The Target Viability Standard, Graded on Live Demand

You have picked a role, a location, and an industry to aim at. Before you spend weeks writing a tailored resume and cover letters for that target, you need to know whether it has enough live hiring to sustain a pipeline. This guide is a pass/fail standard for that one decision: it gives you explicit thresholds and a graded checklist so that two people checking the same target reach the same verdict on whether to commit, widen, or drop it.

The common advice is to pick a target with "hundreds or thousands of opportunities." That sounds like a bar but it is not one, because it never says how to count, what to discount, or what number is too low. People commit months to targets that never had the live demand to carry a weekly pipeline, and they find out eight weeks in. This standard turns viability into something you grade in about two hours, before you build anything.

What "viable" means for a job target

A target is viable when it has enough fresh, distinct, trending, and payable live demand to feed your weekly application pace across a multi-week search without re-applying to the same postings. That is the definition of done, and every axis below is a condition of it.

The reason a single posting count fails as a standard is that four different things hide inside it. Two targets can both show "200 postings" and have completely different viability: one is 150 real openings across 90 employers with wages above your floor, the other is 200 listings where half are stale, a third are one agency, and the pay tops out below your minimum. A standard has to separate those cases. So viability is graded on five axes, each with its own evidence and its own way of lying to you.

The order matters less than the completeness. A target that passes four axes and fails pay is not "80% viable" - it is a fail, because you cannot take a job that pays below your floor no matter how many openings exist.

Why a raw posting count overstates demand

A raw posting count overstates live demand because a large share of listings are stale, ghosted, duplicated, or evergreen, and none of those will convert into an interview. You must discount the raw number before you grade it.

The age of a posting is the cleanest public signal. Postings 0 to 30 days old are fresh and carry the highest response rates; 30 to 60 days are stale; 60 or more days should be treated as inactive. The response penalty is steep: postings over 30 days old draw about a 48% lower response rate than those under 14 days. And staleness is not rare. The median open posting is already 35 days old, and one study of 96,902 Indeed ads found 34% were posted 30 or more days earlier, with technology highest at 43%.

Then there is the ghost-job tail. Roughly 1 in 7 active posts are ghost jobs, and 4% have been active for four or more months. Evergreen requisitions compound this: 25% of one large posting dataset stayed up for more than 90 days and 10% for more than 180 days. An always-on posting is not churn you can ride; it is a low-yield backdrop. The practical rule is to remove anything 60 or more days old, then haircut the remainder by about a quarter to account for the stale and duplicated listings you cannot individually verify.

Deriving your application target and live-posting floor

Your live-posting floor is not a universal number; it is your weekly application target divided by the share of postings you actually fit and can apply to. You derive the weekly target from your interview goal and your real conversion rate.

The conversion math is public even though the threshold is not. Cold online application-to-interview conversion is low: only about 3% of applicants reach an interview in one 10M+ application dataset, and the average sits around 2 to 3%. Expressed as volume, it takes an average of 42 applications to land one interview, a hit rate of about 2.4%. So if you want one interview per week, you need roughly 40 applications per week (1 ÷ 0.025 = 40). The sustainable human pace is 10 to 20 quality applications per week, which is itself a constraint worth naming early.

Channel changes this sharply. Careers-page applications reach interview 6.87% of the time, against 2.92% on a major board and 1.95% on another. A careers-page-first target needs roughly 3.5x fewer live postings to hit the same interview goal as a board-first one. So your floor depends on where the postings live, not just how many there are.

The honest status: the arithmetic above is published, but no public source sets a pass/fail live-posting threshold. That gap is exactly what this standard fills. The derived floor is: adjusted live postings you fit must be at least enough to supply your weekly application number for the full length of your search without re-applying. If you plan an eight-week search at 40 applications a week and can apply to a third of matching postings, you need on the order of 40 ÷ 0.33 × (new supply each week) to keep the pipeline fed - which is why velocity, below, matters as much as the standing count.

Reading market size without confusing it for demand

Market size and live demand are different measurements, and conflating them is the most common way a target is miscalled. Talent-pool size tells you how many people hold the role; live postings tell you how many employers are hiring it now.

Refolk's index gives the prevalence side cleanly. In Refolk's index of professional profiles, there are 65,397 current Data Analysts in the United States against 16,448 in the United Kingdom - a roughly 4:1 market-size gap for the same title. That is a useful sanity check on whether a geography can support a search at all, but it is supply, not openings.

The seniority effect is the one that quietly kills targets. Moving one band up shrinks the addressable pool far more than people expect.

The Director of Engineering pool is just 12.5% of the Senior Software Engineer pool in Refolk's index - roughly an 8x drop for one rung. A target that is comfortably viable at IC level can fail purely on market size one band up, before you even count a single posting. When you are deciding how senior to aim, run the prevalence check first: it is cheap and it reframes the whole search.

Talent-pool size answers whether a market exists; live postings answer whether it is hiring you this week.

The fast way to see both the breadth and the named employers behind a target is to search the index directly rather than hand-counting profiles. Refolk returns the distinct current employers for a role and geography, which is the breadth number you will reconcile against your live posting pull in the procedure below.

Which free public sources to trust, and for what

Two free public sources carry most of the weight, and each is distorted in a specific way you must correct for. Use them for direction and context, never as a standalone verdict on an occupation target.

The first is BLS JOLTS, which publishes openings, hires, and separations by industry and region. Job openings stood at 7.1 million in August 2026. Its distortion is decisive for this job: JOLTS does not collect information by occupation, and its state estimates are modeled rather than a job-level count. If your target is an occupation in a metro, JOLTS cannot grade it.

The second is the Indeed Hiring Lab Job Postings Index, which reports the percentage change in seasonally-adjusted postings since a February 2020 baseline using a seven-day trailing average. A reading of 101 means 1% above that baseline. Its distortion is that it is an index of change, not an absolute live count, and its historical numbers are revised. Use it for trend direction only. For context, the overall index stood at 101.8 in August 2026, while Software Development postings sat at 74.4, up from a low of 61.1 in May 2025 - a sector well below baseline but recovering.

The defensible verdict combines three things: an index trend for direction, a hand-counted live pull for the adjusted count, and a prevalence check for market size. No single free source does all three.

The grading procedure

Run these eight steps in order; the whole pass takes about two hours. The output is a written verdict with the numbers that drove it, which is what makes the grade reproducible.

One sequencing disagreement is worth flagging. Recruiter write-ups often push building a target-company list first, narrowing to under roughly 50 to 100 employers before counting anything. Funnel-math sources imply counting live volume first. This standard counts volume first because you cannot know whether 50 employers is enough until you have graded the demand they generate - but if you already have a curated company list, run the employer-breadth step against it rather than a board pull.

How this grading goes wrong

Most miscalled targets fail in one of eight predictable ways, and each has a false positive that looks like a pass. This is the part of the standard worth re-reading before you trust any verdict.

Two of these deserve extra weight. The employer-concentration trap is common and quiet: a staffing agency reposting the same mandate across boards can make a thin market look deep. If one employer holds more than 30% of your adjusted postings, the target is one freeze away from empty, regardless of the headline count. And the over-broadening trap is the one that follows a failed grade: you widen the target to span six unrelated titles so it "passes" the count, and then every application you write is weaker because it is positioned for nothing in particular. A broader search often produces a weaker message. Widen one variable at a time and re-grade, or you have traded a demand problem for a positioning problem.

Remediating a target that fails one axis

When a target fails one axis, change one variable at a time and re-grade, because changing everything at once destroys both the diagnosis and the message. The safe order is format, then geography, then adjacent titles, then industry.

Geography and format are the highest-leverage levers. Organizations that widen the candidate pool see shorter fill times, and offering remote roles shortens time-to-fill by tapping talent in other regions - which, read from the seeker's side, means widening to remote adds the most live demand for the least damage to your positioning. Geography radius is the next lever for the same reason. Fill-time context helps you set expectations here: tech and media roles fill in about 20 days, while defense and investment banking average 60 or more, so a slow-filling sector needs a larger standing count to keep a pipeline fed.

Target: [role] + [geography] + [industry]
Date pulled: [date]
Raw postings: [n]  ->  Adjusted live (minus 60+ day, minus 25%): [n]
Distinct employers: [n]   Largest employer share: [%]
New postings <7 days this week: [n]   Index trend 6-12mo: [rising / flat / falling]
Weekly application target: [interviews/wk] / [rate] = [apps/wk]
Pay floor: [min]   Advertised bands clear it: [yes / no]
Verdict: [COMMIT / WIDEN / DROP]
If WIDEN, variable changed this round: [format / geo / title / industry]

Only widen titles and industry after format and geography are exhausted, and treat each widening as a fresh grade rather than a patch. A target that needs two widenings to pass is a signal in itself: the original aim was too thin, and you should decide deliberately whether the widened target is still the career you wanted.

Keeping the grade current

A viability grade decays, so date it and re-run the cheap parts monthly while a search is live. Demand velocity and the posting index trend move fastest; prevalence and your pay floor barely move, so you rarely re-check those.

The practical maintenance loop is short. Once a week, recount postings under 7 days old for your committed target - if that number trends toward zero for three weeks, the target is cooling regardless of its standing count, and you should re-grade fully. Once a month, re-check the posting index direction for your sector and geography, because it is revised and a flat reading can turn falling. Keep the verdict template from the previous section as a dated record; comparing this month's row to last month's is the fastest way to catch a target quietly dying under a healthy-looking headline count.

One last discipline. The moment you widen a target to keep it viable, write down what you changed and re-grade from scratch. The failure that costs the most is not picking a weak target - it is letting a target drift wider, one small concession at a time, until you are applying broadly with a message that fits nothing. The standard exists to make that drift visible.