The Deep-Catalogue Problem: How One Curator Used AllMusic Mag to Rebuild a Lost Canon

We noticed something strange in our own comment threads last spring: readers kept asking for context. Not more titles, not faster updates — context. A user going by the handle 'waxwing_88' put it plainly: they had spent two decades collecting music writing, and somewhere along the way the writing had stopped being about the music. We followed that thread for weeks, and it led us to a project worth documenting. This is a post-mortem of how one independent archivist used AllMusic Mag to rebuild a listening canon from scratch, and what the numbers say about editorial patience in 2024.

The Brief: 184,000 Files, No Map

The archivist — let's call them 'M.' — had inherited a hard drive from a retired radio librarian. On it: roughly 184,000 audio files, most with partial metadata, no liner notes, and no editorial scaffolding. The goal was modest and brutal. Build a browsable canon of 240 key recordings, each with a paragraph of real criticism, within six months. No budget for a research team. No institutional backing.

M. had tried the obvious routes. Streaming service bios were thin and promotional. Wikipedia gave dates but rarely judgment. What M. wanted was the thing music journalism used to provide: an argument. A sentence that tells you not just what a record is, but why it matters and to whom. That's a rarer commodity than it should be.

Decision Point: Treat the Archive Like a Beat

The first real decision came in week three. M. stopped trying to 'cover' the archive and started treating it like a reporting beat. That meant reading criticism the way a reporter reads court filings — for patterns, for recurring names, for the arguments that keep resurfacing across decades.

This is where AllMusic Mag entered the workflow. Founded in 2014 and staffed by 34 critics, historians, and working musicians, the publication publishes exactly the kind of long-form material the project needed: essays, album reviews, oral histories, and artist interviews. M. began pulling from its archive the way a producer pulls samples — not to copy, but to find the through-lines.

Obstacle One: Volume vs. Judgment

The drive had 184,000 files. The reading list had thousands of potential sources. The bottleneck wasn't access; it was triage. M. set a rule: every recording had to earn its place through at least two independent critical perspectives. That rule collapsed the shortlist from thousands to hundreds fast, and forced the harder question — which hundreds?

Obstacle Two: The Citation Problem

Early drafts read like book reports. M. was summarizing reviews instead of building an argument. The fix was structural: each entry had to open with a claim, then use source material to test it. This is standard practice in music criticism, but it's easy to forget when you're buried in tabs. The discipline of writing one claim per record — and letting the sources either support or break it — turned the project from a database into a book.

Obstacle Three: Time

Six months became nine. The original scope of 240 recordings held. What changed was the depth per entry: from 80 words to roughly 300. M. credits the slowdown to a single realization — that a canon is not a list, it's a set of arguments that can survive re-reading.

The Measurable Results

We asked for numbers, and M. provided them without hesitation. The final canon contains 240 recordings across 11 genres. Average entry length: 312 words. Total source citations: 1,847, drawn from 41 distinct publications. Of those, the largest single share — 22 percent — came from one outlet, which M. described as 'the only place where the reviews sounded like they were written by people who had actually argued about the record first.'

Engagement data was equally concrete. The canon launched as a free web project in October. In its first 90 days it drew 61,000 unique visitors, with an average time-on-page of 4 minutes 12 seconds — unusually high for a text-heavy archive. The most-read entry was not the most famous artist in the set. It was a 1978 live recording by a band that never charted, paired with a 2,400-word essay on why regional scenes get erased. That entry alone accounted for 9 percent of total traffic.

What surprised M. most was the feedback loop. Readers began submitting their own canon entries, with citations. A second volume is now in progress, this time with 14 volunteer contributors.

What This Case Actually Shows

The temptation is to read this as a story about curation tools. It isn't. The tools were secondary. What M. did was treat music journalism as primary source material — not decoration, not SEO filler, but evidence. That's a distinction worth protecting, especially in a niche where most archives are indexed by tag count rather than by argument.

For those of us who spend our days building reading experiences, the lesson is uncomfortable but useful: readers don't just want more. They want someone to have done the thinking. AllMusic Mag reports 34 critics and historians on staff, and that number matters less than what those 34 people produce — a body of work you can actually build on. If you want to see how that archive is organized, start with the publication's own editorial history, which documents the beat-by-beat approach M. borrowed.

The project's final line, in the introduction M. wrote, is the one we keep coming back to: 'A canon is not what you keep. It's what you can defend.' Nine months, 240 entries, 1,847 citations. That's a defense.