The four-minute clerk

Every dispute in the United States goes through the same machine. A bureau clerk reads your letter with about four minutes to spend and compresses it into one of 29 dispute codes. An automated form carries that code to the furnisher, and the furnisher's default is to check the code against its own record and answer verified. No one re-reads your prose.

Regulators' studies found that roughly 90% of disputes collapse into the same four generic codes. Whatever you wrote, however long you spent writing it, the machine reduces it to a category.

The industry's answer to this machine has always been volume. Dispute everything, reword it monthly, and hope the deadlines break before the classifier does. The largest practitioner of that model ended with a $2.7 billion CFPB judgment.

#The machine every dispute goes through

The pipeline has six stops, and your letter only controls the first one. Everything after the clerk is automation resolving a code against a database.

A dispute succeeds when its claim survives code compression intact and lands in a form the furnisher's automation cannot resolve against its own record. That is the entire game. Not persuasion. Survival.

#Writing for the machine

Fluid writes for the machine. One letter, one Metro 2® field contradiction, one code for the clerk to select, one claim the furnisher's own data cannot refute. DisputeWell runs 48 field-level compliance checks from the 2023 CRRG against every tradeline, so a claim starts as a data structure, not a sentence.

And when an account gives it nothing to work with, it does not dispute. No grounds, no dispute. That is a rule, not a setting.

#Round two is a different legal instrument

Most software treats round two as round one, reworded. The law treats it as a different instrument. What the bureau did with round one determines which statute round two invokes, and Fluid selects the escalation from the outcome on record.

Deletions feed the next strategy. If an item comes back after a deletion, the system catches the reinsertion on the next import and drafts the §611(a)(5)(B) notice. You are never asking the machine the same question twice.

#What we don't claim

An industry this crowded with overclaiming deserves a short list of what we will not say.

• e-OSCAR access. No one has it. Not us, not anyone selling you an integration. Everyone in this industry mails letters.
• Full Metro 2 visibility. Consumer reports are a projection of what furnishers file. We analyze the projection and keep a ledger of exactly what it cannot show.
• Point predictions. Our scores are calibrated rankings, not FICO point promises. Nobody honest promises points.

The machine is not going away. The letter that beats it is the one written for it.