GoDaddy

IThe business

GoDaddy sells a small business its name and its front door. A customer registers a domain, builds a website, sets up email and takes card payments, and pays by the year or longer. It has 20.5 million customers and 82.4 million domains under management, and 94% of its customers bought a domain from it.³⁴ Most of them are very small: the 10-K says most of its largest group, the Independents, have fewer than five employees.³

In the twelve months to September 2026 revenue was $5.18B C (calculated), up 6.0% in the third quarter alone.³⁴⁵ Two segments make it up. Applications and Commerce (A&C) is the software side: website building, commerce and payments, and third-party email. It brought in $2.04B C (calculated), 39% of revenue, and grew 9.0% in the third quarter. Core Platform is the older side: domain registrations and renewals, the aftermarket for used domains, hosting and security. It brought in $3.15B C (calculated), 61% of revenue, and grew 4.2%.³⁴⁵ Domains alone were $1.82B of the $3.91B in the first nine months of 2026, 47% C (calculated).⁴

Revenue is slow money. A customer pays up front, GoDaddy books the cash as deferred revenue, and recognizes it over a term that runs from a month to ten years. Deferred revenue was $3.53B at 30 September, and $1.77B of it will turn into revenue in 2027.⁴ GoDaddy also pays the domain registry up front. It had $791M of those prepaid fees on the balance sheet C (calculated).⁴ The difference, about $2.74B C (calculated), is cash that customers have put in ahead of the work, and it grows when bookings grow. I come back to it in the valuation, because it is part of cash flow that disappears if growth stops.

Costs split three ways. Cost of revenue is 36.5% of revenue C (calculated), and the 10-Q says it is substantially domain registration fees, third-party productivity software, commissions and payment processing.⁴ The wholesale price of a .com domain is set by Verisign: $10.26 now, and $10.97 from 1 November 2026, up 6.9% C (calculated).²⁴ Verisign earns a 67.6% operating margin C (calculated) on that fee, and GoDaddy earns 25.7% C (calculated) on what it sells above it.¹⁹²⁴²⁵ The people cost is the rest. GoDaddy had 5,845 employees at the end of 2025, 2,414 of them in technology and 2,237 in customer care, plus about 3,400 contract support staff, mostly in India and the Philippines.³

About a third of revenue is outside the US. International revenue was $1.28B in the first nine months of 2026, up 5.9%, but only 3.1% in the third quarter.⁴

Revenue grew from $1.85B in 2016 to $4.95B in 2025, 11.6% a year C (calculated), and that includes a $1.88B acquisition year in 2017.¹⁸¹⁹ Since 2022 the pace has been 4% to 8% a year: 7.2%, 4.0%, 7.5% and 8.3% C (calculated).¹⁹ Operating income went from $50M to $1.13B, a margin of 2.7% to 22.8% C (calculated), and 26.0% in the first nine months of 2026.⁴¹⁹

Under a third of the recent margin gain is accounting. In the first nine months of 2026 operating margin rose 4.0 points, from 22.0% to 26.0%, but depreciation and amortization fell from 2.4% to 1.3% of revenue as acquired intangibles reached the end of their lives. That is 1.2 points C (calculated). The cash measure, normalized EBITDA, rose 1.8 points, from 31.4% to 33.2% C (calculated).⁴⁵ The cash gain came from marketing and general costs falling 4% and technology spending staying flat while revenue rose 6%.⁴

Customers, and what each pays

GoDaddy has about the same number of customers as it did three years ago. The count was 21.0 million at the end of 2023, 20.4 million at the end of 2025 and 20.5 million at 30 September.³⁴¹¹¹³ Revenue per customer, which the company calls ARPU, rose from $203 to $242 over those two years and to $253 in the last twelve months R (reported).³⁵¹¹ So the growth since 2023 comes from price and mix. The company's own words for the third quarter agree: bookings rose on domain registration and aftermarket sales and on pricing actions in email and productivity, partly offset by lower unit volume as customers renewed email accounts with fewer seats.⁴

Domains under management tell a similar story. They were 83.9 million at the end of 2022, 80.8 million at the end of 2025 and 82.4 million at 30 September, up 2.0% in nine months.³⁴¹¹ Verisign's .com and .net base, a different count, grew 6.1% in the year to September.²⁴ I do not read much into the gap, because the two counts differ. The 10-K gives GoDaddy's share of the world's 387 million domains as 21% at the end of 2025, and I found no earlier figure on the same basis.³

This matters because bookings lead revenue. Bookings grew 5.7% in the third quarter and 4.7% in the nine months, below the 6.0% and 6.2% for revenue.⁴ Revenue will keep rising for a while on deferred revenue already collected, but a gap that stays negative would pull it down. The company's own full-year outlook is 6% revenue growth at the midpoint, toward the high end of $5.215B to $5.255B.⁵

What Airo is, and what it is not yet

Airo is GoDaddy's name for its AI-built website, branding and commerce product, which the company now calls an agentic operating system for small businesses. Its annualized bookings run rate was $10 million in the first quarter, $50 million in the second and over $100 million in the third.⁵⁸²⁸ At $100 million that is 1.9% of my $5.18B revenue figure C (calculated). On the October call the CEO said Airo had begun to add to bookings growth, and the CFO said more than 70% of new Airo customers have at least two paid products against more than 50% across the company; about two-thirds of Airo bookings are subscriptions and one-third credit packs.²⁸ The CFO also said gross margin on it would deteriorate a little as it grows.²⁸

What the company does not give is Airo's customer count, its retention, or what share of Airo's bookings comes from customers who moved off an older website plan. On the call the CFO said many existing presence customers moved to annual Airo plans.²⁸ If most of the $100 million is a migration of existing revenue, it adds little. If most is new, it adds a lot. I cannot tell from the filings which it is, and I treat that as the main thing I cannot judge on this page.

Two other efforts matter for the AI story. GoDaddy has opened its domain, hosting, email and commerce functions to software agents through a Developer Platform of APIs, and it runs an Agent Name Service, an open registry meant to let software agents find and verify one another.³⁵⁸ It has not said what revenue either brings. It holds an Investor Night on 1 December 2026 to refresh some financial targets.⁵

IIThe moat

A moat is whatever stops a customer from leaving or a rival from taking the next one. GoDaddy's is made of habit and convenience more than of technology. I look at it source by source, and at what is widening and narrowing it.

Notice what is missing. Nothing on the list stops a customer from buying a domain somewhere else for a lower price. The registry sets the wholesale fee for .com and charges every registrar the same, so GoDaddy's edge is in what it sells around the domain and in the 85% of customers who do not bother to move.³²⁴ That is a real edge, and a narrower one than the market share suggests.

Evidence it is holding or widening

A&C revenue has grown 9% to 11% a year for three years and its segment margin rose to 46.3% from 44.7% in nine months C (calculated).⁴⁵ Technology spending was flat while revenue rose 6%, which suggests the platform is not getting more expensive to run.⁴ Customers spending over $500 a year reached 2.0 million.³ The domain business grew 6.8% in the third quarter.⁴ And the retention figure for long-standing customers, about 90%, has stayed there.³

Evidence it is narrowing

Customers have stopped growing, as the last section showed. Core revenue outside domains fell 3.9% in the third quarter.⁴ On the October call the CEO said the industry is discounting more sharply on products like domains, and a Morgan Stanley analyst asked how GoDaddy will handle the coming Verisign price increase.²⁸ The 10-K lists Cloudflare, Namecheap, Hostinger, Squarespace, Wix, Google, Amazon and Microsoft among competitors, and says AI will bring more of them.³ The next round of ICANN's new top-level domain program, expected to open in 2026, adds supply, and GoDaddy paid $17.5M of application fees in the first nine months to take part.³⁴

The CEO also said that Google's move to AI search has so far had a small effect on GoDaddy's traffic.²⁸ A small effect so far is a fact about one year. It says little about ten.

The test for the moat is simple. If customers keep coming to GoDaddy to name an idea and then ask an agent to build the site, the moat is intact and the agent is a feature. If the agent registers the domain and builds the site from somewhere else, GoDaddy is left with a renewal base that ages. Bookings, customer count and domain count are the three numbers that show which is happening, and I track them in section VII.

IIIThe competitive advantage

The moat says what protects GoDaddy. The advantage asks whether it earns more than the people around it. I compare it with the three companies that matter most to its economics: Verisign, which sells it the .com, Wix, which competes for the website, and Shopify, which competes for the store.

Verisign earns 67.6% on its revenue and turns 61% into operating cash. It owns the one thing every .com buyer must pass through, and GoDaddy is one of its customers.²⁴²⁵ GoDaddy earns 25.7% and turns 34% into operating cash.³⁴ The comparison is not unfair to GoDaddy, because a registrar does work a registry does not. It does show where the pricing power sits. When Verisign raises the fee 6.9%, GoDaddy either passes the rise through or keeps the cost, and the call suggests rivals are discounting.²⁴²⁸

Wix is the more direct rival for the website. It makes about 0.1% operating margin on $1.99B of revenue, with $237M of stock pay, so GoDaddy is the more profitable of the two even before Airo.²⁶ Shopify is much larger in commerce and earns a 13.9% operating margin C (calculated) on $13.3B.²⁷

Where GoDaddy wins

It wins where the customer has no technical staff and wants one bill. The bundle of domain, site, email and payments, a phone line that answers, and a brand known to most small-business owners, is hard to match for a business with five employees. It wins on repeat revenue too: 89% of 2025 revenue came from existing customers.³ And it wins on the cost of selling. Marketing spend was 7.0% of revenue in nine months, down from 7.8%, while bookings grew.⁴

Where it loses

It loses on price to registrars that only sell domains and hosting, and it loses on features to builders that put all their work into one product. It does not own the .com. It has no technology edge in AI that I can verify: its own filings list Google, Amazon and Microsoft among competitors, and each of them can put an agent in front of a small-business owner.³ The honest summary is that GoDaddy is a well-run distribution business that earns decent margins on products it does not make, in a market where the people who make the products are moving closer to the customer.

IVManagement

I judge management on four things: how they are paid, what that pays them to do, how much of their own money is in the stock, and what they did with the company's cash over a full cycle.

Mr. Bhutani has run the company through the whole stretch I can measure: from September 2019, revenue went from $3.0B to $5.2B and operating income from $0.20B to $1.33B C (calculated).⁴¹⁰¹⁹ The CFO came from the audit side of the industry, which fits a company that talks about discipline. The team has been stable. None of the four named executives has changed since March 2024.¹⁰

Pay

The CEO's 2025 pay was $23.0M: a $1.0M salary, a $1.1M cash bonus and $20.9M of stock awards.¹⁰ So 91% of it was stock C (calculated). Half of each stock award is performance shares. They pay by how GoDaddy's three-year total shareholder return ranks among the Nasdaq Internet Index companies: 50% of target at the 25th percentile, 100% at the 50th, 200% at the 85th, and nothing below the 25th. The other half is restricted shares that vest over three years.¹⁰ The 2023 to 2025 performance shares paid 129.35% after GoDaddy ranked at the 60.3 percentile, and Mr. Bhutani received 84,450 shares worth $7.5M on 3 March 2026, when the stock was $89.23.¹⁰ His grant for 2026 is $16M, down from $20M.¹⁰

The annual bonus is 80% corporate and 20% personal. For 2025 the corporate part was half total bookings and half normalized EBITDA. Bookings landed at $5,400M against a target of $5,370M and paid 100%; EBITDA of $1,586M against $1,546M paid 125%; the corporate score was 113%.¹⁰ The committee dropped revenue and unlevered free cash flow from the bonus in 2025.¹⁰

Here is what that pays them to do. Normalized EBITDA is calculated before stock pay, so stock pay costs the bonus nothing. Bookings can be raised by price. Relative return against an internet index pays at target for the median performer, and the proxy describes no absolute-return floor.¹⁰ Nothing in the pay ties directly to cash per share or to customers. That is not unusual, and I would rather see cash per share in there. It is also why I look at what they did with cash and what they own.

Their own money

The CEO held 500,053 shares after his 1 October 2026 sale, worth $58.6M at $117.24 C (calculated), against a guideline of six times salary, or $6M.¹⁰²⁰ That is real money even for a person paid $23M a year. Against it I count what I found in 125 Forms 4 filed since October 2024. Officers and directors sold $53.6M of stock from January 2025 to October 2026 and bought none on the open market C (calculated). Of the $53.6M, $31.9M were automatic sales to cover tax on vesting, $13.3M were under pre-arranged 10b5-1 plans and $8.4M were other sales. The CEO's share was $26.0M, of which $18.4M was tax cover and $7.6M plan sales.²⁰

Most of that is the normal exit of stock pay, and sales by plan or to pay tax say little. The absence of any purchase says something. The stock fell from $214 to $75 in about seventeen months, and in that time the company bought back about $2.4B of stock and not one executive or director added a share with their own cash C (calculated).¹⁴¹⁹²⁰

Capital allocation over a full cycle

The record is a good one on the whole. The 2020 settlement saved cash. And since 2023 the cash goes to repurchases. Since 2019 GoDaddy has spent $7.67B buying its own stock C (calculated), which is 53% of today's $14.4B equity value, and the share count is 30% lower than at the 2018 peak.⁴¹⁷¹⁹ I would add one caution. The spending tracks cash flow and ignores the price. In 2025 the company spent $1.6B, with about $600M of it in one quarter at roughly $148 a share; in 2026 it spent $1.3B at about $90.⁴¹⁹ A buyback is only as good as the price paid, and the shares bought in the third quarter of 2025 cost 26% more than today's price C (calculated).

VFinancials

This section has the ten years, then the question of what the reported numbers are worth. GoDaddy's accounts have noise in them: a tax benefit that made 2023 look like a $1.37B profit, a charge that made 2020 a loss, and a tax shield that makes cash flow look better than it will.

Stock pay is the line to watch. It rose from 3.1% of revenue in 2016 to 7.0% in 2023 and has eased to 5.6% in the last twelve months, with nine-month stock pay down 12% to $212.8M.⁴¹⁹ GoDaddy would otherwise have paid that in cash, so every cash figure below subtracts it.

From reported profit to what an owner can take out

Four numbers describe the same twelve months, and they run from $1.74B to $0.96B. Each one answers a different question.

I use $1,225M, which is $9.97 a share on 122.9M diluted shares C (calculated), or 8.5% of the $117.24 price. The gap between $1,744M and $1,225M is stock pay and tax. Neither is a reason to think the company is worse than it is. Both are things a buyer of the whole business would have to fund.

The tax part needs explaining. GoDaddy paid $16.5M of cash income tax in 2025 and $19.3M in the first nine months of 2026, against income tax charges of $145.0M and $215.0M.³⁴ The difference comes from deferred tax assets, mainly past losses and credits, with $2.4B of federal losses and credits at the end of 2025.³ Those assets fell from $1,053M at the end of 2025 to $846M at 30 September, $207M in nine months, or about $275M a year C (calculated).⁴ At that pace they last about three years. After that, cash tax moves toward the 23% rate in the accounts, which on today's pre-tax income is roughly $280M a year E (estimate).⁴ I do not treat the shield as permanent. I tax every year in the model at 23% and add the shield back as a one-off asset worth about $4.60 a share.

The other thing in cash flow is customers' money. Working capital added $184M to operating cash flow in the last twelve months C (calculated), mostly deferred revenue. That inflow is a function of growth. At zero growth it would be close to nothing, and in a shrinking business it would turn negative.³⁴ In the model it is 53% of each year's change in revenue, and if I strip it out of today's cash flow, owner cash is $1,041M C (calculated).

The balance sheet

GoDaddy has $3.81B of debt and $1.12B of cash, so net debt is $2.69B, 1.6 times normalized EBITDA of $1.73B C (calculated).⁴⁵ Interest was $112M net in the last twelve months, covered 12 times by operating income C (calculated).³⁴ The debt is a mix: two term loans, at effective rates of 6.1% and 5.7%, and two bond issues, one due December 2027 at an effective 5.5% and one due March 2029 at 3.7%.⁴ About $1.9B of the floating-rate debt is swapped to fixed.⁴ The $1.2B revolver is undrawn, raised from $1.0B in July.⁴³⁰

Two items in the balance sheet behave like debt without being debt. The first is equity. It is a deficit of $139M, because GoDaddy has bought back more than it has earned in book terms, and goodwill and intangibles of $4.6B sit against it.⁴ A deficit is harmless while cash flows cover the debt. It also means the equity has no book cushion. The second is a patent case. A jury found in November 2025 that GoDaddy infringed two Express Mobile patents, the court kept the verdict in May 2026, and GoDaddy has appealed. It has recorded no liability, and the 10-Q puts the reasonably possible loss at zero to $170M before interest.⁴ In the model I deduct $85M in the base case, $200M in the bear and nothing in the bull E (estimate).

A securities class action was also filed on 21 August 2026 on behalf of those who bought between 3 September 2025 and 24 February 2026. It alleges false statements about GoDaddy's customer acquisition strategy and promotional pricing. The company says it has strong defenses, and I put no figure on it.⁴ I note it because the allegation matches the debate on this page: how much of recent growth came from price and promotion.

VIValuation

This is a ten-year discounted cash flow, built from the two segments down, on cash an owner can take out. I start with what could go wrong, because the number matters less than knowing how it fails.

How this goes badly wrong

There are three ways I can see this being a poor purchase, and they are not independent.

The first is that AI changes where small businesses start. If a business owner asks an assistant to find a name, build a site and take payments, and the assistant does it through another registrar or builder, GoDaddy loses the front door. Customers have already stopped growing, and domain pricing is being cut across the industry.⁴²⁸ This is the path that gets me to my bear case: revenue shrinking about 2% a year from year six, and margins falling from 33% to 30% E (estimate).

The second is that the answer to the first, Airo, costs more than it earns. The CFO said gross margin will see a little deterioration as Airo grows, and I cannot see what a customer pays after moving from a plan to credit packs.²⁸ A&C, 39% of revenue, earns a 46% segment margin today.⁴ If Airo drags that toward 42% while growth slows, the margin gains of the last two years reverse.

The third is the financial layer. The tax shield ends in about three years and cash taxes rise by roughly $280M a year E (estimate).⁴ $600M of notes fall due in December 2027 and $2.2B of loans in 2029.⁴ The company is spending $1.3B a year on its own shares, and if the business shrinks, each of those dollars buys less than it cost. A $170M patent judgment would not break it. Several of these together would hurt.

Working backward, the business is fine as long as customers do not fall further, Airo does not cut the margin it is supposed to protect, and management does not pay a high price for stock it will later want back. Everything in the cases below turns on those three.

What the price implies

At $117.24 an owner pays 9.9 times EBITDA and gets an 8.5% yield on cash after stock pay and normal tax C (calculated). That is high next to a Treasury at 4.94%, and it is the yield of a business the market thinks will shrink.

Over ten years GoDaddy returned 3.3 times and the index 3.6 times C (calculated).¹²³ The share peaked at $214.35 in January 2025, fell to $74.99 in June 2026 and closed at $117.24. It fell 14% on the day after the fourth-quarter report in February, 17% the day after the second-quarter report in July, and rose 11% the day after the third-quarter report on 29 October.¹⁵⁸⁹

The price needs owner cash flow to stay flat for ten years. At a 9.5% rate, $1.31B a year of flat cash flow for ten years and 3% growth after that adds up to today's enterprise value, once the tax shield is counted and the patent case deducted. If I strip out the prepayments, which only exist while the business grows, the price needs about 2% a year.

The discount rate

I use 9.5%. I build it the way the other pages do: the 4.94% ten-year Treasury yield at the close on 30 October 2026 plus a beta times Damodaran's 4.20% implied equity risk premium of 1 October 2026. GoDaddy's ten-year monthly beta against the S&P 500 is 0.97, the five-year 0.94 and the three-year 1.26. They give 9.0%, 8.9% and 10.2% C (calculated).^1212223 The ten-year window reaches back to the years of the old ownership structure, and the three-year window is mostly one drawdown on one idea. I take a beta of 1.08, which gives 9.5% E (estimate). This model follows the site's usual practice of discounting the whole firm's cash flow at that rate and then subtracting net debt, which treats the debt as costing 9.5% when it costs 3.7% to 6.1% before tax. That is cautious. About 55% of my enterprise value is the value after year ten, so one point on the rate moves the base by 15% to 20%.

Three cases

I build each case from the two segments: revenue growth for A&C and Core, a segment margin for each, corporate overhead, stock pay, tax at 23% on profit after stock pay, capital spending of 0.8% of revenue against depreciation of 1.0%, and customer prepayments at 53% of each year's change in revenue, which turn negative if revenue falls. Years are the twelve months to September of each year, 2027 to 2036, valued at 30 September 2026, with the net debt and the share count of that date.

Bear, bull and stress

The bear case is worth $56, 52% below the price C (calculated). It is a business whose customers and prices both slip, whose margin falls three and a half points, and whose cash flow falls 4% a year through year ten. It still has a value: the cash it produces in the first five years is worth about $4.2B, against $2.7B of net debt, so it is a shrinking cash producer that still has value E (estimate). I put no probability on it.

The bull case is $223, 90% above the price C (calculated). It needs A&C to grow 12% for two years and still 4% in year ten while Core grows 3% to 4.5%, and it needs margins to rise to 39%. It is not a stretch for the first three years, since A&C grew 11% in the second quarter. It is a stretch for ten.

From my value to the price, one change at a time

Which input would move my $159 down to $117.24? I changed the base one thing at a time.

Read that last row as the price. It describes a company whose revenue grows about 1.5% a year, less than inflation, with no margin gain, a buyback that adds nothing and an owner who is paid the same cash as today for ten years. The price describes a company that has stopped growing. It does not describe one that is failing.

The margin of safety

My rule for a discount to base value is 25% for a business with net cash and no customer above a tenth of revenue, 30% when one cycle or two customers drive results, and 35% when fixed charges such as debt sit ahead of the shares. I use 35% here, because one question, whether AI changes where small businesses start, drives the outcome, and $3.81B of debt with $625M due in 2027 sits ahead of the shares. That puts my margin-of-safety price at $103, 35% under my base value C (calculated). The price of $117.24 is 14% above it, so there is no margin of safety in the current price by my own rule C (calculated).

The margin of safety covers errors in my base inputs. It does not cover the bear case: at $103 the bear value of $56 is still 46% lower. I do not mark the value down again for the AI question or the debt, because both are already in the cash flows. The cushion is for what the model does not hold, such as an adverse ruling on the patent, a larger rise in cash taxes than I assume, or my own error.

In the bear case the price returns 0.4% a year, less than a Treasury E (estimate). In the base it returns 13.4%, and in the bull 18.0%. The size of the gap between bear and base is the problem.²¹

Quality is one question, price another

GoDaddy is a good business in the narrow sense. It needs almost no capital: capital spending was $28M against $5.2B of revenue. 85% of customers stay. It collects cash before it delivers. Margins have risen without loss of revenue. It does not look like a business that must be saved from itself. Its weakness is a question about tomorrow that I cannot yet answer.

The price is a separate matter. At $117.24 the stock is 26% below my base and 14% above my margin-of-safety price. So I read it as a good business at a price that works if my base holds, with no cushion if it does not. Here the business is good and the price is below my base, which is not the same as safe.

The alternative use of the same dollar

The simplest alternative is a ten-year Treasury at 4.94%. My base case returns 13.4% a year on the price, and my bear case returns 0.4%, so GoDaddy beats a Treasury in two cases of three E (estimate).²¹ Among the pages on this site, the closest comparable is ServiceNow, which the market also prices on a fear that AI will cut demand for what it sells, and whose price is 11% below my base value. GoDaddy's price is 26% below mine. Shopify, which sells to many of the same small merchants, is 68% above my base. The three bases are built on different cash flows and are not a ranking.

Adobe, Intuit, Booking Holdings and GoDaddy all sit in one group of stocks that fell on the same idea: that AI agents will go around the software and the booking or domain layers people pay for today. If that idea is wrong, all four may recover together, and if it is right, all four may fall further. Together they are one bet. They are not four independent bets, and a reader who owns more than one is more exposed to that idea than the count suggests.

Simple odds

A base case that needs several things to go right at once is worth less than it looks. Mine needs four, and I give my odds for each E (estimate). Core revenue grows at least 2% a year through year ten: 70%. A&C grows at an average above 6%: 60%. Margins rise 2.5 points despite lower Airo gross margin: 65%. No legal, tax or refinancing event costs more than $400M: 90%. Multiplied, that is 25% C (calculated). The four are not independent, so the true chance is higher, and a base that misses one still keeps most of its value: halving A&C growth takes it to $134. I do not read the 25% as a probability of the base case. I read it as a reason not to call $159 conservative.

Testing the base against history and the AI fear

The bear is $103 below the base, so I checked where the gap comes from. I put each bear input into the base on its own. The bear's A&C path takes the value to $117, its growth after year ten of minus 1% instead of 3% takes it to $119, its Core path to $136 and its segment margins to $136. Stock pay, overhead, the tax shield and the patent case together take it to $151 E (estimate). Two judgments make most of the spread: how long A&C keeps growing, and what cash flow does after year ten. Those are the two things the AI question is about.

Against history the base is not aggressive on growth. It has revenue up 5.2% in the first year and 3.9% a year over ten, against 7.2%, 4.0%, 7.5% and 8.3% in the last four years and about 6% in the company's outlook for 2026 C (calculated).⁵¹⁹ Those years did not bring more customers. There were 21.0 million at the end of 2023 and 20.5 million now, so the growth was revenue per customer, up from $203 to $253, about 8% a year C (calculated).⁴¹³ My base needs about 4% a year from price and mix with customers flat. I do not know how long customers will accept 8%, so I do not read the history as a promise.

On price, the base is worth 13.4 times free cash flow after stock pay C (calculated). The market value over that same figure at each year end since 2016 ran from 13.3 times in 2025 to 33.2 times in 2018, with a middle of 24.6 times C (calculated).¹³¹⁹ Today's price is 9.9 times. Those were years of faster growth and rising customer counts, so the comparison flatters the base. It shows the base does not need a multiple the stock has never had.

For the AI fear I cut the base three ways. In the domain case Core revenue falls 1.5% a year from year six and its margin ends at 32%. In the site case A&C growth fades to zero by year five and its margin ends at 44%. In the third both happen, margins stay flat, growth after year ten is 2%, I count 70% of the tax shield and the patent loss is $150M. Everything else is the base.

The price sits between the one-sided cases and the two-sided one. The largest single judgment is after year ten. The base lets owner cash grow 3% a year from there. At 2.5% the base is $151, at 2% it is $145 and at 0% it is $126 E (estimate). The AI fear is a claim that 3% is too high for a business that sells names and sites. I keep it because revenue per customer has risen faster than inflation for three years and the renewal base is paid in advance, but it is the input I would defend least. The history says the base is reasonable. The fear says it is not safe. Both are on the page and the base stays at $159.

What I cannot judge

Three inputs cannot be known from what I can read. The first is what AI assistants will do with domain registration: choose, bundle, route. GoDaddy's API launch is a sign it takes the question seriously and not evidence of the answer. The second is the economics of Airo: how much of the $100 million is new, what it costs to serve, and whether customers stay. The company has disclosed neither. The third is the effect of the 2026 round of new domain names on pricing and demand, which opens a new supply of names to every registrar. For these I show a range, $56 to $223, because a single number would claim more than I know.

Where my numbers differ from the screen

This page began with a screen for stocks that look cheap. The screen's notes put GoDaddy at about $235 base on one method and $172 on another. I get $159. The first used a 9% rate, kept $229M of deferred tax that has not been paid inside the cash flow, and so counted the shield forever. The second took out the same tax add-back as I do, used 10% and grew owner earnings 5% a year, with no allowance for the patent case; my base grows owner cash flow 4.3% a year over years one to ten. Where my numbers differ I trust mine. I tax every year at 23% and count the shield once, as an asset worth $4.60 a share, which explains most of the gap to the first. The screen's figures are a starting sketch and I have re-derived everything from the filings.

Where this could be in 5 and 10 years

The table gives my value per share at September 2031 and September 2036. Each is the model's value at that date: the cash after it, plus the value after year ten, discounted to that date at 9.5%, less today's $2.69B of net debt, per today's 122.9M shares and in that year's dollars E (estimate).⁴ Owner cash is what the business pays out to owners after stock pay, interest and tax at 23%, with the tax shield added and the patent loss taken off. The return is the yearly rate from $117.24, the same measure as in the earlier table.

For the base to be worth $186 in 2031, revenue has to reach $6.5B, growing 4.6% a year, with a normalized EBITDA margin near 35% and owner cash of $1.6B before interest E (estimate). For $219 in 2036 it needs $7.6B of revenue, a 36% margin and $1.87B of cash, and a buyer who then pays about 16 times that cash, which is the 9.5% rate and 3% growth I use today. The value rises only about 3% a year because the cash goes out along the way: $60 a share in five years and $128 in ten.

The bear is a shrinking business. Revenue is $4.8B and falling in 2036, the margin is near 30%, and the value slips from $45 to $41 while it pays out $74. The bull needs A&C growing 12% for two years and still 4% in year ten, with the margin near 39%. That gives $330. Four numbers drive the spread: A&C growth, growth after year ten, the margin and the 9.5% rate. Capital needed is small, 0.8% of revenue, and moves little.

The most likely way the base is wrong is revenue per customer. With customers flat, my 3.9% a year is all price and mix. If A&C grew half as fast, the 2036 value would be $180, not $219.

These are estimates of value, not forecasts of the share price. The price can sit above or below value for years: this stock went from $214 to $75 in seventeen months.¹

VIIThe case for the price

What most people miss

The company has been buying its stock below today's price. In the nine months to September it retired 14.6 million shares for $1.31B, about $90 each, and through 26 October 15.7 million for $1.4B, a 12% gross cut in the diluted share count this year C (calculated).⁴⁵ At the same cash flow, 12% fewer shares is 14% more cash per share.

Roughly a third of next year's revenue is already paid for: $1.77B of deferred revenue is due to be recognized in 2027, against $5.2B of revenue expected for 2026 C (calculated).⁴⁵ Customers can leave at renewal, so it is a floor for the first year and no more.

The software side earns almost half the segment profit: A&C is 39% of revenue and 47% of the two segments' EBITDA, at a 46.5% margin C (calculated).³⁴ Each year it outgrows Core, the average margin rises.

The bull case

The people who hold it best are the company's managers. CEO Aman Bhutani said on 29 October that the AI transformation is gaining momentum, that Airo's annualized bookings run rate had more than doubled in a quarter to over $100 million, and that GoDaddy is positioning for an agentic world, where the 10-K says identity, naming and trust could matter more.³⁵²⁸ The release reaffirmed a free cash flow target of about $1.8B for 2026 and a normalized EBITDA margin above 33%, and CFO Mark McCaffrey said the company repurchased 5 million shares for $477M in the quarter.⁵²⁸ These are company claims; Investor Night on 1 December will refresh some targets.⁵

Their case, in plain words: small businesses will keep needing a name and a place online, whoever sets it up; GoDaddy holds the customer and the billing; Airo sells more to each one; and the old domain base pays for the buybacks.

I found no named investor thesis on GoDaddy that I could read and quote. I have left analyst price targets out, because I could not open the notes behind them and the lists that carry them are not a primary source.

Against reality

If A&C keeps growing near 9% and fading to 3.5%, Core grows 2% to 3% and margin rises 2.4 points, the cash is worth $159 a share, 35% above the price E (estimate). The filings so far show A&C up 10.5% in nine months, Core up 3.7% and a normalized EBITDA margin up 1.8 points.⁴⁵ The price says the next ten years will not look like the last two.

The market has reasons. Customers fell from 21.0 million to 20.5 million since 2023, and bookings grew 4.7% in nine months, slower than revenue.⁴¹³ The CEO said discounting in domains has sharpened.²⁸ The stock rose 10.7% after the third-quarter report and is still 45% below its January 2025 high C (calculated).¹ The market may be right that growth is gone, and I cannot prove my base is the better guess.

What a 10x would take. A tenfold share price is $1,172. Suppose buybacks shrink the share count 3% a year to 91M by 2036, and net debt is $3B. The equity would be worth $106B and the business $109B E (estimate). At 20 times normalized EBITDA, twice today's 9.9 times, EBITDA would need to be $5.5B, which at a 36% margin is revenue of $15.2B, 2.9 times today's and 11.4% a year for ten years E (estimate). At today's multiple it needs revenue of $30.7B, 5.9 times today's. GoDaddy's revenue grew 7.2%, 4.0%, 7.5% and 8.3% in the last four years, and my bull case has 5.6% a year.⁴¹⁹ The one ten-year window I have for the stock, October 2016 to October 2026, returned 3.3 times, and its best stretch, October 2016 to the January 2025 high, returned 6.0 times in eight years C (calculated).¹ A tenfold in ten years is 26% a year. A tenfold is not plausible for this company, and nothing on this page assumes it.

Where I come down

The third-quarter report moved the price 10.7% and left my values where they were, because the quarter is already in my base, and the two weak spots, customers and bookings growth, did not move me. I would raise my values if bookings grow as fast as revenue for two quarters with the A&C margin at 46% or better; one more point of A&C growth in every year takes the base to $168 E (estimate). I would lower them if customers fall below 20 million or domains under management below 81 million.

VIIIRisks, and what would change my view

The main risk is the one in the price: that AI changes how small businesses go online and GoDaddy is not where they start. It is the risk behind my bear case, and the numbers that would show it first are bookings, customers and domains. The 10-K names the same risk in its risk factors and adds that competition is likely to intensify.³⁴

The other risks are narrower. Verisign can raise the .com fee 7% a year in each of the last four years of its six-year period, and GoDaddy has to pass it through or absorb it.²⁴ Regulation can change the domain rules, and ICANN accredits GoDaddy as a registrar.³ An outage or a security breach would cost customers; the 10-K spends several pages on both.³ The key-person risk is moderate: the CEO has run the company for seven years and holds $58.6M of stock.¹⁰²⁰ Debt is manageable but not trivial: $625M due in 2027 and $2.2B in 2029.⁴ A $170M patent loss is possible and unaccrued, and a securities class action is early.⁴

I found one accounting flag. At the end of 2023 management reported a material weakness in its controls over income-tax accounting, the area where this company has the most noise. It was remediated, controls were effective at the end of 2024 and 2025, and the auditor's opinions on the 2025 statements and controls were clean.³¹² Management also judged disclosure controls effective at 30 September 2026.⁴ The other things that look odd are explainable: the 2023 profit was a tax benefit, the 2020 loss was a settlement, and cash flow is flattered by a tax shield that I have removed.

Last, correlation. GoDaddy, Adobe, Intuit and Booking Holdings share the idea that AI will go around what each sells, so a reader holding more than one holds one bet. A change in how the market treats that idea will move all four, whatever each does.