Current corpus
334 companies
Public-safe research already visible before anyone buys the brief.
Method & Boundary
How public-source company research becomes one usable view of buying friction
The method is not to publish company dumps or abstract category commentary. It starts with dense company-level research, reduces that into atomic observations and normalized signals, then ranks recurring blockers before anything becomes a public dispatch or canonical blocker page.
Evaluating whether to buy?
Use one specimen and one diagnosis path. Then check fit or start the brief.
Published blockers
4
Canonical pages live now; dispatch history grows separately.
Reviewed releases
6
Older cohorts stay visible instead of being silently overwritten.
Step 1
Capture dense company-level research
The work begins with messy public-source evidence and the specific buying-friction points already visible inside each researched company.
Step 2
Split notes into atomic observations
One prose blob is not a useful publishing unit. Each note has to be broken into the smallest observations that still carry one decision-relevant idea.
Step 3
Normalize signals and rank recurring blockers
The observations are classified into a small taxonomy, deduplicated, and clustered so the site can publish what keeps repeating instead of what happened to be loudest.
Step 4
Publish only the public-safe pattern layer
The paid brief and the public site inherit the same ranking logic, but the public layer publishes only what survives public-safe abstraction.
Method Principles
Pattern layer, not company dumps
A researched company is an input. A recurring buying-friction signal is the working unit. A public artifact is the output.
Rank before writing
The public topic should be selected by recurrence and commercial weight, not by whichever example feels most vivid.
Stop before the private correction
The site exists to make the thinking legible. It should never become a disguised mirror of private company research.
Before you buy
What a cold buyer can verify before paying
Trust comes from artifacts a skeptical buyer can inspect before the private brief starts, not from process description alone.
Engagement Model
Async-first by design
The first product is buyable without a heavy discovery process. The goal is to reduce decision friction, prove signal quality quickly, and keep calls optional unless they clearly add value.
Current Published Batch
What the public layer can already prove before anyone buys
The current public batch shows how much recurring pressure is already visible before a company buys the brief. The internal clustering layer currently groups 5 recurring clusters, while 4 blocker pages are published on the site right now.
That gap is intentional. A cluster can be ranked internally before it has enough public-safe support, stable wording, or durable evidence to justify its own canonical blocker page.
Companies
334
Current public corpus.
Public signals
2024
Structured public-safe signals carried into the current published layer.
Published history
6
Published cohorts preserved instead of silently overwriting the last one.
Internal clusters
5
Recurring patterns ranked before writing starts.
Published blockers
4
Canonical blocker pages live on the site now. Dispatch history grows separately when a cohort materially changes the public proof.
Publication Profile
Public-safe publication layer
The publication layer sits downstream of stored private research. That boundary is what lets the site prove recurrence publicly without exposing company names, contacts, or the intervention path that belongs in the brief. Blocker pages refresh with the latest active cohort, while dispatches remain a published change log.
Current recurring pattern
In the current published batch, decision drift recurs in 331 of 334 companies. The public layer can prove that recurrence without exposing the private intervention path.
Selected outputs
What the public layer publishes before you buy
The method becomes inspectable through one current dispatch and the blocker pages that keep the stable explanation visible.
In this selected cohort, buyers often had a defensible reason not to move.
In this manually reviewed selected cohort of 334 active companies, decision-drift evidence appeared in 331 profiles and 801 tagged signals. The pattern was narrower than generic indecision. In selected cases, buyers had defensible reasons to delay when the savings model, governance packet, price metric, migration wedge, or deployment owner was unclear. The paid brief is most useful if you need to know which fallback your buyer can defend internally and what proof would make that fallback less safe.
Decision drift
A deal can stay active while the next owner, scope, or exit condition never becomes crisp enough to move.
Proof friction
The buyer still needs concrete proof they can check before the team gets a clean comparison at all.