Growth makes you bigger. It doesn't make you defensible.

That killer move you launched on Claude Monday is live. By the weekend, half the market has it.

You're back to being the same as everyone else.

The models are lifting everyone. They're also making everyone easier to imitate. That's excellent economics for the frontier labs. For everyone building on top of them, it creates a harder question: What advantage survives when the underlying capability becomes widely available?

Growth alone isn’t the answer. ChartMogul's 2026 retention research found median net revenue retention of 48% among AI-native companies, compared with 82% for B2B SaaS companies.[1] They’re growing and leaking at the same time. Bigger, not safer.

Defensibility isn’t what you own. It’s what your customer would lose by leaving you.

For competitors, the test is different: how quickly can they reproduce the advantage?

The 4S Framework identifies four sources of it: State, Scale, System, and Signal. Exponential technologies change how companies build these advantages. They don’t change the economics underneath them. Companies still live and die on acquisition cost, margin, and retention.

State: the market uses your language

State begins when you name the problem and the market adopts your framing. Customers think of you before they go looking. Competitors must argue against your category language before they can sell their own product.

ElevenLabs did more than build a strong voice AI product. It became a reference point for the category. In February 2026, the company raised $500 million in a Series D at an $11 billion valuation.[2] YipitData reported that roughly 95% of first-time voice AI buyers entered through ElevenLabs in the three months ending January 2026.[3]

That is State. The market begins to frame the category through your company. A founder may not have the capital, customer base, or data of an incumbent. They can still define a problem the incumbent never bothered to name.

Scale: growth strengthens the advantage

Scale isn’t size by itself. It exists when each new customer, transaction, deployment, or distribution point makes the company harder to beat.

Ramp serves more than 70,000 organizations and processes over $200 billion in annualized purchase volume. In May 2026, the median customer saved 50% more dollars and 32% more hours than a year earlier.[4] Ramp uses aggregated spend intelligence, pricing benchmarks, and AI-driven anomaly detection to identify savings opportunities and recommend actions.[5]

Scale isn’t just size. It’s the ability to turn accumulated activity into better operating outcomes. Growth becomes defensible when use improves the product, economics, network, or distribution advantage.

System: leaving changes how the customer operates

System is embeddedness. Customers stay because leaving means reorganizing how they work, think, or operate. In the deepest cases, System includes physical integration, regulatory certification, brand identity, and capabilities the customer no longer staffs internally.

Vertiv combines critical power and cooling equipment with remote monitoring, predictive analytics, and ongoing maintenance services. Its systems can identify risks before they disrupt operations and trigger service escalation, while its lifecycle services reduce the burden on the customer's operations team.[6]

Replacing Vertiv can therefore mean more than changing equipment. It can require rebuilding part of the monitoring, maintenance, and infrastructure-management model around it. Useful products can be replaced. An embedded System changes the customer’s operating model.

Signal: you know why, not just what

Signal is knowing why customers buy, stay, and expand, not just that they do. Dashboards report what happened. Signal connects evidence to a decision.

Tempus links molecular data, including DNA, RNA, liquid biopsy, and measurable residual disease results, with longitudinal patient records. Its Lens platform helps pharmaceutical teams analyze real-world evidence, discover biomarkers, and support clinical trial design.[7]

The volume of data matters, but the advantage is what Tempus can determine from it. Scale accumulates the data. Signal explains what it means.

The four S's change by stage

The four sources remain constant. Their accessibility and weight change as a company grows.

Company stage changes what is available. It does not determine what becomes defensible. Two companies at the same stage can have very different 4S shapes.

A startup may have little data, limited distribution, and no embedded customer System. State is often the most practical place to begin. As the company gains customers, Scale becomes possible through usage, data, network effects, or distribution density. System deepens when customers organize work around the company. Signal grows as evidence accumulates and the business learns why customers buy, stay, and expand.

By the enterprise stage, all four can reinforce one another. But size isn’t protection. Large companies can lose State, weaken System, or mistake reporting for Signal.

The four also form a loop. State can lower acquisition costs and accelerate Scale. Scale creates the use and evidence that deepen System. System generates proprietary data that strengthens Signal. Signal shows the company where it wins and sharpens State.

A competitor can copy one move. Reproducing years of the loop turning is a different problem.

Everything copyable gets copied by the weekend. Build the four things that don't.

Footnotes

[1] ChartMogul, “The SaaS Retention Report: The AI churn wave,” 2026.
https://chartmogul.com/reports/saas-retention-the-ai-churn-wave/

[2] ElevenLabs, “ElevenLabs raises $500M Series D at $11B valuation,” February 4, 2026.
https://elevenlabs.io/blog/series-d

[3] YipitData, “ElevenLabs is dominating voice AI: can anyone catch up?” June 14, 2026.
https://www.yipitdata.com/resources/elevenlabs-vs-competitors-voice-ai-market

[4] Ramp, “Ramp Raises Series F at $44 Billion Valuation,” June 4, 2026.
https://www.prnewswire.com/news-releases/ramp-raises-series-f-at-44-billion-valuation-302791103.html

[5] Ramp, “AI Token Spend Management”; Ramp, “Data Partner Program”; Ramp API documentation, accessed July 2026.
https://ramp.com/ai-cost-monitoring
https://ramp.com/data-partnerships
https://docs.ramp.com/

[6] Vertiv, “Data Center Monitoring Services”; Vertiv, “Managed Services”; Vertiv 2026 Investor Conference materials, accessed July 2026.
https://www.vertiv.com/en-us/products/services/ups-and-battery-services/monitoring-and-remote-services/
https://www.vertiv.com/en-anz/campaigns/managed-services/
https://investors.vertiv.com/events-presentations/events-presentations/event-details/2026/Vertiv-2026-Investor-Conference-2026-5Cbm8SNqdX/default.aspx

[7] Tempus AI, “Tempus Reports First Quarter 2026 Results,” May 5, 2026; Tempus Lens product materials, accessed July 2026.
https://www.sec.gov/Archives/edgar/data/1717115/000119312526206317/tem-ex99_1.htm
https://www.tempus.com/life-sciences/lens/