Home  /  Sample output

What a read looks like, on one page.

This is the shape of a Cross Data external read on a single signal. It is an illustrative, anonymized example — qualitative, and read from public information. A real read is scoped to your company and carries your own context.

How to read this. Every read follows the same seven parts, so a decision-maker can see the claim, where it comes from, how sure we are, and what would prove it wrong — at a glance. The example below is generic on purpose. No internal company data is used or implied; nothing here is a specific number we invented. — Cross Data Sample Output · illustrative · read from public information · not investment advice
Illustrative subject · a mid-market vertical-SaaS vendor · public observation

One signal, read end to end.

A frontier-AI lab announces a services arm that embeds engineers inside mid-sized companies — aimed at the exact segment our subject sells into. Here is the whole read, in the standard shape.

01 · Signal

A major AI lab launches a dedicated enterprise-services venture that places its own engineers inside mid-market companies — the same buyers the subject serves. A supplier is becoming a competitor.

02 · Source

Named public reporting and the lab's own announcement. Every load-bearing fact links to a source anyone can open — first-party newsroom posts and Tier-1 trade press, tagged by how strong each is. No private data.

03 · Implication

The "we partner with the labs" story weakens: the partner now competes for the same deals. The defensible ground shifts from access to AI (which is commoditizing) to senior judgment and accountable outcomes in a narrow, regulated niche the lab cannot responsibly staff.

04 · Confidence

How sure we are, stated plainly. The fact of the launch is High. The read that it pressures the subject's positioning is Medium — a real basis, with a credible alternative path. The pace is the least certain part, on purpose.

05 · Scenarios

Branch
What it looks like
Base
The lab serves the largest accounts first; the subject's mid-market niche stays defensible on trust and depth for now.
Upside
Buyers want an independent, accountable partner — not their model supplier grading its own work. The niche strengthens.
Adverse
The lab moves down-market faster than expected and absorbs the segment. The window to differentiate is shorter than it looks.

06 · Move

The no-regret next step — right across all three branches: run a short, cheap test of whether buyers pay a premium for senior, accountable, independent work before committing. Re-stake the story on judgment and a named niche, not on "AI access". Small spend, fast read, reversible.

07 · Counter-reading

The strongest case against us, shown on purpose: maybe the lab's venture targets only the giants and never reaches mid-market, so nothing changes for the subject. We hold that door open — and name the public event (the lab signing a mid-market client) that would settle which read is right.

What makes this decision-grade

It is specific, sourced, confidence-tagged, and falsifiable — it says what would prove it wrong. A leadership team can act on it, and check it later. That is the whole product: not a longer report, but a read you can decide on.

See the real thing, done in public

We publish full reads on real, observable companies — MacPaw, Elixirr, and Axent — sourced and falsifiable. Or start your own.

Read the public cases →