Chapter 3

Moat and Durability

GoDaddy clears a real moat bar: a customer retention rate of about 85% held for five straight years, roughly 90% for customers past their third year, more than 89% of revenue coming from customers who were already customers a year earlier, and 21% of the world's registered domains under one roof [1] [2]. But the engine underneath has quietly changed. Since 2023 revenue rose about 16% while the customer count and the domain base each shrank about 3% — the growth is entirely higher spend per customer, not more customers [3]. Whether year-ten cash flow is higher now depends on how far that per-customer figure can travel in an internet being reshaped by AI.

The moat, measured

A moat that matters shows up in retention, and GoDaddy's does. Its blended customer retention rate has sat at roughly 85% in each of the five years through 2025 (about 84% in 2024, held down by deliberate divestitures and product end-of-life), and it climbs to roughly 90% once a customer is past three years with the company. More than 89% of 2025 revenue came from prior-year customers [4].

The single most telling disclosure is the cohort math. GoDaddy acquired about 5.0 million customers in 2017 for $253.2 million of marketing spend; by the end of 2025 that one cohort had generated roughly $3.0 billion of cumulative bookings, at an average annual revenue retention rate above 91% over the seven years [5]. A cohort that returns roughly twelve times its acquisition cost over eight years, while spending more each year, is the definition of a low-churn subscription franchise with real switching costs.

Source: FY2025 Annual Report (Form 10-K), MD&A and Item 1 Business [6] [7]; "over 89%" and "approximately 90%" shown at their disclosed thresholds.

The switching cost is concrete rather than contractual. A customer who leaves does not just cancel a subscription; they move the domain that is their address on the internet, the email tied to it, the website built on GoDaddy's tools, and increasingly the payments and commerce plumbing behind it. About 94% of GoDaddy customers hold a domain with the company, and the domain is the anchor the rest attaches to [8]. Scale reinforces it: roughly 81 million domains under management, about 21% of the roughly 387 million registered worldwide, the largest single position in the industry [9].

Where the growth actually comes from

The moat holds the base. It is not adding to it. Total customers slipped from 21.0 million at the end of 2023 to 20.4 million at the end of 2025, and domains under management fell from 83.6 million to 80.8 million over the same span [10]. What rose was average revenue per user — from $203 to $242, up about 19% in two years, and up from $170 in 2020 [11]. Revenue growth is a pricing-and-mix story on a flat-to-shrinking unit base, not a customer-acquisition story.

Source: derived from reported operating metrics, FY2025 Annual Report (Form 10-K); revenue, ARPU, total customers and domains under management indexed to FY2020 [12].

Two forces drive the rising ARPU, and both are healthier than a simple price hike. The first is bundling and list-price increases across the domain and presence products — domains revenue rose 7.3% in 2025, from $2,152.7 million to $2,310.5 million, even as the domain count edged lower, which is close to pure price [13]. The second is mix: the higher-margin Applications & Commerce segment (email, websites, payments, commerce software) grew from $1,430 million of revenue in 2023 to $1,889 million in 2025, a 32% two-year gain against 8% for Core Platform, and it carries the richer segment margin [14]. Customers are buying more products, not just paying more for the same one — and the company's own data shows multi-product customers retain at higher rates [15].

Source: FY2025 Annual Report (Form 10-K), Results of Operations [16].

This is where the durability question sharpens. Raising ARPU roughly 8–10% a year on a stable base is a fine model for years, and higher lifetime spend on loyal multi-product customers is exactly what a compounder wants. But it is a narrower base to grow from than a rising customer count, and it leans on continued pricing power and attach — the very levers most exposed if the reason a customer needed GoDaddy in the first place gets easier to satisfy elsewhere.

The competitive set and the disintermediation risk

The near-in competition is fragmented and, on the evidence, contained. GoDaddy names a long roster across its markets — Newfold Digital, Namecheap, Tucows, GMO, Cloudflare and Identity Digital in domains and hosting; Wix, Squarespace, Automattic, IONOS and the hyperscalers Google, Amazon and Microsoft across presence and commerce — and describes the market as "highly fragmented and competitive" [17]. No single rival is taking the domain franchise; 21% global share has been stable, and retention has not cracked under a decade of this competition. Fragmentation, held off by scale and switching costs, is a manageable condition, not the threat that halved the stock.

The threat that matters is stated plainly in GoDaddy's own risk factors. "The widespread acceptance of any alternative system, such as mobile applications, AI-powered products and tools or closed networks, could eliminate the need to register a domain name or to establish an online presence," and as reliance on social channels and AI tools grows, "domain names, websites and online stores and marketplaces may become less prominent, and their value may decline" [18]. The company adds that it expects "more competition… from both traditional and non-traditional competitors" as products fill with AI and large language models [19]. This is the bear thesis: if an entrepreneur can stand up a presence by describing it to an AI agent — with no domain to register and no website to build — the attach-and-upsell funnel that lifts ARPU loses its starting point.

Management's answer: Airo and ANS

GoDaddy is not standing still, and it is worth weighing what it has built rather than dismissing it. Two responses stand out.

The first is monetizing AI inside its own funnel. GoDaddy Airo, its AI experience spanning domains, sites, logos, email and marketing, has moved from a generative toolset toward an agentic one, and the company has begun charging for it — Airo Plus is positioned as a direct monetization vehicle, with paywalls inside the experience [20]. This turns the AI shift from pure threat into a potential new price lever on the same base.

The second is more strategically interesting. GoDaddy has launched Agent Name Service (ANS), an open architecture that anchors AI-agent identity to the domain name system — the same DNS infrastructure its domain leadership already sits on. ANS is in production, with registered agents operational since the end of 2025, an open API for developers, and a public transparency log [21]. The logic is to convert the disintermediation risk into an extension: if AI agents need discoverable, verifiable identities to transact with one another, and that identity is anchored to domains, then a world of billions of agents could need more domain-based identity, not less. On the Q4 2025 call, management framed its edge as more than 20 million customers, decades of proprietary behavioral data, and distribution at the top of the funnel — the assets it would use to deploy agentic capabilities at scale [22].

The honest read is that this is a credible, well-positioned response, not yet a proven one. ANS is months old and unmonetized; Airo Plus revenue is early and undisclosed. Management itself notes that as AI is embedded into its solutions, "monetization trends could be affected" — an acknowledgement that the pricing mechanics could cut either way [23].

The read

The moat is genuine and measurable — narrow-to-wide on the strength of switching costs, brand and domain scale, and it has survived a decade of fragmented competition without losing share or retention. The variable the case is most sensitive to is not whether the base leaves; it is whether GoDaddy can keep lifting revenue per customer roughly 8–10% a year on a base that is no longer growing in units, at the moment AI is lowering the cost of doing without a domain and a website. The strongest fact for the durable view is that ARPU has compounded through the first three years of the generative-AI era while retention held [24]. The strongest fact against it is the company's own admission that AI-powered alternatives could eliminate the need for the domain that anchors 94% of its customers [25]. What would change the read, in either direction, is the customer-count line and the ARPU line in the quarterly metrics: a base that resumes growing, or an ARPU curve that keeps bending up, argues durability; a stalling ARPU alongside a still-eroding count would say the pricing lever is nearing its limit.