research protocol
A Reproducible Remote Job-Search Experiment
Career advice often turns one person's outcome into a universal rule. This protocol shows how to test a narrower question—such as whether a tailored work sample improves qualified responses—while controlling workload, protecting personal data, and reporting uncertainty honestly.
Start with a question that can be measured
“How do I get a remote job?” is too broad for an experiment. A testable question changes one behavior for a defined set of suitable roles: Does adding a short role-specific work-sample note change the rate of qualified recruiter responses? Does applying within forty-eight hours of publication change the response rate among otherwise similar openings?
Write the question, eligibility rules, primary measure, sample target, and stopping date before sending applications. Predeclaring those choices reduces the temptation to redefine success after seeing a few outcomes.
Define a qualified application first
Include only roles you would genuinely accept and for which you meet the essential location, authorization, discipline, and level requirements. Exclude duplicates, roles already contacted through a referral, and postings that close before submission. Record the first-party URL and verify it is active at application time.
This rule protects both the measurement and employers from indiscriminate applications. A higher response rate achieved by switching to much better-fit roles does not isolate the effect of a cover note or submission time, though it may still teach a valuable search lesson.
Use a simple comparison design
For matched roles, alternate between a baseline application and one treatment. Keep the resume quality, role family, seniority, location, and weekly effort as consistent as practical. Do not run several changes at once; if the treatment combines a new resume, portfolio, referral, and application timing, you will not know which mattered.
Random assignment is ideal but can be uncomfortable when opportunities are scarce. A transparent alternating rule is easier: after ordering qualified roles by discovery time, apply baseline to one and treatment to the next. Record exceptions and their reasons rather than quietly moving attractive roles into the preferred group.
- Primary measure: qualified positive responses per submitted application.
- Secondary measures: screen invitations, time to response, and time spent applying.
- Guardrail: never reduce application accuracy or misrepresent experience.
- Stopping rule: a fixed date or predeclared application count.
Track outcomes without overclaiming
Use statuses such as submitted, rejected, no response by cutoff, recruiter question, screening interview, later interview, offer, and withdrawn. Define “positive response” before starting. Separate automated acknowledgements from human interest and keep withdrawals distinct from employer rejections.
A small personal experiment cannot establish a universal causal effect. Companies, recruiters, timing, and role fit differ. Report the raw numerator and denominator, the period, inclusion rules, and deviations. “Six of twenty treatment applications received a qualified response versus three of twenty baseline applications” is more honest than declaring a tactic doubles everyone's odds.
Protect people and confidential information
Do not publish recruiter names, private messages, application questions, or employer assessments without permission. Store only the personal information needed for your own follow-up. Respect application terms and do not automate submissions or create fake candidate identities.
If sharing results, aggregate company data unless a public posting is essential to the method. The experiment evaluates your process, not individual recruiters. Delete unnecessary notes after the search and keep source links only as long as they serve a legitimate recordkeeping purpose.
How this site will report future results
This article publishes the protocol, not invented interviews or outcomes. Any future TechJobsRemote experiment should disclose who participated, dates, selection rules, treatment, measures, missing outcomes, conflicts, and the complete result—including a null or negative finding.
Until such data exists, use the monthly market report to choose a sufficiently active category and the skills analysis to define a truthful, role-relevant treatment. The verification guide provides the first-party source checks for the application sample.
Sources and further reading
continue with live listings
Browse relevant remote jobs
related research