Backbone is contested, not safe

The previous entry advised SaaS founders to own a backbone or own an outcome. Reading it back, the symmetry is misleading. Backbone isn't safe — it's contested infrastructure with three winners per slot. Outcome is a durable position. They aren't equivalent strategies, and the piece should have said so.

The backbone is winner-take-most by structure. Network effects, integration depth, regulatory licensing, and reliability brand mean each slice — identity, compliance, payments, integrations, data — sustains 1–3 winners, not 10. Auth converged on Okta-Auth0, with WorkOS as the AI-native challenger. Payments has Stripe and Adyen. Data has Snowflake and Databricks. Compliance has Vanta, Drata, Secureframe. The seats are mostly taken. Telling a SaaS founder "become the backbone" in 2026 is mostly telling them to fight for a top-three slot in a category that already has its top three.

Three forces compress the backbone category further as more vendors pile in.

First, the differentiator dilutes. When every SaaS claims to be platform infrastructure, "backbone" becomes the new "platform" — a marketing word, not a structural position. By 2020 every SaaS was "a platform" and the word stopped meaning anything except in the few cases where it actually did.

Second, substitution erodes pricing power even for winners. The reason backbones can charge rent is that alternatives are expensive. When AI lowers switching cost and credible alternatives proliferate, leverage shifts back to the customer. Software margins stay high but pricing softens. Public multiples have already moved this direction.

Third, re-bundling at the cloud layer. AWS, Azure, and GCP already sell identity, compliance, integrations, AI infrastructure. Microsoft 365 and Salesforce Platform absorb adjacent backbones. The cloud IaaS commoditisation story is repeating one layer up. In ten years, many of today's pure-play backbones will sit inside a hyperscaler bundle.

Outcome ownership doesn't compress the same way. Outcomes are domain-specific, accountability-bearing, and require absorbing real liability — data integration, model risk, operational responsibility, regulatory exposure. None of that gets cheaper as backbones commodify. If anything it gets more valuable, because cheap backbones plus cheap value-layer composition raise the obvious question: who is responsible when the assembled thing produces a wrong outcome? That position — "I am responsible, here is my data, my model, my SLA on the result" — does not have three winners per category. It has as many winners as there are domains where someone is willing to underwrite the result.

The sharper prescription, then:

• Outcome ownership is the durable position.
• Backbone is a category fight with three winners per slot, and the cloud layer is coming for the rest.
• Pure value layer is the squeezed seat.

For SaaS founders, the practical question is not "backbone or outcome" but "which outcomes can I credibly underwrite in a domain where I know what good looks like." For buyers, the question is whether the SaaS you pay for is taking responsibility for a result, providing genuinely irreplaceable infrastructure, or just renting you a workflow that's about to be composable.

If you're weighing the backbone path against an outcome play, push hard on the outcome side. Reach out — happy to think through what your domain actually lets you underwrite.