 30%, calculated billings 30%, non-GAAP income from operations 40%. PSUs on multi-year ARR. The §25 EPS-pay-plus-buybacks flag does NOT fire (no EPS metric, no buyback);

    ⚠️ the finding is the mirror image — not one pay metric is per-share, cash-flow or return-based, so pay is indifferent to the 3.3%/yr dilution and to the free-cash-flow margin, the single guide this team has missed.

  4. S3Reading the company's reportsDone

    FY26 10-K read end to end and diffed vs FY25.

    Full notes

    NEW risk clause: customers' "use of AI to automate tasks traditionally performed by human workers"; Raj Judge resigned 2026-04-15 (8-K); no S-3/ATM ever EQ: 3 flags (2026-09-25, dossier-2026-09-25.md) — (1) SBC $821.9m = 24.5% of revenue and 105.5% of free cash flow, so FCF after SBC is negative: -$42.8m in FY2026 and -$209.7m across FY2022-26. (2) the non-GAAP bridge widened +27.1% against revenue +25.4%, and the extra width is bought — acquired-intangible amortisation $16.8m to $43.5m (+158%), acquisition-related expenses +280%, and $57.4m of acquisition-related SBC inside the SBC line — while non-GAAP income from operations is the 40%-weighted bonus metric. (3) "restructuring and other charges" in three of the last four years ($7.6m FY23, $4.9m FY25, $10.3m FY26) with a fourth, five times larger, announced two months into FY27 (~3% of employees, $30-33m, Q4 call [27]) — a charge taken in four years of five is an operating cost. Clean: accruals -20.1/-18.2/-16.7% of average assets (conservative). DSO 133.8/131.5/124.0/135.5/125.1 days FY22-26, no two-period run; capitalised software fell to 2.18% of revenue in FY26 after two rising years; no goodwill impairment, no restatement, no 10-K/A, no NT 10-K.

    ⚠️ watch OCF / non-GAAP net income, 1.88x to 1.82x to 1.60x, narrowing three years running.

  5. S4Rivals and suppliersDone

    PANW "taking share from SASE-only", Netskope ARR +34%/NRR 118%, Cloudflare copying the channel model. Microsoft GSA adoption unmeasured.

  6. S5Who owns the stockDone

    BlackRock 5.1% (13G). Q2-26 13F: Invesco/Fidelity/Tiger/T. Rowe out, Jericho/Millennium/Citadel in; insiders sell only (RSU withholding); no shelf overhang.

  7. S6What workers and customers sayDone

    (weak)

    Glassdoor 3.5 (AE 2.9), June outage, "MITM you pay for" customer quote, LinkUp jobs +28% y/y; expert call open OPEN (user)

  8. S7The case for it, and pricesDone

    draft / open OPEN (user)

    Whale Rock memo 22/40 = Watchlist (S-curve 6, moat 6, earnings 4, leader 6); owner conviction not recorded · valuation (bear/base/bull; moved from S8 2026-09-23): 7.7x EV/FY27 rev, 33x EV/FCF, ~29x FY29 non-GAAP EPS; Druckenmiller: three death knells rising, not a fat pitch · valuation, base case re-read 2026-09-25 against S8: the verdict is conservative, so no guidance haircut — but the two halves of the FY2027 guide must not be modelled the same way.

    Full notes

    (a) Revenue and ARR: treat the guide as a floor. Fifteen first guides, fourteen beaten; and the proxy shows the public guide is set 2-6% BELOW the board's own operating plan, so a midpoint model is modelling the sandbag, not the plan. (b) Free cash flow: do NOT take the guide. FY2027 is guided at 23.0-23.5% FCF margin on "low teens" capex. FY2026 was guided at 26-26.5%, raised to 26.5-27%, then cut to 22.8-23.3% and landed at 23% — the only miss in the table, on a cost line management says is set by an AI-driven parts shortage (Chaudhry, Q3 call [51]). Model the FCF margin below the guide, or model the capex directly.

  9. S8Do the bosses keep their word?Done

    15 rows (2026-09-25, dossier-2026-09-25.md): 14 beat, 1 missed

    conservative. 9 rows belong to the current team (Chaudhry CEO throughout. Kevin Rubin CFO since May 2025, first guide 2025-09-02): 8 beat, 1 missed.

    Full notes

    The other 6 are Remo Canessa's, who announced his retirement on the call that opens the window (2024-12-02 [3]) and left 2025-12-31 — all 6 beat, all marked departed. Every quarterly revenue guide in eight quarters was beaten, and FY2025 and FY2026 revenue, EPS, billings and ARR were all beaten against the FIRST guide, not just the last.

    ⚠️ THE BEATS ARE A GUIDANCE SPREAD, NOT UPSIDE — and the proxy proves it. DEF 14A 2025-11-21 prints the board-approved FY2025 plan targets: revenue $2,678.7m, calculated billings $3,273.2m, non-GAAP income from operations $573.2m. The public first guide (8-K 2024-09-03) was set 2.2%, 4.2% and 5.8% BELOW them. Actual: revenue $2,673.1m = 99.8% of plan, billings $3,246.2m = 99.2% of plan, only operating income above at 101.2%. The company's own attainment table funds revenue at 99.5%/98.5% and billings at 95.5%/96.5% of target. Zscaler beats what it tells the market and lands on what it tells its board. Read the guide as a floor; do not read the beats as operating upside.

    ⚠️ THE ONE MISS IS THE COST LINE: FY2026 free-cash-flow margin guided 26-26.5% (2025-09-02 [3]), RAISED to 26.5-27% in February, CUT to 22.8-23.3% in May, landed at 23% — capex went mid-single-digit to high-single-digit to low-teens % of revenue, $218.5m of capex and internal-use software in Q4 2026 alone against $78.7m a year earlier. Cause named on the call and external: "It's created by all these AI data centers that's causing such a shortage of so many parts" (Chaudhry, 2026-05-26 [51]).

    ⚠️

    ⚠️ THE CRO LEFT THREE WEEKS AFTER THE GUIDE — see the failed-gates line.

  10. S9What to watch nextDone

    Investor Day 2026-10-06. Q1 FY27 late Nov; 13F Q3 mid-Nov. Fed Oct meeting / 10y at 5%.

    Full notes

    LinkUp postings monthly · trigger added 2026-09-25 from the dodge log: Q1 FY27 net new ARR against the guided 15% of full-year net new ARR in Q1 (8-K 2026-09-03). That is the first print after the CRO transition and the first number management cannot answer with "prudent"; a Q1 share below 15% says the sales disruption is real and larger than the guide allows for.

Sources and notes

Sources 23 documents
Short nameDocumentWhere it came fromRead on
10K-FY26Zscaler 10-K for FY ended 2026-07-31, filed 2026-09-03Edgar via Bigdata.com doc B794EA8C16EAB2FA398E06ABCAC83AC72026-09-19
DEF14A-25Zscaler DEF 14A filed 2025-11-21Edgar via Bigdata.com doc FFA3E80D4FE57D8D2FF3DB77EFE40C7F2026-09-19
AGM-262026 Annual Meeting transcript 2026-01-12Quartr via Bigdata.com doc 17A1E27686AB9790169FA5560033E2222026-09-19
Q4-CALLQ4 FY26 earnings call 2026-09-03Alpha Vantage EARNINGS_CALL_TRANSCRIPT ZS 2026Q4; Quartr via Bigdata.com doc DF6C3F4792229D084D3982B7201945832026-09-19
Q3-CALLQ3 FY26 earnings call 2026-05-26Quartr via Bigdata.com doc AB1891B363082AFA66518238F5CD2FD12026-09-19
13F-Q2-26Fund holdings, reporting quarter 2026Q2WSRA via Bigdata.com company tearsheet, entity D44D332026-09-19
AV-INSIDERInsider transactions from 2025-09-01Alpha Vantage INSIDER_TRANSACTIONS ZS2026-09-19
PR-SOCPress release 2026-09-09, Zscaler launches Agentic SOCPubT via Bigdata.com doc 16B26F933ED8ED483AC543AF5F20C5D92026-09-19
CONF-0603Investor conference transcript 2026-06-03Quartr via Bigdata.com doc 8FA49E5E21F5829FD66AF9A8F9F198562026-09-19
CONF-0609Customer/investor event transcript 2026-06-09Quartr via Bigdata.com doc 7D882E426C03900C3AAAC6999C3770BA2026-09-19
NEWSNamed news item; the edge locator names outlet, date and doc idBigdata.com premium news2026-09-19
EDGAR-IDXEDGAR submissions index for CIK 0001713683 (874 filings)data.sec.gov/submissions/CIK0001713683.json, downloaded2026-09-20
10K-FY25Zscaler 10-K for FY ended 2025-07-31, filed 2025-09-11sec.gov accession 0001713683-25-000158, downloaded and diffed against 10K-FY262026-09-20
8K-0416Zscaler 8-K Item 5.02 dated 2026-04-15, filed 2026-04-16sec.gov accession 0001713683-26-0000822026-09-20
8K-0903Zscaler 8-K Items 2.02/2.05 dated 2026-09-01, filed 2026-09-03sec.gov accession 0001713683-26-0001562026-09-20
13G-0730Schedule 13G filed 2026-07-30 by BlackRock, Inc.sec.gov accession 0002012383-26-0031392026-09-20
NTSK-S1Netskope Inc. Form S-1 filed 2025-08-22Edgar via Bigdata.com doc 31F037136FE7E560922DC1E3B20E44AE2026-09-20
NTSK-Q3-CALLNetskope Q3 FY26 earnings call 2025-12-11Quartr via Bigdata.com doc 0A78FBAD5C304D16887E8C5D97A930682026-09-20
PANW-Q4-CALLPalo Alto Networks Q4 FY26 earnings call 2026-09-01Quartr via Bigdata.com doc B20326DA02DCAC721A31FB13674355D22026-09-20
PANW-Q3-CALLPalo Alto Networks Q3 FY26 earnings call 2026-06-02Quartr via Bigdata.com doc FFAB1818554EAC176C51E649262386F62026-09-20
NET-Q3-CALLCloudflare Q3 2025 earnings call 2025-10-30Quartr via Bigdata.com doc 9D453B66B0A789AF61054C4240DED9DE2026-09-20
AV-NEWSAlpha Vantage NEWS_SENTIMENT feed for ZS, 50 items to 2026-09-14 (headlines only)Alpha Vantage2026-09-20
8K-0924Zscaler 8-K Item 5.02 dated 2026-09-24, filed 2026-09-24 (CRO transition)sec.gov accession 0001713683-26-0001932026-09-25
Links to other companies 72 links
holds 15officer of 8competitor 8director of 7acquired 6context 5event 4key person risk 2channel concentration 2technology partner 2litigation watch 2governance structure 1shareholder proposal 1insider transaction 1issuer of 1auditor of 1counsel to 1channel partner 1partner and competitor 1supplier dependency 1customer of 1regulator of 1
FromLinkToDateWhere it says so
Jay (Jagtar Singh) Chaudhryofficer ofZS2026-09-0310K-FY26 — signatures page (chunk 425)
Jay (Jagtar Singh) Chaudhrykey person riskZS2026-09-0310K-FY26 — Risk Factors, key personnel (chunk 127)
Kevin Rubinofficer ofZS2026-09-0310K-FY26 — signatures page (chunk 425)
Mike Richofficer ofZS2026-06-16AV-INSIDER — reporting title on 2026-06-16 transaction
Mike RichcontextZS2026-05-26Q3-CALL — Q&A, Brad Zelnick question
Mike Richkey person riskZS2026-09-248K-0924 — Item 5.02, first paragraph
Ross Tackettofficer ofZS2026-09-248K-0924 — Item 5.02, first paragraph
Adam Gellerofficer ofZS2026-06-22AV-INSIDER — reporting title
Robert Schlossmanofficer ofZS2026-01-12AGM-26 — chunk 2
Raj Judgeofficer ofZS2025-11-21DEF14A-25 — Director Independence (chunk 27)
Kimberly Watkinsofficer ofZS2026-09-03Q4-CALL — opening
Karen Blasingdirector ofZS2025-11-21DEF14A-25 — chunk 27
Andrew Browndirector ofZS2026-01-12AGM-26 — chunk 12
Scott Darlingdirector ofZS2026-01-12AGM-26 — chunk 12
David Schneiderdirector ofZS2026-01-12AGM-26 — chunk 12
James Beerdirector ofZS2025-11-21DEF14A-25 — chunk 27
Charles Giancarlodirector ofZS2025-11-01DEF14A-25 — beneficial ownership table (chunk 219)
Eileen Naughtondirector ofZS2025-11-21DEF14A-25 — chunk 27
ZSgovernance structureZS2026-01-12AGM-26 — chunk 12
James McRitchieshareholder proposalZS2026-01-12AGM-26 — chunk 6
Ajay Mangal, as trusteeholdsZS common stock (NASDAQ)2025-11-01DEF14A-25 — beneficial ownership table (chunk 219)
Jay (Jagtar Singh) ChaudhryholdsZS common stock (NASDAQ)2025-11-01DEF14A-25 — beneficial ownership table (chunk 219)
Jay (Jagtar Singh) Chaudhryinsider transactionZS common stock (NASDAQ)2026-09-16NEWS — The Globe and Mail 2026-09-19 summarising Form 4 (Bigdata.com doc 93ECCE0C6F73646DF094A45320947737)
The Vanguard GroupholdsZS common stock (NASDAQ)2025-11-01DEF14A-25 — chunk 219
BlackRockholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — company_funds_holders rank 2
Jericho Capital Asset ManagementholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 1
Millennium ManagementholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 3
Citadel AdvisorsholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 5
Norges BankholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 8
T. Rowe Price AssociatesholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 14
Artisan PartnersholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 11
ZSissuer of0.0% Convertible Senior Notes due 2028 ($1,725m)2026-07-3110K-FY26 — Note 10 Convertible Senior Notes (chunk 602)
PricewaterhouseCoopers LLPauditor ofZS2026-09-0310K-FY26 — Report of Independent Registered Public Accounting Firm (chunks 287, 290)
Wilson Sonsini Goodrich & Rosaticounsel toZS2026-01-12AGM-26 — chunk 2
ZSacquiredRed Canary, Inc.2026-07-3110K-FY26 — Note 6 Business Combinations (chunk 592)
ZSacquiredSymmetry Systems, Inc.2026-07-3110K-FY26 — chunk 592
ZSacquiredSquareX Holdings, Inc.2026-07-3110K-FY26 — chunk 592
ZSacquiredSPLXAI Inc.2026-07-3110K-FY26 — chunk 592
ZSacquiredAirgap Networks Inc.2024-04-1210K-FY26 — chunk 592
ZSacquiredAvalor Technologies Ltd.2024-03-0810K-FY26 — chunk 592
ZSchannel concentrationUnnamed channel partner (13% of receivables)2026-07-3110K-FY26 — Revenue Recognition narrative (chunk 720) and channel risk factor (chunk 93)
ZSchannel partnerCarahsoft2026-09-03Q4-CALL — CFO prepared remarks
ZStechnology partnerAnthropic2026-09-09PR-SOC — chunk 2
ZStechnology partnerOpenAI2026-09-09PR-SOC — chunk 2
ZSpartner and competitorMicrosoft2026-06-03CONF-0603 — chunk 34
ZSsupplier dependencyZS2026-09-03Q4-CALL — CFO capex remarks
Palo Alto NetworkscompetitorZS2026-09-14NEWS — Morningstar 'Zscaler Stock Fell 5%...' (Bigdata.com doc 7C19A3F6FFEF79FFA7DBC2D7ACFEFD6E)
CrowdStrikecompetitorZS2026-09-14NEWS — same
FortinetcompetitorZS2026-09-14NEWS — same
Edward Jonescustomer ofZS2026-06-09CONF-0609 — chunks 45-46
U.S. Securities and Exchange Commissionregulator ofZS2026-09-0310K-FY26 — cover
ZSlitigation watchZS2026-09-0310K-FY26 — Legal Matters (chunk 500)
ZSeventZS2026-09-03Q4-CALL — CFO prepared remarks
ZSeventZS2026-09-03Q4-CALL — Q&A Saket Kalia
Raj JudgeeventZS2026-04-158K-0416 — Item 5.02
ZSeventZS2026-09-018K-0903 — Item 2.05
BlackRockholdsZS common stock (NASDAQ)2026-07-3013G-0730 — cover page rows 9 and 11
ZScontextZS2026-09-0310K-FY26 — Risk Factors, customer retention paragraph (text line 424 of local extraction); 0 hits in 10K-FY25
ZScontextZS2026-09-20EDGAR-IDX — form-type counts over filings.recent
ZScontextZS2026-07-3110K-FY26 — Item 1 Business and Item 7 Key Factors
ZSchannel concentrationZS2026-07-3110K-FY26 — Note on revenue by sales channel and by geographical region
Netskope, Inc. (NTSK)competitorZS2025-08-22NTSK-S1 — Competition (chunks 109, 498, 532)
Netskope, Inc. (NTSK)competitorZS2025-12-11NTSK-Q3-CALL — CFO prepared remarks (chunk 29) and Q&A (chunk 61)
Palo Alto NetworkscompetitorZS2026-09-01PANW-Q4-CALL — CEO prepared remarks (chunks 6, 27)
Palo Alto NetworkscompetitorZS2026-06-02PANW-Q3-CALL — Q&A (chunk 71)
Cloudflare, Inc.competitorZS2025-10-30NET-Q3-CALL — Q&A (chunk 61)
InvescoholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — company_funds_holders rank 20
Fidelity InvestmentsholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 19; top_sells_history Q1
Tiger Global ManagementholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 17
American Century CompaniesholdsZS common stock (NASDAQ)2026-06-3013F-Q2-26 — rank 16
ZSlitigation watchZS2026-08-12AV-NEWS — feed item 2026-08-12 Bloomberg Law News, headline only; Gross Law PR via web search
Federal Reserve (FOMC)contextZS2026-09-16NEWS — Federal Reserve press release via Quartr (Bigdata.com doc F89AA6B1CB09F72EE11E548CC09D3CFD); FXStreet 2026-09-18 (409300D278CFDBC9D65D1A1735972BB0)
History 7 entries
  • 2026-09-19 — first pass, model-drafted in a supervised session; see dossier-2026-09-19.md for evidence and access limits.
  • 2026-09-20 — full pipeline re-run in a supervised session on owner instruction: sec.gov reachable, FY26 10-K read and diffed, competitor calls, 13F detail, Whale Rock and Druckenmiller handoffs; see dossier-2026-09-20.md. S8 proposal added; owner has not committed.
  • 2026-09-20 — correction to the 2026-09-19 dossier's access-limits note: sec.gov, data.sec.gov and efts.sec.gov are NOT blocked by the egress policy. The 403s were a missing User-Agent header, which SEC requires; with one, all three return 200 (verified 2026-09-20). S3 and S5 can be redone against primary EDGAR documents, and S2's share-count trend against the XBRL companyconcept API. Separately, Bigdata.com — the source behind most of that dossier — now reports out of credits. Dossiers are never edited after their date, so this log carries the correction. See ledger/research/WATCHLIST_PLAN.md §1.
  • 2026-09-20 — two gates advanced using the newly authorised Equibles connector, no model judgement involved. S5 shelf/ATM, previously "could not be confirmed": no ATM program captured across roughly the last 13 months of 10-K/10-Q/8-K filings. The source is explicit that absence there is not proof the company runs none, so the scorecard says the same. S3 red-flag sweep gains a real finding: channel-partner concentration at 85% of FY26 revenue, down from 88% in FY25 and 91% in FY24, with one channel partner at 13% of receivables (12% the prior year). Structured XBRL from the 10-K filed 2026-09-03 and the 10-Q filed 2026-05-26; both filing URLs are in the connector's output. Going-concern status checked at the same time: no flag on the newest examined filing.
  • 2026-09-21 — owner decision: Stage 0's circle-of-competence and exclusion-list checks are dropped for this project (WATCHLIST_PLAN.md §16). The exclusion check, open on this scorecard since 2026-09-19, is therefore closed rather than answered, and S0's gate is met on the four machine-checkable parts alone. S6, S7 and S8 owner items are unaffected.
  • 2026-09-23 — stage redefinition, owner decision (WATCHLIST_PLAN.md §25): S8 is now Management credibility and is not yet run for this name; the old S8 valuation figures moved to the valuation: line under S7, and any proposed buy price or position size was removed (the pipeline no longer drafts either). No research was redone. S2 capital allocation and S3 earnings quality are also pending, in the same backfill slot.
  • 2026-09-25 — backfill, slot backfill-2026-09-23-7, model-drafted unattended by the cloud research routine. Wrote the three checks added on 2026-09-23 (WATCHLIST_PLAN.md §25) into dossier-2026-09-25.md and updated this file: S8 management credibility (15 rows, 14 beat / 1 missed — conservative), S2 capital allocation (neutral) and S3 earnings quality (3 flags). No other stage was redone. Sources: the Ex. 99.1 outlook tables of six earnings 8-Ks (2024-09-03 through 2026-09-03), the Item 5.02 8-Ks of 2025-06-13 and 2026-09-24, the DEF 14A of 2025-11-21, XBRL companyfacts, and all eight archived market data transcripts (FY2025-Q1 to FY2026-Q4), each of which was checked to name its own fiscal quarter in its opening turn. Run finding 1: the beat record is a guidance spread. Fourteen of fifteen first guides were beaten, but the proxy prints the board's own FY2025 plan — revenue $2,678.7m, billings $3,273.2m, non-GAAP operating income $573.2m — and the public guide was set 2-6% below it while the results came in at 99.8%, 99.2% and 101.2% of plan. Zscaler beats the market's number and lands on its board's. Run finding 2: the single miss is the free-cash-flow margin, which was guided at 26-26.5%, raised, then cut to 22.8-23.3% as capex went to low teens as a share of revenue — the one guide that did not depend on selling, and the one the FY27 model should not take at face value. Run finding 3, found in a filing one day old: an Item 5.02 filed 2026-09-24 says CRO Mike Rich steps down effective 2026-10-01, three weeks after the Q4 call on which both the CEO and the CFO named him as the executive building the FY27 sales organisation, and after two of his leaders had already left in Q3 FY26. That fires the kill criterion "another senior sales departure before 1H FY27 ends" written on 2026-09-20. Also opened: the dodge log, whose two-quarter thread is the same sales-organisation question. Nothing here is human-verified.
Full notes, 25 Sep 2026 15 min read

ZS — backfill dossier, 2026-09-25

Slot backfill-2026-09-23-7, model-drafted unattended by the cloud research routine. Nothing here is human-verified.

Scope. Only the three checks added on 2026-09-23 (WATCHLIST_PLAN.md §25): S8 management credibility, S2 capital allocation and S3 earnings quality. No other stage was redone; the 2026-09-19 and 2026-09-20 dossiers stand. Dossiers are never edited after their date, so anything this pass corrects is stated here and carried into scorecard.md.

Sources used, all primary.

TagDocumentHow it was read
8K-2409038-K filed 2024-09-03, Ex. 99.1 (Q4/FY2024 results and first FY2025 outlook)accession 0001713683-24-000097, zs-07312024_991.htm, sec.gov
8K-2509028-K filed 2025-09-02, Ex. 99.1 (Q4/FY2025 results and first FY2026 outlook)accession 0001713683-25-000138, zs-07312025_991.htm
8K-2511258-K filed 2025-11-25, Ex. 99.1 (Q1 FY2026)accession 0001713683-25-000203, zs-10312025_991.htm
8K-2602268-K filed 2026-02-26, Ex. 99.1 (Q2 FY2026)accession 0001713683-26-000047, zs-01312026_991.htm
8K-2605268-K filed 2026-05-26, Ex. 99.1 (Q3 FY2026)accession 0001713683-26-000095, zs-04302026_991.htm
8K-2609038-K filed 2026-09-03, Ex. 99.1 (Q4/FY2026 results and first FY2027 outlook)accession 0001713683-26-000156, zs-07312026_991.htm
8K-2609248-K filed 2026-09-24, Item 5.02 (CRO transition)accession 0001713683-26-000193, zs-20260924.htm
8K-2506098-K filed 2025-06-13, Item 5.02 (Canessa transition agreement)accession 0001713683-25-000103
DEF14A-25DEF 14A filed 2025-11-21accession 0001713683-25-000198
XBRLdata.sec.gov/api/xbrl/companyfacts/CIK0001713683.json10-K annual facts only, FY2020–FY2026
ARCHIVEledger/research/_transcripts/ZS/FY2025-Q1 … FY2026-Q4 (8 calls, market data)read in the §3.1 order; turn numbers below are line indices in the archived content string

Transcript integrity check (the guard the 2026-09-24 AX pass asked for): each of the eight ZS files names its own fiscal quarter in the opening turn, and all eight agree with the key they are filed under. No AX-style mislabel here.


S8 — management credibility

Who made the promises. Two CFOs cover the window. Remo Canessa announced his retirement on the Q1 FY2025 call itself (2024-12-02, turn [3]: "after eight amazing years, I have decided to retire as Chief Financial Officer of Zscaler") and signed a transition agreement on 2025-06-09 running to 2025-12-31 (8K-250609). Kevin Rubin was appointed CFO in May 2025 (DEF14A-25, "In May 2025, upon his appointment as CFO, Mr. Rubin received…") and gave his first guide on the FY2025 Q4 call, 2025-09-02. Jay Chaudhry has been Chairman and CEO throughout. Roles are taken from EDGAR, not inferred from the call; the call date is the locator.

The say-vs-do table

15 scored rows. [n] is the turn index in the archived transcript; an 8K tag is the Ex. 99.1 outlook table.

#PromiseMadeByFirst guidePathResultVerdict
1FY2025 revenue8K-240903Canessa (CFO)$2.60–2.62bn→$2.623–2.643bn (Q1) →$2.64–2.654bn (Q2) →$2.659–2.661bn (Q3)$2,673.1m (8K-250902)beat (departed)
2FY2025 non-GAAP EPS8K-240903Canessa$2.81–2.87→$2.94–2.99 →$3.04–3.09 →$3.18–3.19$3.28beat (departed)
3FY2025 calculated billings8K-240903Canessa$3.110–3.135bn→$3.124–3.149bn →$3.153–3.168bn →$3.184–3.189bn$3,246.2mbeat (departed)
4Q2 FY2025 revenue2024-12-02 [3]Canessa$633–635m—$648m (2025-03-05 [3])beat (departed)
5Q3 FY2025 revenue2025-03-05 [3]Canessa$665–667m—$678m (2025-05-29 [3])beat (departed)
6Q4 FY2025 revenue2025-05-29 [3]Canessa$705–707m—$719.2m (8K-250902)beat (departed)
7FY2026 revenue8K-250902Rubin (CFO)$3.265–3.284bn→$3.282–3.301bn (Q1) →$3.309–3.322bn (Q2) →$3.3295–3.3325bn (Q3)$3,353m (8K-260903)beat
8FY2026 ARR8K-250902Rubin$3.676–3.698bn→$3.698–3.718 →$3.730–3.745 →$3.740–3.749$3,771mbeat
9FY2026 non-GAAP EPS8K-250902Rubin$3.64–3.68→$3.78–3.82 →$3.99–4.02 →$4.10–4.11$4.21beat
10Q1 FY2026 revenue8K-250902Rubin$772–774m—$788.1m (8K-251125)beat
11Q2 FY2026 revenue8K-251125Rubin$797–799m—$815.8m (8K-260226)beat
12Q3 FY2026 revenue8K-260226Rubin$834–836m—$850.5m (8K-260526)beat
13Q4 FY2026 revenue8K-260526Rubin$875–878m—$898.2m (8K-260903)beat
14FY2026 free-cash-flow margin2025-09-02 [3]Rubin~26% to 26.5%→26.5–27% (raised, 2026-02-27 [3]) →22.8–23.3% (cut, 2026-05-26 [3])23% ($779.1m on $3,353m, 8K-260903)MISSED
15FY2026 Red Canary ARR2025-09-02 [3]Rubin~$95m→~$130m (Q2) →~$137m (Q3)$141m (8K-260903)beat

Split by who is still here: 9 rows belong to the current team (8 beat, 1 missed), 6 to Canessa, who has left (all 6 beat). The verdict rests on the 9.

Verdict: ✅ conservative — with one finding that changes how the beats should be read

Fifteen first guides, fourteen beaten. Under §25's definition that is conservative. But the proxy shows what the beats are made of, and this is the run's main S8 finding.

The public guide sits below the board's own plan, and the plan was not beaten. DEF14A-25 prints the internal FY2025 targets the compensation committee set from the board-approved operating plan:

FY2025 metricPublic first guide (8K-240903)Board plan target (DEF14A-25)Guide vs planActualActual vs plan
Revenue$2.60–2.62bn$2,678.7m−2.2% at the high end$2,673.1m99.8% — short
Calculated billings$3.110–3.135bn$3,273.2m−4.2%$3,246.2m99.2% — short
Non-GAAP income from operations$530–540m$573.2m−5.8%$580.1m101.2% — ahead

The proxy's own attainment table says the same thing in the company's words: revenue funded at 99.5% and 98.5% of target in the two halves, billings at 95.5% and 96.5%, only non-GAAP income from operations above 100. The full-year bonus pool funded at 100.9% of target.

So the record is: Zscaler beats what it tells the market and lands on what it tells its board. Both readings are true and the second is the one a model should carry. The guide is a floor, not a forecast — which is the right way to read the FY2027 guide of 16.6–17.4% ARR growth (8K-260903) — but the beats are not evidence of operating upside against plan.

The one miss is the one guide that did not depend on selling. Free-cash-flow margin was guided at 26–26.5% for FY2026, raised to 26.5–27% in February, then cut to 22.8–23.3% in May against capex going from "mid-single digits" to "high single digits" to "low teens" as a percentage of revenue. Q4 alone carried $218.5m of capex and internal-use software against $78.7m a year earlier (8K-260903). Management named the cause on the call and it is external — Chaudhry, 2026-05-26 [51]: "this situation you're talking about is fairly unique. It's created by all these AI data centers that's causing such a shortage of so many parts." FY2027 is guided at 23.0–23.5% FCF margin on "low teens" capex. That is the one line in the FY2027 guide this management has a miss against, and it is a cost line they do not control.

Promises belong to people — and one left the day before this pass

Item 5.02, filed 2026-09-24 (8K-260924): "Zscaler, Inc. … is promoting Ross Tackett to be the Company's Chief Revenue Officer effective October 1, 2026. Mr. Tackett will succeed Mike Rich, who will step down as an officer effective October 1, 2026, but will remain with the Company as a strategic advisor for a transition period through December 31, 2026."

That filing is three weeks after the Q4 FY2026 call on which both officers named Mike Rich as the person executing the FY2027 sales plan:

  • Chaudhry, 2026-09-03 [26]: "Our CRO, Mike Rich and his team evolved the idea of take-up teams that we have started about 3 years ago."
  • Rubin, 2026-09-03 [27]: "Mike is adding people to the organization, both to address opportunities we have with new logos as well as upsell."

It is also the third departure in this reporting line inside two quarters. On the Q3 FY2026 call (2026-05-26 [6]) Chaudhry said of two departing sales leaders: "these 2 leaders are part of our CRO, Mike Rich's team"; Rubin confirmed on the Q4 call [7] that "we did have 2 sales leaders depart last quarter. One was a geo leader and the other was a vertical leader", that the vertical leader was backfilled internally and the geo role only recently accepted, and that "we expect that the leadership transition will play out in the first half of this fiscal, and that is reflected in our guidance." The FY2027 guidance reflects two departures. It was set before the third.

This fires an existing kill criterion. The scorecard's list, written 2026-09-20, includes "Another senior sales departure before 1H FY27 ends." 1H FY2027 ends 2027-01-31. The CRO's departure is effective 2026-10-01. The criterion is met — as a fact, not a judgement; what to do about it is the owner's call.

No S8 row is scored against Mike Rich, so no row moves to departed on this. The exposure is prospective: the FY2027 ARR guide (row for a future pass) was built on a sales organisation that has since lost its head.

Dodge log roll-up

Opened this pass from the two most recent calls (§3.1). Four dodges, and one topic dodged in two consecutive quarters: what inside the guide produces the net-new-ARR deceleration, and what the sales-leadership disruption is worth in numbers.

  • FY2026-Q3 (2026-05-26) · quantified effect of the sales-leadership departures · Brad Zelnick [5] — asked three things: were the departures voluntary or involuntary, how senior were they, and "What would your guidance have been if these leaders were still in place?" Chaudhry [6] answered none of the three ("we just want to be prudent"); Rubin [7]: "I don't have much more to offer other than we are taking a prudent approach."
  • FY2026-Q3 (2026-05-26) · sales productivity inside the FY2027 guide · Fatima Boolani [43] — asked for quantification of productivity, attrition and hiring assumptions behind the 16–17% guide. Rubin [44] gave a directional statistic ("sixth straight quarter of sales productivity growth") and declined the quantification as "a bit early".
  • FY2026-Q4 (2026-09-03) · what decelerates net new ARR in the FY2027 guide · Ittai Kidron [41] — asked for the puts and takes that take net-new-ARR growth "to 4% at the midpoint". Rubin [42] answered with the accelerants (momentum, NRR 115%, new-logo hires, product breadth) and never addressed the decelerating components.
  • FY2026-Q4 (2026-09-03) · composition of organic net new ARR · Fatima Boolani [9] — asked to "stack rank" the discrete drivers of the organic acceleration. Chaudhry [10] and Rubin [11] answered "broad-based" with a product list and no ranking.

Not dodges, logged so a later pass does not re-file them: the refusal to give a FY2027 growth assumption for ZIA/ZPA specifically (Suppiger, 2026-05-26 [34]/[35] — an explicit "we didn't provide that level of granularity"), and the Symmetry/SquareX contribution question, which was answered as immaterial (Rubin, 2026-05-26 [41] and 2026-09-03 [23]).


S2 — capital allocation and returns (FY2022–FY2026)

All figures from XBRL 10-K annual facts unless tagged. $ millions.

Where five years of operating cash flow went

FY2022FY2023FY2024FY2025FY20265-yr
Operating cash flow321.9462.3779.8972.51,129.73,666.3
Capex (PP&E)(69.3)(97.2)(144.6)(164.3)(277.3)(752.7)
Capitalised internal-use software(21.3)(31.5)(50.3)(81.5)(73.2)(257.8)
Free cash flow231.3333.6585.0726.7779.12,655.7
Acquisitions, cash net(25.3)(15.6)(374.7)(0.8)(918.1)(1,334.5)
Buybacks000000
Dividends000000
Debt raised / (repaid)———+1,725.0 / (1,150.0)—net +575.0
Stock-based compensation409.6444.8527.7661.4821.92,865.4
FCF less SBC(178.3)(111.2)+57.3+65.3(42.8)(209.7)
Diluted weighted-average shares (m)140.9144.9149.6154.4160.2+13.7%
Goodwill impairment000000

The 2025 convertible line: the $1,150.0m 2025 Notes were repaid in cash and the premium settled by issuing 3.8 million new shares; $1,725.0m of 2028 Notes were issued for $1,700.0m net (8K-250902). Goodwill went $78.5m (FY2022) → $1,218.1m (FY2026) and acquired intangibles $31.8m → $214.4m, all of it bought, none of it yet impaired.

Did the buybacks more than offset SBC? There were no buybacks. The last repurchase on the XBRL record is $22 thousand in FY2019. Dilution ran at +3.3% a year for five years and was funded entirely by issuance.

What it earned

ROIC is not meaningful here and saying otherwise would be a fiction. GAAP operating income was negative in every one of the five years (−327.4, −234.6, −121.5, −128.5, −133.3) and the cumulative five-year GAAP net loss is $755.0m, so NOPAT is negative throughout and the incremental return is a narrowing loss, not a return: operating loss improved by $194.1m while shareholders' equity grew $573.3m → $2,598.3m and total assets $2,832.7m → $7,866.7m. The company also runs a net-cash balance sheet, which makes an invested-capital denominator arbitrary.

What can be measured is the cash return per share, and it is good:

Per diluted shareFY2022FY2026Change
Revenue$7.74$20.92×2.70
Free cash flow$1.64$4.86×2.96
Free cash flow less SBC−$1.27−$0.27still negative

Revenue grew ×3.07 and free cash flow per share grew ×2.96 through 13.7% dilution. That is the case for the equity issuance having bought something.

What the acquisitions were worth

No goodwill has been impaired in five years, so the accounting says nothing yet. The relevant disclosure is management's own, on the largest one. Red Canary cost $675m of the FY2026 cash (announced 2025-05-27, closed 2025-08-01) and exited FY2026 with $141m of ARR against $83m recognised at close. Asked about it on the Q4 call, Rubin said (2026-09-03 [33]): "as it relates to Red Canary, we're not expecting any net new ARR contribution" in FY2027, because it becomes part of an integrated offering — and [35]: "We did see elevated churn in Red Canary's business in '26." The book was held through integration; whether it compounds is a FY2027 question and the console should not credit it before then.

What the bonus is paid on

DEF14A-25, FY2025 plan: cash bonus funded on revenue 30%, calculated billings 30%, non-GAAP income from operations 40%; PSUs on multi-year ARR targets ($4bn / $5bn / $6bn ladders). The §25 flag (EPS-based pay plus buybacks) does not fire — there is no EPS metric and no buyback. The observation worth writing down is the opposite one: not one pay metric is per-share, cash-flow or return-based. All four are gross growth measures, and the 40%-weighted one is struck before stock-based compensation by construction. Pay is therefore indifferent to the dilution (+3.3%/yr) and to the free-cash-flow margin — the single guide this management missed.

Verdict: neutral

Not a destroyer: the core compounds, revenue and free cash flow per share both roughly tripled through the dilution, there is no impairment and no debt problem. Not a compounder on returns either: five years, no GAAP profit, $755m of cumulative GAAP losses, stock compensation larger than all the free cash flow it produced, nothing returned to owners, and $1,334.5m spent on acquisitions whose largest member is guided to add no net new ARR next year. Neutral, and the thing that would move it is the FY2027 capex year — $752.7m of capex over five years with $277.3m of it in FY2026 alone is a different business model from the one the earlier years describe.


Earnings quality

EQ: 3 flags.

1. Stock-based compensation is bigger than the free cash flow. FY2026 SBC $821.9m = 24.5% of revenue and 105.5% of free cash flow ($779.1m). FCF after SBC is −$42.8m for FY2026 and −$209.7m across the five years. §25's test — FCF after SBC below half of reported FCF — is not merely failed, it is negative. Flag.

2. The non-GAAP bridge is widening faster than revenue, and the new width is bought. Reconciliation from 8K-260903:

Add-back to GAAP loss from operationsFY2025FY2026Change
Stock-based comp and related payroll taxes685.5841.6+22.8%
Amortisation of acquired intangibles16.843.5+158.4%
Restructuring and other charges4.910.3+108.4%
Acquisition-related expenses1.35.0+279.9%
Total add-backs708.6900.4+27.1%
(memo) revenue2,673.13,352.5+25.4%

Add-backs grew faster than revenue, and the acceleration is entirely acquisition-related — including $57.4m of acquisition-related stock compensation inside the SBC line (FY2025: $33.1m; footnote 2). The measure being widened is the same non-GAAP income from operations that carries the 40% weight in the bonus pool. Flag.

3. "Restructuring and other charges" is a recurring one-off. It appears in FY2023 ($7.6m), FY2025 ($4.9m) and FY2026 ($10.3m) — three of the last four years — and a fourth, five times larger, was announced on the FY2026 Q4 call before the fiscal year was two months old: a reduction of about 3% of employees at "about $30 million to $33 million in restructuring charges" (Rubin, 2026-09-03 [27]). A charge taken in four years out of five is an operating cost. Flag.

Clean, with the locator:

  • Cash conversion. Net income is negative, so OCF/NI is not defined; against non-GAAP net income it is 1.88× (FY2024), 1.82× (FY2025), 1.60× (FY2026) — still strong, drifting one way, worth a second look next year. Accruals, (net income − operating cash flow) ÷ average total assets, are deeply negative and therefore conservative: −20.1%, −18.2%, −16.7%.
  • Working capital. Days sales outstanding, receivables ÷ revenue × 365: 133.8 (FY2022), 131.5, 124.0, 135.5, 125.1 (FY2026). It swings with the Q4-weighted billing year and does not rise two periods running. No flag. Deferred revenue $2,480.5m current at FY2026 against receivables $1,149.1m.
  • Costs moved onto the balance sheet. Capitalised internal-use software rose two years running (2.32% then 3.05% of revenue in FY2024–25) and then fell to 2.18% in FY2026. The balance-sheet growth in FY2026 is in data-centre property, not software. No flag, but this is the series to re-check next year given the capex step-up.
  • No goodwill impairment in five years, no restatement, no 10-K/A or 10-Q/A, no NT 10-K.

What this changes on the scorecard

  • S8 ⬜ → ✅ conservative, 15 rows, 14 beat / 1 missed, with the proxy finding beside it.
  • S2 stays ✅; gains the capital-allocation table and the verdict neutral.
  • S3 stays ✅; gains EQ: 3 flags.
  • S7 valuation: no guidance haircut (the verdict is conservative), but the FY2027 revenue and the FY2027 free-cash-flow guides must be treated differently — see the scorecard's · valuation continuation line.
  • ⚠️ the CRO departure of 2026-09-24, and the fact that it fires an existing kill criterion.
  • Dodge log opened, with the two-quarter thread flagged.
Full notes, 20 Sep 2026 18 min read

Zscaler (ZS) — research dossier, 2026-09-20

Second pass, drafted by Claude in a supervised session on the owner's instruction to run the stock-sourcing pipeline against every available source. Nothing here is human-verified. Retrieved sources are named per item; items marked [background] come from Claude's prior knowledge, not from a document retrieved this session. This file supersedes nothing: dossier-2026-09-19.md stays as the first snapshot and is referenced where its findings were restored rather than redone.

Access this session (what changed since 2026-09-19)

  • sec.gov is reachable now (it was blocked yesterday). The EDGAR submissions index (CIK 0001713683, 874 filings) and the full FY26 and FY25 10-K documents were downloaded and read mechanically (grep and heading diff), plus the 8-Ks of 2026-04-16 and 2026-09-03 and the Schedule 13G of 2026-07-30.
  • FMP: only profile-symbol works on the current plan; statements, analyst, filings, insiders, quotes and 13F endpoints are all denied. The Bigdata.com company tearsheet carries FMP-sourced fundamentals and estimates, which is how the 5-year table below was obtained.
  • Alpha Vantage: quote, insider transactions, news feed, fed funds and WTI worked; the free-tier daily cap was then hit, so the 10-year yield and 200-day average did not return. The 50-day average did (167.61 on 2026-09-18). The Q4 FY26 transcript was not re-pulled (one call per ticker per run; it was pulled 2026-09-19).
  • Bigdata.com: filings, Quartr transcripts, premium news, podcasts, WSRA 13F, LinkUp jobs. Its investment-research category returned nothing for Zscaler in Aug–Sep 2026.
  • Blocked by egress policy: cnbc.com, zscaler.com, glassdoor.com.
  • Owner's Gmail and Google Drive were searched for "Zscaler": no owner notes, models or saved research exist. The only hits were unrelated (Cboe newsletters, generic 2026 outlook PDFs, personal correspondence). The idea has no prior owner work behind it.
  • The board contributes nothing: no podcast transcript in ledger/pipeline/gold/transcripts/ mentions Zscaler, and no claim row references it. Bigdata's podcast index has Zscaler only in company-hosted appearances (Schwab Network CFO interview 2026-09-11, Motley Fool CEO interview 2026-01-04) and one customer aside on The Stack Overflow Podcast.

Snapshot (price 2026-09-18 close; Alpha Vantage GLOBAL_QUOTE, FMP profile)

ItemValueSource
Price / market cap$197.31 / $31.9bnFMP profile-symbol
52-week range$114.63 – $336.99FMP profile-symbol
Average daily volume4.07m shares (≈$800m/day)FMP profile-symbol
50-day SMA$167.61Alpha Vantage SMA
Shares outstanding 2026-07-31163,055k (158,301k FY25; 152,490k FY24)10-K FY26 balance sheet
Cash & ST investments$3,474m10-K FY26 via Bigdata tearsheet (FMP)
Debt$1,725m 0% converts due 2028-07-15, carrying $1,696m10-K FY26
Net cash / EV≈ $1.75bn / ≈ $30.2bnderived
Short interest (2026-08-31 settlement)6.35m shares, 2.92 days to coverFINRA via Equibles (web search)
Consensus (FMP via Bigdata)42 buy / 10 hold / 1 sell; PT median $200, high $298 (Bernstein 2026-09-17), low $150tearsheet analyst_data
Consensus FY27 / FY28 revenue$3,927m (34 analysts) / $4,566m (31)tearsheet estimates
Consensus FY27 EPS$4.88 (22 analysts)tearsheet estimates
Next eventsInvestor Day 2026-10-06 NYC (Q4 call); Q1 FY27 call late Nov (prior dossier)Q4-CALL

Five-year financials (FMP via Bigdata.com tearsheet, FY ends July 31, $m)

FYRevenueGrowthGAAP op incGAAP net incSBCOCFFCF (FMP)Wtd dil. shares (m)
20221,091—-327-390410322231140.9
20231,61748%-215-202445462334144.9
20242,16834%-124-58528780585149.6
20252,67323%-128-41661972727154.4
20263,35325%-133-638221,130852160.2

Company-defined FCF (OCF less capex less capitalised software) was $779m in FY26, $727m FY25, $585m FY24 (10-K FY26 non-GAAP reconciliation). Non-GAAP income from operations $767m / $580m / $442m. Share count compounds at about 3.5% a year; FY27 guide assumes 173m diluted shares. SBC was 24.5% of FY26 revenue and $1,806.5m is unrecognised.

Stage-by-stage findings

S0 Universe & fit — restored from 2026-09-19 (3/4), one refinement

The 10-K FY26 states the scale claim directly: "over 750 billion transactions, blocks over nine billion policy violations and threats and performs over 250,000 unique security updates per day" and that this "creates a structural data advantage" (Item 1). Liquidity is not a constraint at ~$800m traded per day. The falsifier from yesterday stands and got a second leg this session: the 10-K's competition risk factor says "AI technologies may lower barriers to entry for new competitors, including large AI providers and hyperscalers", and Futurum (2026-06-12) frames Microsoft Agent 365 as the natural home for agent inventory, which would make SASE platforms "enforcement layers that depend on inventory they don't own". Exclusion-list check remains the owner's.

S1 Source — owner request, no thesis attached (restored). No prior owner work found in Gmail or Drive.

S2 Kill test — passes, with the same two flags

Growth 25% (20% ex-Red Canary), 77% GAAP gross margin, net cash, company FCF $779m. Flags: (1) SBC $822m ≈ FCF, so SBC-adjusted FCF is roughly zero and the GAAP loss has persisted since IPO (accumulated deficit $1,252.7m); (2) the share count rises ~3.5% a year with no buyback offset. Neither kills the idea at this stage; both cap what the owner should pay.

S3 Filings deep dive — 10-K now read end to end (mechanically), FY25 diffed

Three biggest risks management discloses (unchanged from yesterday): channel concentration (top five partners 31% of revenue, up from 28% and 25%; one partner 13% of receivables; 85% of revenue through channel); key-person and sales-leadership dependence; larger competitors bundling "at zero or negative margins" with AI lowering entry barriers.

What changed vs the FY25 10-K (heading diff plus grep):

  • New clause in the retention risk factor: customer retention may decline because of "our customers' increasing use of AI to automate tasks traditionally performed by human workers". The phrase does not appear in the FY25 10-K (0 hits) and appears once in FY26. This is the company's own seat-compression disclosure and the single most important addition; the FY26 Q4 call answered it with the 30%-of-new-ACV non-seat figure.
  • New heading on sales productivity: risk that "new sales personnel are unable to achieve desired productivity levels in a reasonable period of time" (context: two sales-leader departures, restructuring, new-logo hunters being hired).
  • New heading: "A broad reduction in IT security spending would have a material impact".
  • Export-control language moved from "potential" to "actual or potential implementation".
  • The FY25 headings on interest-income sensitivity, FX hedging effectiveness and indemnification obligations no longer match as headings in FY26 (reworded or folded; not verified line by line).
  • Word counts: "uncertain" 31/31, "macroeconomic" 12/13, "elongat*" 0 (FY25: 2), "agent" 62 (FY25: 28). The FY25 text on customers "carefully considering purchasing decisions ... elongating our sales cycles" is gone from the FY26 MD&A.
  • Customers "approximately 11,000" at 2026-07-31 vs 9,400 and 8,650 (10-K). The 10-K does not say whether Red Canary's customers are in that count; treat the +17% as unverified organic growth. Also: "over 4,600 of the estimated 20,000 organizations with over 1,500 employees" and "over 50% of the Fortune 500".
  • Employees "over 8,700" (FY26), ~60% outside the US; R&D centres in India, Israel, Spain.
  • Geography (10-K table): US $1,792m 53%, EMEA $923m 28%, APAC $519m 15%, other 4%. US share rose from 50% to 53% over two years, i.e. international grew slower.
  • RPO $7,365.4m, 45% within 12 months, 90% within three years.
  • Goodwill $1,218m from $418m (four FY26 acquisitions); SPLX developed-tech valuation was cut from $14.1m to $3.3m within the measurement period (prior dossier, Note 6).
  • Auditor PwC since 2015; one CAM (subscription revenue recognition); Red Canary, SPLX, SquareX and Symmetry excluded from ICFR. No 10b5-1 plans adopted or terminated in Q4.
  • Legal (Note 12): "no pending or threatened legal proceeding ... likely to have a material adverse effect". Contrast with plaintiff-firm PR: The Gross Law Firm (2026-06-09, Morningstar/PR Newswire) refers to an "upcoming securities class action deadline", which in that genre means a complaint was filed after the 2026-05-27 drop. The docket was not retrieved, so the filing is recorded as reported, not confirmed. Separately, Bloomberg Law (2026-08-12, via Alpha Vantage news feed) headlined "Zscaler Accused of Patent Infringement Over AI Security Feature"; body not retrieved.

8-K sweep (EDGAR index):

  • 2026-04-16: Raj Judge, EVP Corporate Strategy and a non-independent director, resigned as officer and director effective 2026-04-15 (employment to 2026-07-31) with severance under the Change of Control and Severance Policy; "not the result of any disagreement". relationships.yaml had him as current; corrected this session.
  • 2026-09-03 Item 2.05: ~3% workforce reduction, $30–33m charge, mostly 1H FY27, "reallocating resources ... to support our AI and growth initiatives".
  • No S-3, no ATM, no NT 10-K/10-Q ever; the only shelf-type filings are annual S-8s (latest 2026-09-03). The last SEC CORRESP/UPLOAD letters date from the 2018 IPO.
  • Schedule 13G 2026-07-30: BlackRock 8,255,213 shares, 5.1%.
  • Twelve Form 4s and ~40 Form 144s since June 2026, all sales.

S4 Industry & supply chain map — competitor calls now read

Value chain (unchanged in shape): component and colocation suppliers (memory pricing is the disclosed pressure; capex guide low teens % of revenue ex-software) → Zero Trust Exchange (200+ data centres, 185 countries) → channel (85% of revenue; Carahsoft for US commercial/SMB; GSIs, telcos, marketplaces) → ~11,000 customers; pricing power sits at renewal (NRR 115% for four quarters) and in Z-Flex ($1.7bn TCV in FY26, ~30% ARR uplift).

What competitors said, from their own transcripts and filings (Bigdata.com/Quartr/Edgar):

  • Palo Alto Networks, Q4 FY26 call (2026-09-01): a $126m telecom deal "displacing legacy proxy providers with Prisma Access for SASE"; SASE "bookings and ARR growing well ahead of the overall market"; Prisma AIRS >$100m ARR in year one. Q3 call (2026-06-02), Nikesh Arora: "We are taking share from some of the, I'd say, SASE-only customers ... firewalls are not dead." Q2 call: Prisma Browser 1,500+ customers, 9m+ licences; platformised-customer NRR 119%.
  • Netskope S-1 (2025-08-22): "Our primary competitors include companies such as Broadcom, Cisco, Fortinet, Palo Alto Networks, and Zscaler"; claims May-2025 latency tests beat Palo Alto and Zscaler. Q3 FY26 call (2025-12-11): ARR $754m +34%, NRR 118%, net new ARR +44%, "all these new logos were won against primary competitors ... in bake-offs". A later Netskope quarter was not found in the index. The "wins ~70% of bake-offs vs Zscaler" figure surfaced only in a web-search summary attributed to Third Bridge and is not verified.
  • Cloudflare Q3 2025 call: Matthew Prince on Netskope's S-1 showing 95% channel sales, "seeing the same thing for a while ... for Zscaler ... the way that you sell these products is through partners. We are doubling down on that."
  • Fortinet 10-K (via Q4 2025 report): lists Zscaler among named competitors.
  • Gartner SSE Magic Quadrant 2026: Zscaler's own landing page title says "Leader in 2026 Gartner SASE & SSE Magic Quadrants" (page blocked; title from search result). The SDxCentral and BankInfoSecurity articles returned by search describe an earlier year (Palo Alto joining as third Leader) and must not be read as 2026 positioning.
  • Microsoft: no adoption data for Entra Global Secure Access was found in any indexed source; the only evidence is management's "not a meaningful factor" (self-reported, 2026-06-03) and the Futurum Agent 365 framing above.

S5 Ownership & smart-money check — complete

  • Proxy (2025-11-01): Chaudhry 16.8%; a trustee line (Ajay Mangal) 18.2%; all officers and directors 17.2%; Vanguard 7.1%. BlackRock 5.1% by 13G (2026-07-30).
  • 13F Q2-26 (WSRA via Bigdata): 747 holders, 51.2% fund ownership (55.5% prior), 137 new positions vs 133 closed, 4 funds hold it top-10 (from 1). Sellers: Invesco -4.17m shares (-97%), Fidelity -1.84m (-84%), Tiger Global closed (-1.58m), T. Rowe Price Investment Mgmt closed (-1.69m), T. Rowe Associates -63%, American Century -64%, Two Sigma -99.8%, UBS -40%, Artisan and Westfield closed. Buyers: Jericho new 1.36m, BlackRock +1.32m, Millennium +0.79m (+237%), CC&L +0.58m, Citadel +0.50m, HRT new, Voloridge +106%, Norges new. Fund put value $1.54bn vs calls $1.42bn. Reading: long-only growth managers left after the May crash; multi-strats and one concentrated fund (Jericho) came in. The other side of the trade is Fidelity/T. Rowe.
  • Insiders: every Form 4 since June is a sale; the 2026-09-15/17 cluster (Rubin 5,957, Rich 4,108, Geller 4,584 + 5,578, Schlossman 2,325, Chaudhry 2,852, Beer 177) matches RSU vest tax withholding at $192–193. No open-market purchase in twelve months.
  • Overhang: none from a shelf or ATM (none filed). The 2028 converts ($1,725m, conversion price $439.52, capped at $784.85, cash settlement of principal intended) are the only dilution instrument; the 10-K adds that note holders "may engage in short selling to hedge their position".
  • Short interest ~6% of float; Equibles model estimates 6.72m for 2026-09-15 (+5.8%).

S6 Scuttlebutt — better than yesterday, still no primary conversation

Evidence not from company materials:

  • Glassdoor (page blocked; search summaries): overall 3.5/5 on 2,550 reviews; Account Executive sub-rating 2.9/5 on 50 reviews; 2026 reviews cite layoffs (Feb, June) and "toxic management structure" (July). Weak but directionally negative on sales culture.
  • Outage: International Business Times 2026-06-12 reported service disruptions "affecting hundreds of users" with admin-UI and latency complaints; the 10-K says past disruptions were not material.
  • Customer voice: on The Stack Overflow Podcast (2026-09-11) a Zscaler customer said a colleague called it "a man in the middle attack you pay for" (context: TLS inspection).
  • Review sites (Capterra/G2/AWS Marketplace via search): latency at peak hours, configuration complexity, limited troubleshooting visibility, "premium" pricing.
  • Hiring (LinkUp via Bigdata): 276 average active postings TTM, +27.8% y/y; Q2-26 +10.7% q/q; August -3.4% m/m. Hiring did not slow into the 3% cut.
  • Sell-side channel checks: Oppenheimer 2026-08-19 "mixed ... broadly tracking in line/better" (prior dossier); Morgan Stanley 2026-05-27 "penalty box until ... new sales leader roles are filled"; Evercore downgrade 2026-05-27 on "net new logo softness".
  • Expert call / product test: OPEN (user).

S7 Written thesis + framework pressure test — Whale Rock memo (Sacerdote)

Zscaler — Whale Rock Framework Memo

Verdict: Watchlist · Conviction: 22/40 Zscaler owns the best inline zero-trust exchange in a security category whose first S-curve (users) is past its steep phase; the stock only works if the second curves (non-user workloads, branches, AI agents) inflect before Palo Alto, Netskope and Microsoft compress the price of the first.

1. S-Curve — 6/10

  • Trend: enterprise zero-trust / security service edge replacing firewall-VPN perimeters; a sub-S-curve of cloud, with a new mini-curve (securing AI agents and models) forming on top. Dishwasher model: plumbed into identity, networking and endpoint stacks.
  • Barriers removed: cloud/SaaS adoption, remote work, board-level breach fear (all done). Remaining: inertia of firewall estates (Arora: "firewalls are not dead"), latency and cost objections (review sites), procurement cycles (Z-Flex is the workaround).
  • Penetration: Zscaler serves 4,600 of ~20,000 target organisations (23%) and >50% of the Fortune 500 (10-K). For the user curve that is the 30–40% zone where exponential goes linear, and the numbers say so: revenue growth 48% → 34% → 23% → 25% (with M&A) → 17% guided. The agent-security curve is <5% penetrated [background] and unsized.
  • How tall: FY26 ARR $3.8bn against 20,000 targets × the current $1m+ cohort economics is a multi-tens-of-billions opportunity, but it is dynamic in both directions (seat compression from AI is now a disclosed risk).
  • Red flag: core curve past 30–40%; second curve real but pre-inflection.

2. Competitive Advantage — 6/10

  • Moats: scale/platform (200+ DCs, 750bn transactions/day, single policy engine), switching costs (inline proxy embedded in every user's path; NRR 115%), brand with CISOs (Gartner Leader). Not a network effect in the classic sense.
  • Leader check: leader in SSE, but contested on three fronts with evidence: Palo Alto is displacing "legacy proxy providers" in $100m+ deals and says it takes share from SASE-only vendors; Netskope grows ARR 34% at $754m with NRR 118% and wins bake-offs on data protection; Microsoft and hyperscalers own agent inventory. Ecosystem losers are the firewall-VPN incumbents (Cisco, Check Point), not Zscaler, but the winner's share is no longer uncontested.

3. Underappreciated Earnings Power — 4/10

  • Build (base): FY27 revenue $3.92bn (guide) → FY28 $4.57bn (consensus, +16%) → FY29 $5.25bn (+15%). Non-GAAP operating margin 27% → $1.42bn; tax 21%, plus ~$0.14bn other income → ~$1.25bn; ~180m diluted shares → non-GAAP EPS ≈ $6.9 in FY29. At $197 that is 28–29x three years out. GAAP: SBC of ~$1bn takes GAAP operating income to ~$0.4bn, GAAP EPS ≈ $2, ~100x.
  • Bull build: 20% growth both years → FY29 $5.65bn, 30% margin → EPS ≈ $8 → 25x.
  • Consensus already models the linear case; there is no exponential gap unless net new ARR re-accelerates above the 17% Q4 print, and management guided FY27 net new ARR to roughly +4%. The tenet fails the "cheap on the out-year number" test.

4. Super Leader — 6/10

  • Chaudhry: founder-CEO, 16.8% owner, five security companies founded [background], shipped Security for AI, SquareX browser, Agentic SOC within a year; commercial cadence is real. Against: two sales leaders left in Q3, EVP Strategy/director Raj Judge resigned in April, CFO changed in 2025, restructuring in September, Glassdoor AE rating 2.9. Talent-magnet mark is not evidenced; second-curve jump (AI agents) is being attempted, not proven.

Pre-Mortem

  • Organic net new ARR growth below 10% for two consecutive quarters (kills S-curve and earnings tenets).
  • NRR below 112% or the 10-K seat-compression clause showing up in numbers (users per customer falling while non-seat ACV fails to offset).
  • A Palo Alto or Netskope displacement of a named $10m+ Zscaler account, or Gartner 2027 moving Zscaler out of the top execution position.
  • Capex guide above low teens % of revenue or FCF margin guide below 23% at Investor Day.

Where on the Curve / Exit Discipline

User curve ~30–40% penetrated among targets (already linear). Hold only for the second curve; sell if Investor Day targets imply FY28–30 growth ≤15% with no non-seat acceleration, because then it is a 25–30x out-year P/E on a linear business.

Short-Side Mirror

Closest archetype: "in the sweet spot without an airtight moat" is too harsh; "mature and being disrupted" does not fit either. It is a block of ice in Massachusetts: $7.4bn RPO, 115% NRR and net cash mean the cash flows last for years even if growth fades. Not a short at 7.7x EV/revenue with a hiking Fed, but not a fat long either.

Research To-Do

  • Investor Day 2026-10-06: FY28–30 targets, non-seat ACV mix, net new ARR bridge.
  • Two customer or channel calls (a GSI and a Fortune 500 CISO) on Security for AI uptake and on Palo Alto/Netskope displacement pricing.
  • Gartner 2026 SSE MQ document itself (blocked this session).
  • Netskope Q2 FY27 results (due ~Sept 2026) for NRR and win-rate claims.
  • Docket check for the securities class action and the AI-feature patent suit.

Framework exercise in the style of Alex Sacerdote / Whale Rock — not investment advice.

S8 Valuation, entry plan & sizing — Druckenmiller pressure test

Macro & liquidity context. The Fed hiked 25bp to 3.75–4.00% on 2026-09-16 (12–0, first hike since 2023, Chair Warsh; Fed press release via Quartr, Benzinga), the 10-year closed the week at 5.0% after touching 5.04%, the highest since 2007 (FXStreet, Benzinga, Nasdaq), WTI October futures trade ~$100 with the Middle East keeping oil bid (MT Newswires; Alpha Vantage monthly average was $83.9 in August, $102 in May), and DXY is at a seven-week high above 100 (FXStreet). All three of Ed Hyman's death knells (dollar, oil, rates) are rising together. That is the regime in which long-duration growth multiples compress regardless of the company, and the sell-side's FY27 consensus has not been marked for it. Effective fed funds averaged 3.63% in August (Alpha Vantage).

Thesis stress-test. The charitable long: net new ARR accelerated 7% → 10% → 17% and management guided FY27 to ~4%, so the bar is on the floor; Security for AI bookings +50% q/q, Z-Flex $770m TCV in Q4, non-seat 30% of new ACV; the stock is 41% below its high at 7.7x EV/revenue and 33x EV/FCF. The questions: (1) Where is the puck in 24 months, when Palo Alto, Netskope, Microsoft and Cloudflare all have an agent-security SKU and the seat count per customer is shrinking because of the very AI Zscaler sells protection for? (2) If SBC-adjusted FCF is zero today, what is the owner actually buying at 33x reported FCF? (3) Is the 17% Q4 print organic acceleration or Q4 renewal seasonality plus Z-Flex pull-forward, and would the owner know the difference before Investor Day?

Sizing. Not a fat pitch. It is a good company at a fair-to-full price in a hostile liquidity regime, with the two facts that make Druckenmiller-style conviction (a variant perception and a macro tailwind) both absent. A starter is legitimate under "invest then investigate" only if the investigating (Investor Day, two channel calls) actually happens within the quarter.

Proposed entry plan (Claude's proposal; the owner commits, S8 stays OPEN):

  • Do not chase at $197: the stock is +56% from the May 27 close of $126.41 and above the 50-day average of $167.61 with the three death knells flashing.
  • Tranche 1 (1% of portfolio): at or below ~$170 (50-day average; ≈6.6x EV/FY27 revenue).
  • Tranche 2 (up to 2% more): at or below ~$140 (the May–June base the Schwab Network technician called a six-month double bottom; ≈5.4x EV/FY27 revenue).
  • Maximum 3% of portfolio before Investor Day; expand to 5% only if the Investor Day targets imply organic net new ARR growth ≥10% for FY27 and non-seat ACV ≥35% of new.
  • Liquidity: at ~$800m/day, a 5% position in a personal portfolio is a same-day trade.

Exit plan. Thesis-based, not price-based: any kill criterion in the scorecard; or the Investor Day sets FY28–30 growth ≤15% with no non-seat acceleration; or a second sales-leadership departure. Macro override: if the 10-year holds above 5% and the Fed signals a second hike in October, cut the starter regardless of the company, because the multiple will not hold. On the owner's known weakness: if the position works and runs to $300, the only question is whether the reason for buying (second-curve inflection) is still true, not whether the gain feels large.

Verdict. Decent business, wrong regime, no variant perception yet: watchlist with a pre-committed accumulation plan, not a buy at $197.

S9 Watchlist & monitoring

  • 2026-10-06 Investor Day: FY28–30 targets vs the 17% FY27 guide; non-seat mix; capex.
  • Late November: Q1 FY27 call (Q1 net new ARR guided to 15% of the FY27 total).
  • Mid-November: 13F Q3-26 for Jericho (add or exit), Fidelity/T. Rowe (return or not).
  • Fed 2026-10 meeting and the 10-year at 5%: macro kill switch.
  • Leading indicators between quarters: LinkUp job postings (monthly), Zscaler trust portal incidents (outage frequency), Gartner Peer Insights review velocity, Netskope's quarterly NRR and net new ARR growth as the competitive thermometer.
  • Filing alerts: 8-K Item 5.02 (any further sales-leadership change), Form 4 open-market buys (none in 12 months; one would be a signal), any S-3.
  • Litigation: class-action docket and the AI-feature patent suit.

Corrections to the 2026-09-19 files made this session

  • relationships.yaml: Raj Judge is a former officer and director (8-K 2026-04-16); his node roles and a dated event edge now say so. Yesterday's edge marking him current was left in place with its as_of date because it was true on that date.
  • Customer count: 10-K says ~11,000 at 2026-07-31; yesterday's ">9,400" was the FY25 figure read from a Q3 call.
  • Shelf/ATM status, which could not be confirmed yesterday, is now confirmed absent from the full EDGAR index.

This is process support, not investment advice.

Full notes, 19 Sep 2026 6 min read

Zscaler (ZS) — research dossier, 2026-09-19

Drafted by Claude in a supervised session, no owner input. Nothing here is human-verified. Retrieved sources are named per item. Items marked [background] come from Claude's prior knowledge, not from a document retrieved this session.

Access limits this session

  • FMP connector: only profile-symbol and peers allowed on the current plan. Statements, insiders, 13F, analyst, news, filings all denied.
  • sec.gov, efts.sec.gov, ir.zscaler.com, stockanalysis.com: blocked by the session's network egress policy.
  • Used instead: Bigdata.com (Edgar filings, Quartr transcripts, premium news, WSRA fund holdings, LinkUp jobs), Alpha Vantage (Q4 FY26 transcript, insider transactions, institutional summary, company overview), open web search for Glassdoor, Gartner Peer Insights, short interest.
  • Zscaler is not a subject in ledger/subjects.yaml and no podcast row mentions it. The board contributes nothing yet.

Snapshot (price data as of 2026-09-18 close, FMP via Bigdata.com)

ItemValue
Price / market cap$197.31 / $31.9bn
52-week range$114.63 – $336.99
1y / 3m / YTD-33% / +59% / -12%
Cash & ST investments (2026-07-31)$3.47bn
Debt$1.725bn 0% converts due 2028-07-15 (carrying $1.70bn)
Shares outstanding / float (Alpha Vantage)163.1m / 106.5m
Insiders / institutions (Alpha Vantage)34.7% / 53.2%
Short interest (FINRA 2026-08-31 via Equibles search result)6.35m shares, ~5.9–6.2% of float, 2.9 days to cover
Analyst consensus (FMP)42 buy, 10 hold, 1 sell; PT median $200 (low 150, high 298 Bernstein)
Next eventsInvestor Day 2026-10-06 NYC; Q1 FY27 call 2026-11-24

FY2026 results (10-K filed 2026-09-03; Q4 call 2026-09-03)

MetricFY26Note
Revenue$3,352.5m, +25%ex-Red Canary $3.2bn, +20%; Red Canary $144m
ARR$3,771m, +25%net new ARR ex-RC $232m in Q4, +17% (7% FY25, 10% 1H26)
GAAP gross margin77% flatnon-GAAP 80.3%
GAAP net loss-$63.2mFY25 -$41.5m, FY24 -$57.7m; accumulated deficit $1,252.7m
Non-GAAP op margin22.9%Q4 24.3% record
Free cash flow$779m, 23%down from 27%; capex $277m (8% of rev) + $73m capitalised software
Stock-based comp$821.9m24.5% of revenue; unrecognised $1,806.5m over 2.5 yrs
Weighted shares (GAAP)160.2m154.4m FY25, 149.6m FY24 (+3.5%/yr); FY27 non-GAAP guide 173m diluted
RPO$7,365m, +27%45% current
NRR115% every quarter of FY26CFO, Q4 call
Customers>9,400 (Q3 call); 785 >$1m ARR; 4,182 >$100kFortune 500 penetration 45%→50% (Q4 call)
GeographyAmericas 57% (+31%), EMEA 28% (+16%), APJ 15% (+23%)47% international; ~60% of employees outside US, India significant
Goodwill$1,218mfrom $418m; four FY26 acquisitions
DSO (FMP ratios, TTM)~125 daysannual-in-advance billing model

FY27 guidance (Q4 call): revenue $3.908–3.938bn (+16.6–17.5%), ARR $4.396–4.426bn, non-GAAP op margin ~23.7%, FCF margin 23–23.5%, non-GAAP EPS $4.86–4.90, capex ex-software low teens % of revenue (memory prices). Q1 FY27 revenue $935–939m. Restructuring 3% of staff, $30–33m.

Revenue history [background]: FY24 ≈ $2.17bn, FY25 ≈ $2.67bn. Not retrieved this session; consistent with the +25% FY26 growth in the 10-K.

Pipeline scorecard

Lives in scorecard.md next to this file and is updated there; this dossier is a dated snapshot.

S0 detail

  1. Distribution lock-in = 1. Installed base >9,400 customers, RPO $7.4bn, ~160 exchanges/200+ data centres peered to major providers (10-K), channel via GSIs and telcos.
  2. Unscrapeable data = 1. Inline telemetry, "over 750 billion transaction logs per day" (Chaudhry, Q4 call); physically generated, not on the internet.
  3. Liability absorption = 0. Zscaler does not take breach liability onto its balance sheet.
  4. Bottleneck = 1 (judgement). Global peered exchange footprint plus government authorisations; the last point is [background] (FedRAMP) and unverified this session.

Surplus: NRR 115% four quarters running, Z-Flex ARR uplift ~30%, gross margin stable at 80% while op margin expands. Third-party complaints of 35%+ renewal increases (dope.security, a competitor's blog) cut both ways: pricing power exists, and it is resented. Falsifier: a hyperscaler or Microsoft bundling SSE into an identity/endpoint suite at zero marginal price, which the 10-K itself flags ("AI technologies may lower barriers to entry ... including large AI providers and hyperscalers"). Chaudhry says Microsoft is "not a meaningful factor" (self-reported, June 2026 conference).

S3: the three biggest risks management itself discloses (10-K FY26)

  1. Channel concentration and non-exclusive resellers: top five channel partners 31% of revenue and rising; one partner 13% of receivables.
  2. Key-person and sales-leadership turnover: "highly dependent on ... Jay Chaudhry"; "two sales leaders recently departed."
  3. Competition intensified by AI lowering entry barriers, including hyperscalers; larger vendors can sell "at zero or negative margins" and bundle.

Also: 2028 converts ($1.725bn, intends cash settlement), international/India exposure, sustained GAAP losses. Red flags checked: no auditor change (PwC), no restatement or going-concern language seen, one CAM (subscription revenue recognition), acquired entities excluded from ICFR audit (0.1% of assets, 4.4% of revenue). SEC comment letters: could not confirm; sec.gov blocked.

S4: value chain in words

Suppliers: memory/storage/processor vendors and colocation (unnamed; component pricing is the disclosed pressure) → Zscaler Zero Trust Exchange (ZIA, ZPA, Zero Trust Branch/Cloud, data security, Security for AI, Agentic SOC with Anthropic/OpenAI models) → channel (GSIs, telcos, VARs, Carahsoft, cloud marketplaces; 31% via top five) → enterprise and federal customers (federal = high single digits of new ACV). Pricing power sits at renewal and in Z-Flex commitments. Competitors named by Morningstar and management: Palo Alto Networks, CrowdStrike, Fortinet, Microsoft; Netskope, Cloudflare, Cisco are [background].

S5: who is on the other side

Insiders: Chaudhry 16.8% plus an 18.2% trustee line (Ajay Mangal; proxy footnote not retrieved). All officers and directors 17.2%. Twelve months of Form 4s show only RSU tax-withholding sales and small 10b5-1 lots; no open-market buys. 13F Q2-26 (WSRA via Bigdata.com): BlackRock +15%, Millennium +237%, Citadel +97%, Jericho Capital and Norges Bank new; T. Rowe cut 63%, Artisan and Westfield exited. 747 holders, 51% institutional per that dataset (Alpha Vantage says 58%). Short interest ~6% of float.

S6: outside-in evidence

  • Glassdoor 3.5/5 on 2,550 reviews, down 2% in 12 months; 63% recommend (web search summary).
  • G2 4.7/5; Gartner Peer Insights complaints: 10–20% throughput drop, certificate breakage with Docker/npm, ZPA latency, renewal sticker shock.
  • Job postings (LinkUp via Bigdata.com): 275 average active, +27% y/y trailing 12 months, despite the 3% cut.
  • Oppenheimer (2026-08-19): partner checks "mixed ... broadly tracking in line/better versus partner plans."
  • Company-hosted testimonials (Edward Jones, June 2026) are not independent evidence.

S7: draft thesis (not the owner's, not scored)

Bull: organic net new ARR accelerated 7%→10%→17% through FY26 while guidance embeds ~4% net new ARR growth for FY27; if Security for AI and Agentic SOC land, 17% is a floor and the stock re-rates from ~7.7x EV/FY27 revenue. Bear: the acceleration is Q4 renewal seasonality plus Z-Flex pull-forward; new logos have lagged (CFO, Q3 call); Red Canary churn and capex inflation compress FCF; SBC absorbs the whole FCF. Kill criteria: organic net new ARR growth below 10% in any two consecutive quarters; NRR below 112%; Q1 FY27 capex guide above low teens; another senior sales departure before 1H FY27 ends.

S8: valuation frame (owner to decide)

EV ≈ $30.1bn (market cap $31.9bn less $1.77bn net cash). On FY27 guide midpoints: EV/revenue 7.7x, EV/FCF ~33x, non-GAAP P/E ~40x. SBC-adjusted FCF ≈ zero. Scenario sketch: bear FY29 revenue $5.0bn at 5x = $25bn; base $5.6bn at 8x = $45bn; bull $6.2bn at 11x = $68bn. These are illustrative, not a model.

S9: monitoring

Investor Day 2026-10-06; Q1 FY27 2026-11-24; Agentic SOC uptake; geo sales-leader onboarding; Red Canary renewals; memory pricing/capex; 13F Q3-26 for Jericho/T. Rowe; any filed (not "investigated") securities suit; 2028 convert conversion conditions.

Connection map seed

See relationships.yaml: 45 nodes, 52 sourced edges (officers, directors, holders, auditor, counsel, six acquisitions, channel concentration, partners, competitors, converts, litigation watch).

This is process support, not investment advice.