"We need to raise our Domain Authority" is a sentence that reveals the misunderstanding in real time. The scores — DA, DR, AS, whichever vendor — aren't Google metrics. They're third-party estimates of link strength, built by companies that don't have access to search data, modeled from crawlable links. They describe something real. They are not the thing itself.
What the scores are actually for
Used correctly, authority scores are filters: comparing candidate sites for outreach (is this a real publication or a link farm?), spotting a sudden drop that might mean lost links or deindexing, and sizing up a competitor set cheaply. As a diagnostic lens, they're fine. As a KPI, they're fiction with a spreadsheet.
Why chasing the score backfires
Optimizing for a modeled number means optimizing for the model's inputs — and the models weight raw link counts heavily enough that link-buying looks efficient. So teams buy packages, the number ticks up, nothing in traffic moves, and the disavow file eventually grows. Meanwhile the inputs that actually move rankings — relevant links from real publications, content depth, entity consistency — don't respond to package deals. You can win the proxy and lose the game.
What to track instead
The measurable version of "authority" is behavior: branded search volume (are people looking for you by name?), links from sites your audience actually reads, pages ranking for competitive head terms, and — the accountant's version — organic conversions. Every one of those is real, reportable, and immune to model updates.
The honest use case
Before outreach: score the target once, skim their content once, decide. After that, ignore the number and evaluate links the way engines do — by relevance and audience. The moment authority scores become monthly targets, they've stopped being tools and started being taxes.
