Research
Measured, not assumed.Findings from reading real postings and real application forms.
Most job search advice is written from memory. The work below started the other way around: iapplyai.app reads employer job boards and application forms as part of doing its job, so the questions people argue about can be answered by counting instead of guessing. Each area links to the full write up.
What application forms actually ask
The same six fields appeared on every form measured, and the questions beyond them were the same handful reworded rather than a long tail of surprises.
How it was measured. Reading the published form definitions of real matching roles across several employers.
The screener questions that repeat
Seven questions recur across employers in different words: work authorization, sponsorship, location, referral source, prior employment, demographics and start timing. Answer them once and you have answered most forms.
How it was measured. Comparing the non standard questions on real forms and grouping them by what they are actually asking.
What listings disclose about pay
Roughly twenty eight percent of listings publish pay. Most searches are therefore run on inference, which is worth knowing before you treat a stated band as the norm.
How it was measured. Counting published compensation on the postings iapplyai.app reads.
Whether cover letters are still required
Across the roles measured, the cover letter was optional on almost all of them and required on none. That is a different question from whether one helps.
How it was measured. Recording the required, optional or absent status of the letter field on real matching roles.
Company boards against aggregators
A role appears on the employer board first, and the employer board is the only place the application form can be read before you commit to it.
How it was measured. Comparing where the same roles appear, and what information is available at each source.
The real supply of remote roles
In one industry sample, the roles that scored well were concentrated almost entirely at the fully remote employers, and the rest advertised widely without matching.
How it was measured. Scoring every advertised role at a set of employers in one industry against one candidate profile.
Every finding links to its working.
The write ups state what was measured and how, so you can disagree with the conclusion on the evidence rather than on the claim.