The Cobalt Glacier B2B SaaS brand-readiness checklist.

The temptation in any new-brand process is to treat validation as a checkbox. A pitch deck, a few customer calls, a rough model, greenlight. The temptation gets stronger the more excited everyone is about the idea, because nobody wants to be the one asking the question that stalls it.

On a twenty-five-year hold, the cost of skipping a question before we commit build capital is measured in years of a small team's time and, more importantly, in the opportunity cost of not building the idea we should have built instead. We owe the operator we're building with a process honest enough to say no before we start, and we owe ourselves a process that does not produce surprises after the brand is live. That is why the Cobalt Glacier validation checklist is longer than most seed-stage decisions and shorter than most private-equity diligence processes: every item earns its place by connecting to a specific decision we have to make before we build.

How pre-build validation differs from acquisition diligence or a venture pitch

A venture investor is mostly underwriting an upside scenario on someone else's execution. An acquirer is mostly underwriting an existing business's numbers. A studio committing its own build capital is underwriting whether a team it assembles, on a category it chose, can produce a business that survives every macro environment, every operator transition, and every competitive entrant for the next twenty-five years, starting from nothing. The question set is therefore different. We care less about a polished pitch and more about which structural characteristics of the idea will still be true in a decade.

We described the underwriting frame in How we underwrite a brand idea for a 25-year hold. This essay is the operational counterpart: the actual checklist that translates that frame into a process an operator can prepare for before we commit a team.

Validation is not a search for reasons to build. It is a search for reasons not to. Everything else is planning, not validation.

The six workstreams

Every Cobalt Glacier brand idea goes through the same six workstreams, run in parallel, with named owners on both sides. We share the structure with any co-founding operator early so the process feels legible from the first conversation.

1. Commercial validation

The commercial workstream tests whether real budget exists for this problem today. It starts with structured conversations against a target buyer list, not a survey. We want to hear, in the buyer's own words, what they currently pay to solve this problem, what breaks when they don't solve it, and what would make them switch. Soft interest ("this looks interesting") does not clear the bar. A named budget line and a described switching trigger does.

• A target buyer list of at least twenty-five real prospective customers, with named roles and current spend on adjacent tools.
• Willingness-to-pay evidence from structured conversations, ideally including a small number of prospects willing to commit to a paid pilot before the product exists.
• A credible go-to-market motion that matches how this buyer actually buys software today, not an aspirational motion borrowed from a different category.
• A competitive map of incumbent tools and workarounds, including the honest answer to why they haven't already solved this.

2. Product and engineering validation

The product workstream answers two questions. First, can a small, AI-native team actually build a credible version of this in the window we're budgeting, typically three to six months to a sellable first version. Second, is there a real proprietary data or workflow asset available to us here, or would we be building a thin wrapper on a foundation model that any competitor could replicate in a sprint.

• A scoped architecture for the first sellable version, with an honest build-time estimate from the engineers who would actually build it.
• A data-moat assessment against the framework in Where the AI moat actually lives in a brand we build.
• A dependency map of third-party services and models the product would rely on, with a plan for what happens if any one of them changes terms or pricing.

3. Customer validation

The customer workstream is the one that most reliably kills a promising-sounding idea. We insist on direct conversations, run by the operator who would run the brand, across a spread of the target buyer list: enthusiastic prospects, skeptical prospects, and at least a few who explicitly said no. Every conversation is anchored on the same five questions, scored independently by two people on the team, and rolled up into a validation memo before we commit a team to building anything.

4. Financial validation

The financial workstream builds the bottoms-up model described in our underwriting essay and tests it against the free-cash-flow conversion bar we hold every brand to, the same one we wrote about in Free cash flow conversion is the underwriting bar. Ideas that cannot plausibly reach that bar at a realistic scale don't clear the process, no matter how good the demo looks.

5. Legal and IP validation

The legal workstream looks for the small number of items that can actually block a brand before it starts: an outside operator's existing non-compete or IP assignment obligations from a prior employer, open-source license obligations that would creep into the production codebase, and any regulatory posture the category requires from day one. None of these are common. All of them are launch-blocking when they appear.

6. Operator continuity validation

The continuity workstream is the one that distinguishes a studio's process from a venture fund's process. We are building a brand we default to running indefinitely, which means we have to know, before we commit a team, exactly who will run this brand day to day and whether that person genuinely wants to do it for a decade. We described the philosophy in Who runs a brand we start, and why they stay. The validation version is a named operator, a written commitment conversation, and an honest read on whether they're excited about the launch or the decade.

What an outside operator should prepare before the first call

The single best predictor of a clean validation process is whether the operator pitching a brand idea has already done the legwork themselves before the conversation started. The evidence exists either way; the question is whether it has been organized into a form the studio can evaluate quickly.

• A written one-page thesis: the buyer, the workflow, the current alternative, and why now.
• Notes or recordings from at least ten real conversations with prospective buyers, including the ones who said no.
• A rough bottoms-up model of the first three years, even if it's wrong in the details.
• A short, honest written narrative of the three things that worry the operator most about the idea. We are going to find them anyway; hearing them in the operator's own voice changes the tone of the entire process.

The disqualifiers

Most ideas that don't get built are disqualified early, in the commercial or customer workstream, for reasons that were visible on the first call but only confirmed in week three. The most common disqualifiers we see are a target buyer who says the problem is real but not urgent, a category where the honest answer to "why hasn't this been solved" is that it's genuinely hard to build a durable business around, a data asset that would arbitrage to zero the moment the underlying model improves, and an operator who is excited about the idea but not committed to running it for years.

None of these are moral failings. They are structural mismatches with a twenty-five-year build. The right outcome in each case is a fast, honest no, delivered with enough specificity that the operator can use the feedback elsewhere.

The bottom line

A validation checklist is not a substitute for judgment; it is a scaffold that lets judgment focus on the items that matter. Run across six workstreams in parallel, scoped to the questions a patient-capital studio actually has to answer before it commits a team, the checklist compresses a process that would otherwise sprawl across months into one that respects everyone's time and produces a defensible decision either way.

If you're an operator with a brand idea and the domain depth to run it, we would like to hear from you. The first call is short, the process is structured, and the decision either way is delivered with specificity. If you'd rather run a brand we've already conceived, the Operating Partner program is the door.