Anthropic’s cross-selling playbook offers a retention lesson and a lock-in warning

Anthropic’s IPO filing shows that more than half of customers spending at least $100,000 annually now use both its Claude Platform and a first-party application. For UAE founders, the strategy offers a practical lesson in expanding customer relationships, alongside a warning to monitor how deeply one AI supplier becomes embedded in the business.

What happened

According to TechCrunch’s reporting on Anthropic’s S-1 filing, adoption across its platform and applications has risen from about 30% of these high-spending customers at the end of last year to more than half. The route is straightforward: a company can start with Claude’s API, introduce Claude Code to its engineers, then extend adoption to its wider workforce through Cowork. Anthropic can sell additional products into an existing account rather than find a new buyer for every offering.

The company supports that expansion with hands-on implementation. Its Applied AI team helps startups and enterprises integrate Claude into their existing technology. It has also launched Ode with Anthropic, a venture with Blackstone and other private equity firms that places engineers inside businesses to develop bespoke Claude-based tools. Distribution agreements with AWS, Google Cloud and Microsoft Azure provide another route to customers, while Anthropic retains the direct customer contract and sets pricing.

The filing also identifies vulnerabilities behind that distribution model. Anthropic says its cloud partners increasingly view it as a competitor and could prioritise their own computing needs or those of other companies they back. It also warns that losing a major customer, or a substantial reduction in that customer’s spending, could hurt its business unless other revenue replaces it.

Why it matters to business owners

For a UAE SME selling services or software, the useful lesson is not to build a sprawling product catalogue. It is to identify the next problem an existing customer needs solved. Anthropic’s platform, coding tool and workforce application give it several entry points into the same organisation. A founder can apply that logic by offering an adjacent service that supports the customer’s original purchase, rather than pushing an unrelated add-on.

Implementation is part of the retention strategy, not just an after-sales task. Anthropic invests in helping customers put its technology to work inside their organisations. For smaller businesses, the practical equivalent could be clearer onboarding, integration support or a review of where customers struggle to get value. These are options to test, not evidence that every SME needs a dedicated implementation team.

The buyer’s perspective is different. Using one AI supplier across development and everyday work may be convenient, but it can also increase the effort required to change providers. The reported spending figures show how far account usage can expand; they do not establish that every customer receives a matching productivity gain. UAE founders assessing these tools should separate useful adoption from growing dependence and measure outcomes alongside expenditure.

Key details

• As of September 30, Anthropic reported more than 11,900 customers spending at least $100,000 annually.
• Its customer base included more than 1,900 accounts spending over $1 million annually and about 250 spending at least $10 million.
• Anthropic attributed net revenue retention of about 900% at the end of the third quarter to expanding usage across customer organisations.
• In the third quarter, the highest-spending 5% of Claude Enterprise customers using Claude Code spent roughly 20 times as much monthly as the average enterprise user without it. This is a comparison involving a top-spending group, not a typical-user result.
• Under its cloud distribution arrangements, Anthropic says it pays providers a fee while keeping the direct contractual relationship with customers.

What founders should do

• Map the next customer problem: ask existing buyers what work remains difficult after using your core offering, then test one closely related service before expanding further.
• Review onboarding friction: identify where customers need help implementing your product and trial targeted support at those points.
• Create an AI usage inventory: list which teams use the API, coding tools and other applications, what workflows depend on them, and who owns each budget.
• Evaluate total account spending: assess platform and application costs together, and require each expansion to demonstrate a specific business outcome.
• Prepare an exit checklist: document integrations, bespoke tools and workflows that would need changing if you switched suppliers. Review relevant contract terms before adding further dependencies.

The MoreProductive takeaway

The MoreProductive view: earn retention by solving the customer’s next real problem and helping them implement the solution. As an AI buyer, apply the same discipline in reverse: expand usage where value is clear, while keeping costs visible and the practical work of switching suppliers understood.

Keep reading on MoreProductive

• Customer Experience & Retention topic hub
• The Mom Test

Sources

• TechCrunch