Customer Workflows Are the New Moat for AI Companies

Quick Facts

• Schneider Electric agreed to acquire PTC for $22.6 billion, buying workflow access across 30,000 industrial customers globally.
• ElevenLabs hit $600 million in ARR by June 2026, powered by deep integrations with Revolut, Klarna, and 41% of Fortune 500 companies.
• Strategic adviser Itay Sagie argues the real AI moat is not the model itself but how difficult the product is to replace inside a customer's operations.

The AI company that wins is not always the one with the best model. It is the one that becomes hardest to remove. That is the central argument Itay Sagie, a strategic adviser to tech companies and investors, makes in a Crunchbase News column published Oct. 9, 2026.

Sagie opens with a concrete scenario: two AI assistants tested by an insurance company. Both perform well in demos. Six months later, one handles policy renewals inside the insurer's own systems, follows approval rules, and escalates exceptions to staff. The other is a tool people occasionally open. A better model could arrive tomorrow, but replacing the first assistant means changing how the company operates.

Buying Workflow Access

Schneider Electric's agreement to acquire PTC for approximately $22.6 billion in equity value is Sagie's primary case study. The all-cash offer of $205 per share represents a 42% premium to PTC's last closing price and is expected to close by Q3 2027.

PTC provides computer-aided design, product lifecycle management, and service lifecycle management software to more than 30,000 customers globally. The Boston-based company generated 2.4 billion euros in revenue in calendar year 2025, with an adjusted EBITA margin of roughly 40%. Schneider projects 250 million euros in annual run-rate cost synergies by year three, alongside approximately 800 million euros in anticipated revenue synergies.

Schneider CEO Olivier Blum called the deal "an important step forward in our ambition to lead the new era of energy and industrial intelligence." Investors responded cautiously. Schneider's shares fell nearly 10% in early Paris trading, erasing close to 15 billion euros in market value. Jefferies noted that AI disruption fears allowed Schneider to acquire PTC at a decade-low valuation but warned that dynamic could also weigh on Schneider post-close.

Sagie's read: Schneider is not buying software. It is buying a seat inside customer decisions across the full lifecycle of physical products.

The Synopsys Model

A separate deal reinforces the same logic. OpenAI and Synopsys signed a multi-year preferred partnership to develop GPT-Synopsys, a specialized model for chip design that pairs OpenAI's AI with Synopsys' established electronic design automation tools. The agreement includes revenue sharing and a joint global go-to-market strategy.

Synopsys now forecasts 15% revenue growth for the fiscal year, up from analyst expectations of 11.19%. Shares climbed as much as 7% following the announcement. Sagie frames this as another case of an AI developer needing a workflow incumbent to commercialize effectively. Synopsys brings customer relationships, domain expertise, and trusted processes that OpenAI cannot replicate quickly.

Recurring Work as a Retention Engine

ElevenLabs illustrates how deep workflow integration drives revenue. The company hit $600 million in ARR by June 2026, up from $330 million at the end of 2025, according to estimates from Sacra. Its AI voice platform is used by 41% of Fortune 500 companies.

In January 2026, Revolut deployed ElevenLabs Agents for customer support across the UK and Europe, covering more than 4 million customers and reducing resolution time by 8x across 30-plus languages. In February, Klarna launched an ElevenLabs voice agent as first-line phone support for 35 million U.S. customers, reporting up to 10x faster resolutions.

The risk is real. OpenAI and Google are shipping native voice APIs. If voice generation becomes a commodity, ElevenLabs' defensibility rests entirely on its enterprise workflow layer.

What Founders and Investors Should Track

Sagie draws clear lines for each audience. Founders should build for a specific workflow and measure how hard they are to replace, not how impressive their demo looks. Established businesses with existing workflow access have an opportunity to convert that position into bargaining power.

For investors and acquirers, the standard has shifted. AI infrastructure is absorbing hundreds of billions in annual spend. The market is no longer rewarding ambition alone. It is demanding proof that capital turns into durable cash flow.

The question Sagie says every founder should be able to answer: not how good is the technology, but what would it cost the customer to remove it?

Read more: Why Customer Workflows Are Becoming The Moat In The AI Era