The Ranked Target Company List, Built From Live Hiring Signals

This guide turns your role and area into a shortlist of companies that are actually hiring for you right now, ranked by which ones are worth an application. It is for a job seeker deciding where to aim who is tired of lists built from brands they admire rather than evidence they can act on. By the end you will have 20 to 40 named companies, each verified against observable hiring signals, ordered by expected payoff per application.

Most target-list advice builds the list forward from fit and aspiration: industry, culture, values, logos you would be proud to wear. That produces a list of places you wish worked at you, not places hiring you this month. This playbook runs the other way. It starts from live hiring evidence - posting recency, req cadence, growth-versus-backfill signals, freeze and layoff news - so every company on the final list is provably hiring, and the ranking predicts where your applications convert.

Why build the list backwards from hiring signals

Build the list from what companies are doing, not from what you admire, because a company you love that is not hiring for your role converts at zero. The aspirational list optimises for how good it feels to send the application; the evidence list optimises for whether anyone reads it.

The math forces the point. Roughly 1 in 5 postings on major platforms never leads to a hire. Greenhouse's own 2024 data classified 18 to 22 percent of its listings as possible ghost jobs, and a 2025 LinkedIn analysis put 27.4 percent of US ads and 24.9 percent of Canada ads in that bucket. Revelio Labs found hires per posting fell from about 8 per 10 ads to about 4 per 10. So when you apply blind, a meaningful share of your effort lands on reqs with no hiring behind them at all.

That is why verification is a payoff multiplier, not a nicety. Filtering out even a quarter of a list of dead reqs materially raises your expected conversion per application, because every hour you would have spent on a ghost gets redirected to a live one. The whole guide is one idea applied repeatedly: trust the observable signal, distrust the field a company can game.

An aspirational list optimises for how it feels to apply; an evidence list optimises for whether anyone reads it.

How big the list should be, and how geography changes that

Target 20 to 40 named companies, with 20 as a practical floor and roughly three dozen as a ceiling. That range spans the practitioner consensus: one coach sets 10 to 20, another 15 to 20, another 20 to 30 with a mix of top-tier and backup options. Tiering lets you honour all of them at once - a tight Tier 1 to work now, a wider set to monitor.

But the right raw net depends on how many companies employ your role where you are looking, and that varies more than most job seekers expect. In Refolk's index, "Product Manager" is a current title for 66,558 professionals in the US versus 15,140 in the UK - a 4.4x gap for the same title.

The lesson is not that the UK target should be smaller. It is that a list-size rule calibrated for a US market may demand a much wider net elsewhere to reach the same 20 companies. If your market is thin, expect to pull 150 candidate companies to land 30, not 100 to land 30.

Adjacent roles also share a market, and the split between them is often close enough to matter. In Refolk's index of US professionals, "Data Engineer" narrowly outnumbers "Data Scientist."

If your history could read as either, that near-even split is a reason to run the whole procedure twice, once per title variant, and merge the results. You are not choosing an identity; you are widening the pool of companies that could be hiring for a version of you.

The signals that prove a company is hiring

A company is genuinely hiring when its own careers page carries a recent, specific req that clusters with other hiring activity and is not contradicted by layoff news. Any one signal can lie; the case is built when two or three agree.

Here are the signals worth reading, what each proves, and what it looks like when it lies.

• Posting recency. A fresh req proves an open cycle. Most legitimate roles fill or disappear within 30 days, so act within 7 days, treat 30 as caution, and treat 45 to 60 as a red flag. It lies when the displayed date has been reset by a repost.
• Req cadence and clustering. Several similar roles posted together prove budget was approved at once. A company posting five sales engineering roles within two weeks signals compressed urgency and real headcount commitment, because simultaneous reqs need simultaneous budget that backfills never require. A single isolated posting proves far less.
• Growth context. Funding, leadership hires, and new-region launches within six months prove expansion intent. Read the rate against base size: 20 percent growth on a 40-person company is a real shift, while the same percentage on a 4,000-person company is ordinary churn.
• Posting specificity. A pay range, location details, shift or start timing, and a named hiring manager prove the req is tied to a real need. Mismatched requirements - "entry-level, 7+ years, PhD preferred, $45,000" - suggest the posting exists to justify not hiring anyone.
• The absence of contraction. No recent WARN filing or layoff news proves the company is not quietly frozen. WARN filings are public and give 60 days notice, so you can often see contraction before the careers page reflects a freeze.

The recency threshold deserves defending because it looks arbitrary and is not. The median role across industries closes in 36 to 44 days, based on SHRM's benchmark covering thousands of organisations. A posting past about 45 days has outlived one full hiring cycle, which is exactly why practitioner tools flag 45 days as stale. The threshold is one median, not a hunch.

Adjust for your role. Engineering runs 50 to 62 days, so a stricter 45-day cut would wrongly kill live technical reqs. Move the line to match one median cycle for the specific role you are chasing.

The procedure, start to finish

Run these eight steps in order. Steps 1 and 2 build the universe, steps 3 through 6 filter and score it, and steps 7 and 8 rank and confirm. Budget roughly a working day and a half for a first pass.

On step 8's placement: sources disagree on order. Some coaches verify with a human before ranking, on the logic that a confirmed req deserves to outrank an unconfirmed one. Most rank first and verify only the top tier, because messaging every candidate is expensive and you only need certainty where you are about to spend an application this week. I favour ranking first. Reserve your outreach budget for the companies your scores already say are worth it.

Generating the universe in step 2 and confirming reqs in step 8 are the slowest manual parts. This is where Refolk removes friction: instead of assembling a company list by hand from scattered funding news and careers pages, I can return companies that match your role, stage, and hiring signals in one query, so you start step 3 with a populated universe rather than a blank sheet.

Turning three scores into one ranking

Rank by expected payoff per application, which is the product of how likely the req is real, how strongly the company is growing, and how well the posting is written. A company that scores high on all three earns Tier 1; one that fails recency drops out no matter how much you admire it.

Keep the scoring simple enough to run across 60 companies without stalling. A workable rubric:

RECENCY (pick one)
  Fresh, under 30 days, true first-seen verified ......... 2
  Caution, 30 to 45 days ................................ 1
  Stale or repost, over 45 days or date reset ........... 0

GROWTH (pick one)
  Clustered reqs or funding within 6 months ............. 2
  Isolated backfill, no contraction signals ............. 1
  WARN filing or layoff news in the window .............. 0 (drop)

QUALITY (0 to 3)
  +1 pay range present
  +1 location and start timing specific
  +1 named hiring manager or human reply
  -1 mismatched requirements (cap at 0)

TIER
  Tier 1 (apply now):     total 6 to 7
  Tier 2 (apply + network): total 4 to 5
  Tier 3 (monitor):       total 2 to 3
  Drop:                   any GROWTH = 0, or total under 2

The matrix below is the fast version when you want to eyeball a company before scoring it in full. It plots hiring evidence against how well you fit, and tells you what to do in each corner.

How this goes wrong

The failure modes here all share a root: a signal that looks real but is not, or a real signal read in isolation. This section is the most valuable part of the playbook, because the whole method's payoff comes from what you correctly exclude.

Trusting the displayed date. A repost shows "3 days ago" but hides a months-old search, because every repost resets the visible date to zero. The false positive is a fresh-looking listing you rush at. Fix it by fingerprinting the role - company, title, location - and tracking when you first saw it, not the counter. If the same role reappears every few weeks with identical copy and a new date, the position has not been filled and may never be.

Aggregator lag. A role looks live on a big board but is already gone from the company's careers page. If you cannot find it on the company's own website, it may have been filled, frozen, or pulled while the board caught up. Always verify against Greenhouse, Lever, Workday, or Wellfound, not the aggregator alone.

Confusing hard-to-fill with fake. An old posting can be a genuine but stalled search, not a ghost. Past 45 days the explanation mix shifts toward hard-to-fill roles, stalled processes, evergreen pipelines, and ghost jobs, all at once. Check posting history and specificity, not age alone, before you write a company off.

Single-signal scoring. One data point misleads. The ghost case is strong when two or three signals appear together - stale date, generic template, no named human - and weak when only one does. Require the cluster.

Reading growth off raw headcount. A large company adding people may just be churning. Twenty percent on a 40-person company is expansion; the same on a 4,000-person company is noise. Always read rate against base size.

Missing a quiet freeze. A req can stay live after a hiring freeze simply because nobody pulled it down. Cross-check WARN filings and layoff news before ranking any company Tier 1. Because WARN gives 60 days public notice, the contraction is often visible before the careers page reflects it - an asymmetry that works in your favour.

Internal-candidate postings. A legally required external posting where the internal hire is already chosen looks perfectly real. Some jurisdictions now force the distinction: Ontario's Working for Workers Act requires employers with 25 or more staff to disclose whether a posting is an existing vacancy, and California passed a 2025 law requiring employers to disclose whether a posting reflects an active vacancy. Where no such disclosure exists, a named hiring manager and a human reply are your best tells.

Verify before you call the list done

Before you commit to the list, confirm that every tier survives its own test and that the top tier is human-confirmed. A list that passed the filters on paper but was never checked against a live human is a list of hypotheses, not targets.

Keeping the list current

A target list is a snapshot, and hiring signals decay fast, so treat the list as something you refresh weekly rather than build once. Top candidates are off the market within about 10 days, and reqs close near the 44-day median, so a list built a month ago is describing a market that has already moved.

Rerun the recency filter on Tier 1 every week and promote from Tier 3 as fresh reqs appear. Watch for two changes that force a re-tier: a new WARN filing or layoff announcement drops a company out immediately, and a new cluster of reqs plus funding can lift a Tier 3 monitor straight to Tier 1. When a company reposts the same role you already flagged, do not reset its clock in your own tracker - the search is older than it looks, and that is exactly what the date field is hiding.

The discipline is worth naming plainly. This method does not tell you where you would be happiest. It tells you where an application has the best chance of being read by someone actively hiring. Keep the two questions separate, run this one on evidence, and spend your applications where the signals say the door is open.