Six questions every buyer asks before they pay — answered straight. With the real artifacts, the real cases, and the pattern library behind them. Where we can't prove something yet, we say so plainly.
Not another report. A map of the environment, a set of options, and the logic for action. Four things land on your desk:
Decision-grade means: every claim sourced, every option scenario-backed, every recommendation with a no-regret floor.
We don't post client revenue charts. The practice is young, and we won't dress up a number we can't yet stand behind. Two things we can show honestly: how the decision itself changes, and how our outside-in reads have held up against what companies later disclosed.
MacPaw at a platform-shift moment: six patterns intersecting around its mission-led bet (Eney) and the Apple-Intelligence shift. High
Five of those reads were echoed by MacPaw's own Social Impact Review 2025. Consistency, not independent proof — but the outside-in read lined up.
Elixirr (LSE: ELIX): on the winning side of a consulting industry splitting in two — with three dated, falsifiable predictions about its next moves, counter-case shown. Medium
When Elixirr's own disclosures land, we mark each prediction right, partly right, or wrong — in public. A bet placed before the fact beats a chart drawn after it.
One read already echoed by a company's own disclosure. One bet placed before the fact, to be graded in public. That is the honest version of proof for a young practice — and not investment advice.
One disciplined loop, run on your real situation. AI does the scanning. We do the judgment, the confidence tags, and the decision.
AI is genuinely good at scanning. What it can't do is take accountable judgment, remember your past cases, or own the decision. Here is the honest comparison — no put-downs.
We sit in one corner on purpose: mid-market affordable, accountable cross-case judgment. Not the cheapest, not the biggest — the one that owns the decision.
Yes — and we're honest about what they are. Two named public companies, read end-to-end from public information only. Not paid engagements: this is the method on display. The confidential version is your company, your data, under NDA.
Three more readings, anonymized — the inside frame, and the move an external read names:
Top-2 benchmark, no commercial moat.
Lock one buyer thesis for 12 months; convert free accounts to design partners before the category closes.
AI tutors undercutting price 5–10×; "is AI eating us?"
Stop competing in the AI-comparable middle; re-anchor on human trust AI can't manufacture; deploy AI inside the marketplace, not against it.
Treat a 5M-person diaspora as temporary demand?
Read it as a structural demand layer; separate product and brand; open a cross-border payments line; move in 12 months, not 3 years.
Anonymized to keep attention on the strategic shape, not the identity. These are external reads, not engagements we delivered.
Not mockups. Real artifacts you can open and read right now.
A short, confidential read of your external position. No commitment, no exposure. The diagnostic ends with you holding the document — there's no next step we own.
We can promise a sharper decision: sourced, confidence-tagged, with the counter-case shown. We won't promise a number we can't yet measure. If the read only confirms what you already believed, it still earns its keep — it turns a hunch into evidence.