GoDaddy is the world’s largest domain-name registrar: ~81 million domains under management at the end of 2025, roughly 21% of every domain registered worldwide, serving 20.4 million customers — about a third of revenue from outside the United States. A domain is a small, annually renewing purchase, and it is the top of the funnel: from the name, GoDaddy sells the website, the email, the store, the payments terminal and, increasingly, an AI layer (Airo) that assembles the whole presence from a prompt.

Two segments report the economics. Core Platform — domains, hosting, security — is the annuity: $3,062.1M of FY2025 revenue at a 33.0% segment EBITDA margin. Applications & Commerce (A&C) — websites, email and productivity, commerce and payments — is the growth layer: $1,889.0M at a 45.4% margin, growing in the low-to-mid teens while Core grows in the low single digits. The company’s whole strategy is the mix shift between those two lines: every point of revenue that migrates from Core to A&C carries roughly twelve more points of margin with it.

Customers prepay: bookings were $5,400.0M in FY2025 against $4,951.1M of recognized revenue, and the balance sheet carries $3,510.9M of deferred revenue — cash collected before the service is delivered. Capital needs are minimal (capex was $23.9M in FY2025, under 0.5% of revenue), which is why the model converts so much of its profit into free cash flow — and why what management does with that cash, covered under Management, matters more here than what it builds.

1. The customer base is shrinking. Total customers fell from 21.0 million (2023) to 20.5 million (2024) to 20.4 million (2025), and domains under management slid from 83.6 million to 80.8 million. Growth has come from charging each remaining customer more — ARPU $203 → $220 → $242 — and extracted price always meets its limit eventually. Conceded: the volume line is genuinely negative, and no quarter of customer growth at the old rate is in evidence.
   Bull answer: The shrinkage is substantially a chosen purge — management ended deep discounting, divested low-value lines and let low-intent customers lapse, then said so at the time (Q4 2024 call). The economics kept: revenue from prior-year customers is above 89% of the total, retention of three-year-plus customers is ~90%, and the count stabilized in 2026 — 20,422K at year-end 2025 to ~20.5 million by Q2 2026 — while ARPU kept compounding (+9% to $250 in Q2 2026). A smaller, richer register is the model, not the accident.
2. AI builds the website for free — and agents may skip websites entirely. The entry product (a small business’s first website) is exactly what generative AI commoditizes, and the 10-K itself names the risk: AI-driven changes in how customers use the internet may reduce demand for domains and sites. Contested: this is the live strategic question, and two years of Airo are not proof either way.
   Bull answer: So far the layer is monetizing, not cannibalizing: A&C grew +11% in Q2 2026 with Airo attached, and management reported Airo’s annualized bookings run-rate reached $50 million in Q2 2026, up from $10 million one quarter earlier (Bhutani, Q2 2026 call). The identity layer — the name, the email, the payments account — is what an agent still needs a registrar for, and GoDaddy is positioning its Agent Name Service there. The honest limit: Airo’s revenue is not broken out, so the size of the offset is company-described, not filed.
3. The growth was bought with a promo nobody disclosed. After two years of telling investors deep discounting had ended, GoDaddy ran a $4.99 one-year .com promotion in late 2025 and first described it on the Q4 2025 call — the same call where bookings growth printed +5%, its softest in the set — framing it retrospectively as a success. A securities class action over that disclosure followed (allegations unproven; the stock fell ~14% the day after the print, as reported). Conceded: the disclosure lagged the decision, and it lagged exactly where guidance is vaguest — bookings carries no hard range.
   Bull answer: The damage was to bookings optics and trust, not to the revenue machine: the promo shifted term mix (one-year instead of multi-year terms), which defers bookings without changing revenue recognition — and the revenue guide was then raised twice during 2026 while the promo lapped. Q2 2026 bookings grew +6% against +7% revenue, on the convergence path management guided. The episode is graded where it belongs — in the Management section’s credibility score — not hidden from it.
4. The buyback is a treadmill over a stock-comp leak. Stock-based compensation was $301.9M in the last twelve months — about 18% of free cash flow — and holding the share count down cost $1,633.8M of buybacks in the same period, essentially the entire free cash flow. If the repurchase ever pauses, dilution resumes and the per-share story stalls. Answered on the record: the net shrink is demonstrated, not promised — diluted shares fell 3.2% in FY2025, and the latest quarter’s diluted count (131.0M) is down 7.4% from a year ago.
   Bull answer: Owner earnings on this page are already net of stock compensation — the leak is charged before the bar is set, and the bar still solves near zero growth. The treadmill framing also inverts the alternative: over the last four years more than 95% of free cash flow went to repurchases (company-reported) at prices that averaged well below today’s — the Q2 2026 tranche alone retired 6.6 million shares for $554M, about $84 a share. The remaining authorization is $1,331.6M through 2027.
5. Core is a regulated pass-through with at-cost rivals. Registry operators (Verisign for .com) set wholesale prices politically, renewals reprice against GoDaddy’s highest-in-class renewal pricing, Namecheap is gaining registrations faster relative to its size, and Cloudflare sells domains at cost — registry price plus the ICANN fee, no markup, no upsell. Contested: the pricing umbrella is real, and so is the slow share arithmetic.
   Bull answer: Wholesale increases have historically passed through to retail across the industry, and the Core segment’s margin has held ~33% through the recent cycles while the mix shifted toward aftermarket and services. Cloudflare’s at-cost register is structurally an acquisition channel for its network products, still roughly a tenth of GoDaddy’s annual .com adds; Namecheap competes for the price-sensitive tail GoDaddy has been deliberately shedding. The defense is the installed base: 81 million domains renew where they sit unless a customer does paperwork to leave.

The record in six lines: revenue compounded from $3,815.7M (FY2021) to $5,104.1M (trailing twelve months); the A&C share of it rose from about three-tenths to about four-tenths; operating margin expanded from 10.0% to 24.8% as the mix shifted and restructuring costs fell away; owner earnings — free cash flow after stock compensation — compounded +24.3%/yr across FY2022–25; and the diluted share count fell every year, from 171.1 million to 140.6 million, then to 131.0 million in the latest quarter. Growth here is three engines running at once: modest revenue growth, a mix shift into a higher-margin segment, and a melting share count.

The latest quarter against the same quarter a year ago:

Diluted share figures for FY2021–22 are as reported in those years’ 10-Ks (a dual-class structure was still in place; the Class B effect on the diluted count was under half a percent). The TTM share figure is the Q2 2026 quarter’s diluted count, the latest filed. Stock compensation for FY2023 differs slightly between the cash-flow statement ($296.3M, used throughout) and the segment reconciliation ($294.0M).

Free cash flow (orange) beside stock-based compensation (blue), by quarter: the cash line climbs from ~$150M to ~$440M a quarter while the comp charge stays flat — owner earnings are the gap between them, and the compounding the price is tested against.

Shares outstanding, by quarter: the repurchase is large enough, every year, to shrink the count through the stock-comp leak — from ~168M in late 2021 to ~131M by mid-2026.

Scale and default status. With ~21% of all domains worldwide, GoDaddy is the registrar a new business finds first — the brand is the generic word for the category for its customer class, and the funnel starts there. Switching friction. A domain is cheap and the paperwork to move it is annoying; bundled email, site, and payments raise the friction further, which is what ~85% overall retention and ~90% retention among three-year customers look like from the inside. The attach ladder. The register is the low-margin door into a 45%-margin applications layer — the moat is not the domain, it is the sequence domain → site → email → store → payments, each step sold to a customer already inside. What the moat is not. There is no network effect between customers and no pricing power over the registry: wholesale prices are set upstream, retail renewal pricing is already the highest among majors, and the moat defends volume, not price.

Competition — the short list:

• Namecheap — Real. The clear #2 registrar, adding .com registrations at roughly two-thirds of GoDaddy’s volume from a base a third of its size, at lower first-year and renewal prices.
• Cloudflare — Emerging. Sells domains at cost (registry price plus the ICANN fee, no markup, no upsell), a structural attack on the promo-and-renewal model — but still only about a tenth of GoDaddy’s annual .com adds.
• Wix — Real. Several times GoDaddy’s share of websites in the builder layer where A&C’s growth lives, on ~$2.0B of revenue; its site share growth has recently stalled to roughly flat.
• Squarespace — Real. Design-led sites plus a top-tier registrar book acquired with Google Domains (~10 million domains in 2023) — it now competes on both of GoDaddy’s layers at once.
• Shopify — Limited. A different league in commerce (~$11.6B revenue, +30%/yr); it caps GoDaddy’s commerce ceiling with larger merchants rather than contesting the micro-merchant base GoDaddy serves.

B+ Guidance credibility B+ · Execution A− · Capital allocation A− · Alignment B

Aman Bhutani has run GoDaddy since September 2019 (from Expedia, where he ran the flagship brand and, before that, its technology platform); CFO Mark McCaffrey since June 2021. The pair’s operating record is the strongest part of this page. On guidance — the evidence that matters most — the last eleven delivered quarters show zero revenue misses and zero margin misses: delivery clusters at or just above the top of the guided range, full-year guides are raised inside the year (FY2024 revenue was guided $4.48–4.56B, raised repeatedly, and delivered at $4,573.2M; FY2025 was guided $4.86–4.94B, raised to $4.93–4.95B, delivered at the top), and dated multi-year promises were kept — the ~31% exit-2024 Normalized EBITDA margin arrived at 32%, and the 33%-by-2026 margin target is being met on schedule (33.4% in Q2 2026). Management tells you how the small beats are manufactured — high-value aftermarket sales are excluded from guidance — which is the honest version of sandbagging.

The deduction is for the one episode where disclosure ran the other way. After telling investors on the Q4 2024 call that the company had “ended deep discounting at the top of our funnel,” and on the Q2 2025 call that it had “largely moved beyond the impacts of eliminating deep discounts,” GoDaddy introduced a promotional $4.99 one-year .com price in late 2025 — and described it only on the Q4 2025 call (Feb 24, 2026), in the same breath as the bookings softness it caused: “the shift in term mix, combined with the promotional price reduced up front bookings and near-term revenue.” Asked by Baird’s Vikram Kesavabhotla why the go-to-market change was made, McCaffrey explained the mechanics — bookings hit, revenue timing largely intact — and Bhutani called it “a bit more successful than we thought.” Both may be true; neither was said while the promo was live. A securities class action over that sequence is pending (allegations unproven). Guidance credibility grades the numbers A and the candor C+ — the blend is B+.

Execution (A−). Operating margin expanded from 10.0% (FY2021) to 24.8% (trailing) while the mix shifted toward the 45%-margin segment; the FY2023–25 restructuring program wound down on schedule; AI tooling is being deployed against the cost base with margins guided up, not down. The October 2026 departure of the CTO — reported in trade press, with no successor named — is the open item here, not a grade input yet.

Capital allocation (A−). The simplest record on the page: more than 95% of free cash flow over four years went to buybacks (company-reported); the 2022 authorization ($4.0B) was completed in April 2025 and replaced with $3.0B through 2027 ($1,331.6M remaining at Jun 30, 2026); recent buying has been counter-cyclical — $554M in Q2 2026 at about $84 a share into the post-print weakness. The blemish is the January–April 2025 accelerated repurchases: $767.4M at a $176.02 average, near the all-time high — the same dollars a year later would have retired roughly twice the shares. No dividend has ever been paid, and none is intended; M&A since Poynt (2021, $320M cash plus deferred) has been tuck-ins only.

Alignment (B). Insiders own ~0.9% of the company. Selling is persistent but mechanical — Form 4s across the team are 10b5-1-plan sales and tax-withholding dispositions on vesting dates; no open-market purchases appear in the last two years of filings reviewed. Bhutani still holds ~500,000 shares outright (about $53M at the price on this page) — real exposure, steadily diversified down by plan. Nothing here is a red flag; nothing here is a co-investment signal either.

With credit for buybacks shrinking shares 3.21%/yr, the requirement falls to −2.28%/yr — owner earnings could shrink modestly for a decade and the price would still clear a 10% return.

At $106.70, what’s priced in before that credit is about +1.0%/yr of owner-earnings growth — against delivered compounding of +24.3%/yr across FY2022–25 and a trailing year at +11.4%. The FY2026 guide implies owner earnings near $1.5B (the ~$1.8B free-cash-flow guide, less stock compensation at its ~$0.3B recent pace) — roughly +20% over FY2025’s $1,257.7M in a single year. The price does not ask for the record to continue; it asks for the record to have essentially stopped.

Method, in one line: we work backwards from the price at a 10% required return — the solved growth rate holds ten years, then eases gradually toward 2% over the following forty years, terminal struck at year 100, owner earnings = free cash flow − stock-based compensation, solved on enterprise value of $16,174.7M; no exit multiple. No analyst estimates. All we assume are the required rate of return, and the timeframe of predictable earnings growth.

Dividends: GoDaddy pays no dividend, so there is no yield to add: at the priced-in bar the ~10% total return the model solves for would come entirely from price appreciation and the shrinking count.

The inputs, in one place: base = trailing-twelve-month owner earnings $1,401.0M = free cash flow $1,702.9M (operating cash flow $1,728.8M − capital expenditures $25.9M) − stock-based compensation $301.9M, through Q2 2026; enterprise value = market cap $13,513.3M (126,647,704 filed shares at $106.70) + debt principal $3,816.9M − cash $1,155.5M. Memo: on the FY2025 base instead ($1,257.7M of owner earnings), the flat bar is +2.09%/yr. The credited bar is the multiplication form (1+g)(1−s)−1 at the demonstrated shrink s = 3.21%/yr: diluted weighted-average shares fell from 145,287,000 (FY2024) to 140,621,000 (FY2025), and the basic count fell too (141,250,000 → 138,100,000) — both filed tests pass. Terminal value struck at year 100 is 0.05% of enterprise value.

What you must believe for the price to be merely fair: that owner earnings roughly flatline for a decade (flat bar) in a business whose recurring base renews at ~85–90% and whose guide points the other way — or, on the credited bar, that the buyback stops entirely while owner earnings shrink. The honest bear version is not in the arithmetic; it is in the Kill attempt: that the customer base’s slow shrinkage is the first chapter of the AI story, and the flat bar is the market pricing a melting annuity correctly. The next print grades the first piece of that.

Every tranche outstanding at Jun 30, 2026 (principal):

The maturity ladder is gentle until it isn’t: $624.6M comes due in 2027 (the senior notes plus revolver expiry), only $24.6M in 2028, then $2,210.3M in 2029. Against that stand $1,155.5M of cash and a business whose customers fund it in advance — deferred revenue of $3,510.9M ($2,527.9M current + $983.0M long-term) is the largest liability on the balance sheet and it is settled by delivering services, not by writing checks. Net debt of $2,661.4M is ~1.6× trailing free cash flow. Two honest cautions: the term loans float, so the interest line re-prices with rates — net interest expense was $114.2M in FY2025, covered many times over by segment EBITDA of $1,867.2M; and goodwill plus intangibles are $4,573.8M of the $7,996.6M asset base, the residue of the acquisition era. Book equity of $6.7M is an accounting artifact of repurchasing $6B+ of stock — it says nothing about solvency, but it does mean the balance sheet offers no book-value floor.

Q3 2026 print — Thu Oct 29, 2026, after the close (webcast 5:00 p.m. ET). The checkpoints below are graded PASS / FAIL / MIXED exactly as written when the print lands.

• Revenue: Q3 revenue at or above the $1,315M guided floor — FAIL below $1,300M.
• The FY2026 guide: full-year revenue guide held or raised (floor at or above $5,215M) — FAIL on any cut.
• Decisive — the cash guide: the ~$1.8B FY2026 free-cash-flow guide reaffirmed — FAIL if it is cut or withdrawn.
• Buyback pace: at least $400M repurchased in Q3 and diluted shares at or below 130M — FAIL below $200M of repurchases.
• The growth layer: A&C revenue growth at or above +9% — FAIL below +6%.
• Margin: Normalized EBITDA margin at or above 32% — FAIL below 30%.
• The register: total customers at or above 20.5M with ARPU at or above $250 — FAIL if customers fall below 20.4M.
• The legal overhang: no material charge booked for the promo class action or the Express Mobile patent verdict — FAIL if a charge above $100M is taken.

After the print: the class-action process runs through 2027 (lead-plaintiff deadline Oct 20, 2026); the Dec 2027 senior notes and revolver expiry become a 2027 refinancing story; and the FY2026 print (expected Feb 2027) carries the first full-year test of whether bookings and revenue growth reconverge as guided. The Investor Day frame — 33%+ Normalized EBITDA margin and 6–8% revenue growth — is the standing multi-year yardstick.

• Airo’s revenue is not broken out. The only quantification is an annualized bookings run-rate (~$50M, Q2 2026 call) — bookings, not revenue, and a run-rate, not a period figure. The AI offset in Kill point 2 is therefore company-described, not filed.
• No domain renewal rate is disclosed. The 10-K gives customer retention (~85%; ~90% for three-plus-year customers). Anyone quoting a GoDaddy “domain renewal rate” is quoting a third-party estimate, not the company.
• Segment cash flow does not exist in the filings. Segments report revenue and EBITDA only; corporate overhead ($281.3M in FY2025), stock compensation and interest are unallocated, so segment-level owner earnings cannot be computed from the record.
• Two free-cash-flow totals circulate. The company’s Q2 2026 materials cite trailing free cash flow of $1.73B; the filed cash-flow statement on the standard definition (operating cash flow − capital expenditures) gives $1,702.9M. The ~$27M difference is not reconciled in the materials reviewed; this page uses the filed definition throughout.
• The litigation outcomes are unquantifiable from the record. The promo class action is at the lead-plaintiff stage (allegations unproven, no accrual disclosed in the filings reviewed), and the Express Mobile jury verdict (Nov 2025) is in post-trial motions with damages enhancement sought — neither has a filed number to underwrite.
• The CTO departure is trade-press sourced. No company release or 8-K was located, and no successor has been named in the sources reviewed.

• [T1]Tier 1 — filings: Forms 10-K for FY2021–FY2025 (FY2025 filed Feb 25, 2026), Forms 10-Q through Q2 2026 (filed Jul 31, 2026), and the Q2 2026 earnings presentation — all SEC EDGAR, CIK 1609711. Segment revenue and EBITDA from the segment notes; the debt schedule and balance sheet from the Q2 2026 10-Q (Note 9 and the consolidated balance sheet); share counts from the 10-K EPS notes and the Q2 2026 10-Q cover (126,647,704 shares, Jul 24, 2026); retention, customer, bookings and ARPU figures from the FY2025 10-K MD&A and business sections; the repurchase authorization from Note 4 of the FY2025 10-K and Q2 2026 10-Q ($3.0B through 2027; $1,331.6M remaining at Jun 30, 2026).
• [T1]Model: house reverse DCF — the price is taken as given and the model solves backwards for the owner-earnings growth rate that makes it fair at a 10%/yr required return. Owner earnings are free cash flow minus stock-based compensation, valued against enterprise value. The solved rate holds for ten years, then growth eases exponentially toward 2% over the following forty years (a seven-year half-life) and holds at 2% to year 100, where the terminal value is struck.
• [T2]Tier 2 — transcripts: 12 quarters of earnings calls (Q3 2023–Q2 2026) read in full — Motley Fool, MarketBeat, AlphaSpread and EarningsCall transcripts as published; guidance figures cross-checked against the earnings-release outlooks. The Q3 2026 print date (Thu Oct 29, 2026, after the close) is from the company’s Oct 8, 2026 announcement.
• [T3]Tier 3 — reported, color only: the class-action allegations and the Feb 25, 2026 share-price reaction (law-firm releases via GlobeNewswire/Pomerantz); insider-ownership aggregates (~0.9%) and Form 4 tallies; competitor share and pricing comparisons (Domain Name Wire, Cloudflare, W3Techs compilations); the CTO departure (DomainGang). Nothing in the model or the graded record rests on a Tier 3 item.
• [T3]Price: close Mon Oct 19, 2026, $106.70; 52-week range $71.59–$137.93 (Finnhub/Macrotrends; one aggregator’s table shows a $106.62 close — the $0.08 difference does not move any figure on this page at the precision printed).