The Account Hiring-Signal Read: Pitch Now, Time the Hire, Backfill, or Ghost
Score one account's live postings on seven dimensions in about ten minutes and defend a verdict, a named contact, and a dated next action to your manager.
Key takeaways
- In a study of 1 million B2B software purchases, companies with a 20%+ headcount increase averaged 38% more purchases, while a 20%+ rise in job openings came with only 7% more, so headcount should gate the read rather than break ties.
- Judge posting age against the employer's own median time to fill: treat a role as a likely ghost when it stays open past three times that median (never under 90 days) and has been reposted, or has been reposted three or more times.
- Junior careers-site postings carry roughly double the ghost share of senior ones (20.8% against 10.5%), so volume-level reqs deserve a heavier age penalty.
- Normalise cluster velocity as openings in a function divided by that department's current headcount; a ratio above about 0.2, with at least 2 to 3 openings, marks a genuine build-out at any company size.
- In Refolk's index, US Revenue Operations leaders (2,054 profiles) outnumber US RevOps individual contributors (1,925), so a first RevOps IC posting under an existing leader can be a real first-in-function build.
- A competitor named as required experience in a posting signals an entrenched user, not an open evaluation; only 'migrate', 'replace' or 'consolidate' next to the name signals switching.
This is the read I run on one target account when its live job postings look related to what I sell. It answers three questions: is this a real buying window, who do I contact, and when. It is for founders, account executives, SDR leads and partnerships teams who have to explain why an account got a sequence. You score seven dimensions from public evidence in about ten minutes and land on one of four verdicts, each with a named contact and a dated next action.
What does a job posting prove about a buying window?
A job posting proves that a company intends to hire. It doesn't prove the hire happened, that budget exists, or that anyone will buy what you sell. So this read uses postings to decide which accounts to look at, and uses headcount to make the call.
The clearest published comparison comes from a study of 1 million B2B software purchases. It counted team purchases only (not individual subscriptions) and limited the sample to companies with 200 to 1,000 employees. It then compared average purchases for companies with and without each signal:
| Signal | Avg purchases with | Avg purchases without | Lift |
|---|---|---|---|
| Headcount up 20%+ | 3.01 | 2.18 | +38% |
| Job openings up 20%+ | 2.82 | 2.64 | +7% |
| VP hire | n/a | n/a | +28% |
The study's author explains the gap: "job postings are just intent to hire, and there's a massive lag before software is needed."
Know three limits before you lean on these numbers:
- It compares group averages. It is not a correlation coefficient, even though some pages repeat it as a "+38% correlation".
- The threshold gets restated. Some pages quote it as "10%+ in 90 days". The original used 20%. Check which version a tool's alert is built on.
- Stacking is untested. Nobody has published a figure for a posting combined with a second signal. Stacking is sensible, but its lift is not established.
That doesn't make postings noise. A Management Science paper found that rising job postings are positively associated with future growth in headcount, sales and earnings. Postings are early, not wrong. The read works out how early, and for whom.
The evidence ladder for one account
- Posting textTitles, build language, named tools. Cheapest to read, easiest to misread.
- Careers-page date and repost historyWhether the req is live, recycled, or evergreen.
- Normalised clusterWhether one req is part of a build-out sized to the team.
- Headcount trendWhether hires actually closed. This layer gates the read.
The seven dimensions and how to score them
Score each dimension 0, 1 or 2 from public evidence. Three dimensions are gates: a 0 on any of them ends the read with a verdict. The other four measure strength, decide who you contact, and rank the accounts that get through the gates.
A buying seat is the role that owns the purchase decision for your product. For a RevOps tool, that is usually the Head or VP of Revenue Operations.
The three gates
| Dimension | Scores 2 when | Scores 0 when | How it lies |
|---|---|---|---|
| Posting age and reposts | Careers-page date within the employer's normal fill time, no reposts | Open past 3x the employer's median fill time (90-day floor) plus a repost, or 3+ reposts, or general-application wording | An aggregator re-index makes an old post look fresh |
| Net-new vs backfill | No matching departure, build or first language | Same or adjacent title left in the last 90 days, flat headcount | A relevelled backfill looks new |
| Headcount corroboration | Function headcount growing over the quarter | Repeated postings across a quarter, flat headcount | Too early for a first hire to show yet |
A 1 on a gate means "not proven either way". Headcount scores 1 more often than anything else, because a single new seat takes weeks to show up on profiles. A 1 lets the read continue. Write down why it is a 1.
The four strength dimensions
| Dimension | Scores 2 when | Scores 1 when | How it lies |
|---|---|---|---|
| Role proximity | Posting is the buying seat or reports into it | Adjacent function that uses your product | Title inflation at small firms |
| First-in-function | Nobody holds the function today | Function exists but is thin | A "Head of" who is the only person in the function |
| Cluster velocity | Openings divided by department headcount above about 0.2, with 2 to 3+ openings | One or two openings | Percentage growth from a tiny base |
| Stack language | "Migrate from", "replace", "implement", "build out" | A tool named only as required experience | A named competitor read as an evaluation |
Role proximity is also the entry condition. If none of the account's postings scores at least 1, the postings aren't about your product. Stop and don't log a verdict.
How to use the strength total (0 to 8): use it to rank accounts that passed the gates against each other, not to make a pass/fail call. I have no outcome data that supports a fixed cutoff, and I don't know of a published one. Ranking holds up when your manager asks. A made-up threshold doesn't.
How old is too old for a job posting?
How old is too old depends on the employer. A posting is suspect if it has been open past three times that employer's median time to fill (never less than 90 days) and has also been reposted, or if the same role has been reposted three or more times. Universal cutoffs of 45, 60 and 90 days all circulate, they disagree with each other, and none has been validated against whether the role was actually filled.
Benchmarks from published indexes:
- Typical age: one report on careers-site postings puts the median open posting at 35 days old. The same report found 28.0% of 668,355 careers-site postings had been up for more than 90 days.
- Tail age: an index of active postings found 9.5% visible for 45+ days, 6.8% that old and also reposted, and 3.5% carrying two or more strong ghost signals.
- Ghost share: Greenhouse found that 18% to 22% of job posts are ghost listings.
- Fill time: SHRM has put average time to fill at 41 days. Re-check this figure, because it is republished periodically.
- Population matters: one regional workforce board reported a median posting duration of just 11 days, against 35 days on careers sites. Compare an account against benchmarks from the same kind of source.
The median careers-site posting (34 to 35 days) is younger than most practitioner ghost cutoffs (45 to 60 days). A 45-day rule only catches the tail. That is why the gate uses the employer's own baseline.
| Seniority | Open over 90 days | Ghost share | Median age (days) |
|---|---|---|---|
| All platforms | 26.5% | 9.3% | 34 |
| Junior | 40.2% | 20.8% | 58 |
| Mid-level | 37.8% | 11.9% | 60 |
| Senior | 36.5% | 10.5% | 57 |
| Lead | 36.7% | 11.3% | 60 |
Source: a public ghost-jobs index of careers-site postings.
Junior postings have about twice the ghost share of senior ones. My guess at the reason is that evergreen pipeline reqs cluster at volume levels. Either way, a junior SDR post deserves a heavier age penalty than a director post. Another index found that seniority moves the numbers less than the job board or employer size does. The share of long-open postings rises steadily with employer size, and at companies under 50 employees 8.7% of careers-site postings had been up 45+ days. Some boards run long-open by default: SmartRecruiters (61.6%), Y Combinator (49.6%) and Workday (41.7%) led on the 45+ day share. Note which board the posting lives on before you read anything into its age.
Evergreen ads are rarer than people think. A count of 10,630 companies' Greenhouse, Lever and Ashby boards found 324,634 open jobs, of which 2,178 (0.7%) were general-application ads. They are rare, but they are the easiest discard you'll ever make.
Is it a backfill, net-new, or first-in-function seat?
A backfill replaces someone who left and inherits their tools. A net-new seat adds headcount to a function that already exists. A first-in-function seat creates a function that didn't exist before. Only the last two open a buying window, and first-in-function opens the widest one, because the seat has no stack to inherit.
The cues practitioners use:
- Backfill: one posting with the same title as someone who just left, and flat headcount.
- Net-new: "build out" or "build from scratch" language, more than one opening, and rising headcount.
- First-in-function: no prior leader of the function anywhere in the company's history, and nobody holds the title today.
There is one trap from the HR side. A backfill that changes level, location or pay range is usually treated as a new role. For selling, treat it as a backfill unless the posting also says migrate or replace: the people changed, but the tools stayed. Nobody has published how often each cue misleads, so record your misses as you go.
Before you score first-in-function, know how the market is shaped. These counts come from Refolk's index of professional profiles:
| Slice (Refolk's index) | Profiles |
|---|---|
| US RevOps leaders (VP, Head, Director) | 2,054 |
| UK RevOps leaders (same titles) | 226 |
| US RevOps ICs (manager, analyst, specialist) | 1,925 |
In the US, RevOps leaders outnumber the ICs: 0.94 ICs per leader across the market. So many RevOps leaders have nobody under them. That means a first RevOps Manager posting under an existing Head of RevOps can be a genuine first-in-function build of the team, not a backfill. It also means a lone "Head of" title at a small company can overstate the budget.
Titles differ by market too. In a small 25-profile sample from the same index, the most common US title was "Vice President of Revenue Operations" (11) and the most common UK title was "Head of Revenue Operations" (12). The sample is small, so treat this as a reminder to search both forms in each market, not as a distribution.
The departure check and the headcount-under-the-leader check are both people-search questions. Careers pages can't answer them. This is where Refolk saves time: I ask in plain English for the people who hold the function at the account, and I get the profiles back, so I can see who left and who is still there without stitching searches together.
How do cluster velocity and stack language change the read?
Cluster velocity tells you whether one req is part of a build-out sized to the team. Stack language tells you whether you are displacing an incumbent tool or filling an empty seat. Both raise the strength score. Neither replaces the gates.
Cluster velocity
Cluster velocity is how many openings a function posts in a window, usually 30 days. Published cutoffs vary:
- 1 to 2 postings reads as backfill.
- 10+ SDR postings in 30 days reads as a full outbound build-out.
- Five roles within two weeks is another rule of thumb.
None of these has been validated against purchases, and fixed counts break across company sizes. A 30-person company with 3 open roles is at about 10 openings per 100 employees. A 5,000-person company with 25 open roles is at 0.5. Normalise instead: divide openings in the function by that department's current headcount. A ratio above about 0.2 indicates a genuine build-out at any company size, provided there are at least 2 to 3 openings in absolute terms.
If you sell into smaller markets, the ratio is the only fair comparison. Three RevOps openings at a UK company can mean more than ten at a US company of similar size.
Stack language
Read the posting text the way a buyer wrote it:
| Phrase in the posting | What it usually means | Score |
|---|---|---|
| Tool named in responsibilities | Current stack, in use today | 1 |
| Competitor named as required experience | Entrenched user, not switching | 1 |
| "Implement" or "build out" | Upcoming purchase or greenfield build | 2 |
| "Migrate from" or "replace" a named tool | Switching intent, and rare | 2 |
One sales-intelligence writer calls "migrate from" a competitor "the highest-intent signal you will ever find in a job posting", and also notes that it is rare. No published correlation links any of these phrases to an open evaluation. Quote the exact words in your scorecard so someone else can check how you scored them.
The four verdicts: who to contact, and when
The gates decide whether a window exists. Role proximity decides who owns it. If the posted role reports into the buying seat, pitch the hiring manager now. If the posted role is the buying seat, wait for the new hire and time your outreach to their first 60 days.
| Verdict | When it applies | Named contact | Dated next action |
|---|---|---|---|
| Pitch the Hiring Manager Now | Gates pass; role reports into the buying seat | The leader the role reports to | First touch this week, while the req is live |
| Time to the New Hire's First 60 Days | Gates pass; role is the buying seat | The incoming hire, once named | Check whether filled at day 35; first touch at start date plus 15 days |
| Backfill: No Window | Net-new gate scores 0 | Current seat holder, logged only | Rescore 90 days out |
| Ghost or Evergreen: Discard | Age gate or headcount gate scores 0 | None | Rescore only when a new, distinct posting appears |
The day-35 check matches the median age of a careers-site posting. Around then, look for whether the req has come down and someone new holds the title.
Where the account lands
Why the timing splits this way
Published views on the window disagree:
- One argues the best window is before the new hire starts.
- Another says director and manager hires make tool decisions in their first 60 to 90 days.
- A third puts first decisions at around 90 to 120 days.
- A fourth says new leaders audit the stack within 30 to 60 days, and buying is strongest 30 to 90 days after they join.
These views fit together once you ask who will sign. When the posted role reports into the buying seat, the hiring manager is building the team and its tooling right now, so the before-the-start argument applies. When the posted role is the buying seat, the decision waits for that person. The same purchase study found VP hires came with 28% more purchases, which supports waiting for leadership hires rather than pitching around them.
Timing a buying-seat hire
- Req liveLog the careers-page date and check around day 35 whether the seat is filled.
- Hire namedFind the person and confirm their start date from their profile.
- Days 1 to 14Map the mandate from the posting text, the team, and what they publish.
- Days 15 to 30Test one hypothesis with a short, question-led touch.
- Days 31 to 60Run the full sequence only when a supporting signal appears.
Finding who started is the step where reps lose the thread, because the careers page goes quiet once the req closes. Ask Refolk for the person by title, company type and start window, for example heads or VPs of Revenue Operations who joined a US B2B SaaS company in the last 60 days. You don't have to keep refreshing a page that no longer exists.
A posting is a sentence about the future; headcount is the receipt.
The ten-minute procedure
Run these seven steps in order on one account. Each step produces a field on the scorecard. The order follows the evidence ladder: confirm the req is live before you analyse what it says.
The account hiring-signal read
- Verify the first-published date on the careers pageList the postings that sit in or under the buying seat, score proximity, and log each first-published date from the employer's careers page, not the aggregator.
- Check repost history and evergreen wordingCount reposts of the same role, flag general-application wording, and score the age gate against the employer's own median time to fill.
- Label the seat: backfill, net-new, or first-in-functionCheck for a same or adjacent-title departure in the last 90 days, whether anyone holds the function, and any build or first language.
- Count the 30-day cluster and normalise itDivide openings in the function by the department's current headcount and confirm at least 2 to 3 openings in absolute terms.
- Parse the stack languageQuote every named tool and every implement, build out, migrate from or replace phrase, and label each one incumbent, migration, or greenfield.
- Corroborate with the headcount trendCompare the function's headcount over the quarter with the postings, and score it as growing, too early to see, or flat.
- Pick the verdict, the contact, and the dateApply the gates in order, use proximity to choose between pitching now and timing to the hire, and write one name and one dated action.
Where sources disagree on the order: some practitioners keep a watch list from week 0 and confirm the cluster in week 2 before parsing anything. That works if you watch many accounts. For a decision on one account, the order above gets you to a verdict in a single pass. My estimate is about two minutes for the date check, the slowest step at first. That figure is mine and unverified.
ACCOUNT HIRING-SIGNAL READ Account: Read date: Postings in scope (title :: careers-page first-published date :: repost count :: board) - - GATES (a 0 ends the read) Posting age and reposts (0/1/2): evidence: Net-new vs backfill (0/1/2): evidence: Headcount corroboration (0/1/2): evidence: STRENGTH (ranks accounts that pass) Role proximity (0/1/2): seat or reports-into-seat: First-in-function (0/1/2): who holds the function today: Cluster velocity (0/1/2): openings / dept headcount = Stack language (0/1/2): phrases quoted: Strength total (0-8): VERDICT: Pitch the Hiring Manager Now | Time to the New Hire's First 60 Days | Backfill: No Window | Ghost or Evergreen: Discard Named contact: Next action: Next action date: Rescore trigger:
One per account. Quote evidence verbatim so a manager can re-run the read from the scorecard alone.
Subject: the {role} you are hiring
Hi {first name},
I saw the {role} opening on your careers page, posted {date}. It reads like {function} is being built out rather than backfilled ({one quoted phrase from the posting}).
When a team adds a {role} under {their title}, the question I usually hear first is {the one problem your product solves}. Is that on your list before {role} starts, or after?
{your name}Use only when the posted role reports into this person; swap the bracketed fields and cut anything you cannot evidence.
How does the hiring-signal read go wrong?
Most bad sequences come from a signal that looked real and was misread. Each failure below has a cheap check, and running all seven costs less than one wasted sequence.
- Age read from an aggregator. A repost or refresh resets the date, so a ghost looks fresh. Check: compare the aggregator date with the careers-page date. If they differ, the careers page wins, and a gap is itself evidence of reposting.
- A fixed age cutoff applied to an enterprise. The share of long-open postings rises with employer size, so 60 days at a 5,000-person company doesn't prove a ghost. Check: compare against that employer's own median time to fill, and note the board. Some boards run long by default.
- A changed-role backfill scored as net-new. A relevelled or relocated seat looks new, but the incumbent stack stays. Check: look for a departure in the last 90 days with an adjacent title, not just the same one. If one exists, score the seat as backfill unless the posting says migrate or replace.
- A cluster counted without normalising. Going from one open role to three is 200% growth and means almost nothing. Check: divide by department headcount and apply the 2 to 3 opening floor before you call it a build-out.
- A named competitor read as an open evaluation. "Required experience" usually means an entrenched user. Check: only "migrate", "replace" or "consolidate" next to the name counts as switching intent. Everything else scores 1.
- Postings with flat headcount. If a company posted eight roles last quarter and its headcount is flat, disregard the postings entirely. They are churn, a pipeline, or a plan that stalled. Check: the headcount gate scores 0 and the verdict is Ghost or Evergreen: Discard, however good the posting text reads.
- Title inflation in Refolk's index. A "Head of RevOps" at a 20-person firm may be the only RevOps person there, with a title bigger than the budget. Check: count the ICs under the leader. Since US RevOps leaders outnumber ICs in Refolk's index, zero is common and doesn't make the seat fake, but it should temper your deal-size expectations.
There are two quieter failures to watch for. First, vendor pages restate the same study with different thresholds (10% in 90 days against the original 20%), so an alert can fire on a weaker condition than you think. Second, no published source gives a miss rate for any backfill cue or stack phrase. If your manager asks how accurate the rubric is, the honest answer is: "accurate enough to rank, not proven enough to threshold, and here is my log".
Keeping the read current
A verdict expires. Rescore an account when a scheduled trigger fires, not when you happen to remember it, and re-check the benchmarks this guide relies on, because the indexes behind them are republished.
Rescore triggers by verdict:
- Pitch the Hiring Manager Now: rescore when the req comes down or a second posting in the function appears.
- Time to the New Hire's First 60 Days: rescore around day 35 from posting, then at the start date.
- Backfill: No Window: rescore 90 days out, or sooner if the role is reposted with migrate or replace language.
- Ghost or Evergreen: Discard: rescore only when a new, distinct posting appears or headcount moves.
Benchmarks to re-check periodically:
- The median careers-site posting age and the share open past 90 days, from the ghost-job indexes. These set your sense of normal.
- Average time to fill. This sets the floor under the three-times rule.
- Your own market's buyer-seat supply. The RevOps counts in this guide come from Refolk's index and will move. If you sell into a different function, rebuild the supply table for that function before borrowing anyone's cluster threshold.
Keep every scorecard. After a quarter, sort them by verdict and see which ones produced meetings. That log will tell you more about timing in your market than any published window.
Before the account gets a sequence
- Every in-scope posting has a first-published date taken from the employer's careers page.
- Repost count, job board, and any general-application wording are logged for each posting.
- Posting age was judged against the employer's own median time to fill, with a 90-day floor.
- The departure check covered the last 90 days and included adjacent titles, not only exact matches.
- Cluster velocity is recorded as a ratio to department headcount and clears 2 to 3 openings.
- Every stack phrase is quoted verbatim and labelled incumbent, migration, or greenfield.
- The headcount trend is recorded, or marked too early with a written reason.
- The scorecard names one verdict, one person, one next action, and one date.
Questions practitioners ask
Are job postings a good buying signal for outbound?
On their own, job postings are a weak buying signal, and they become useful once something else backs them up. In a study of 1 million B2B software purchases, a 20%+ rise in job openings came with only 7% more purchases, while a 20%+ headcount increase came with 38% more. Use postings to choose which accounts to investigate. Then check the careers-page date, the repost history and the headcount trend before you commit a sequence.
How do I tell if a job posting is backfill or new headcount?
Look for a departure in the last 90 days with the same or an adjacent title. One posting matching someone who just left, plus flat headcount, reads as backfill. 'Build from scratch' language, rising headcount and no prior holder of the function point to net-new. A backfill that changes level, location or pay is often treated as a new role by HR. For selling, it still inherits the old stack unless the posting says migrate or replace.
How can I tell if a job posting is a ghost job?
Read the date on the employer's careers page, not on an aggregator, because reposts and refreshes can reset dates. A posting is suspect when it has been open past three times the employer's own median time to fill (never under 90 days) and has been reposted, or when the same role has been reposted three or more times. General-application wording is an automatic discard. Universal 45 or 60 day cutoffs have not been validated against hiring outcomes.
When should I reach out after a new hire starts?
Published views disagree. Different sources put the window at 30 to 60 days for a stack audit, 60 to 90 days for tool decisions, and 90 to 120 days for first decisions. I use days 1 to 14 to map the mandate, days 15 to 30 for one short, question-led touch, and days 31 to 60 for a full sequence only if a supporting signal appears. Nobody has published outcome data that settles the window, so record your own results.
Should I contact the hiring manager or wait for the new hire?
It depends on who will own the purchase. If the posted role sits under the person who buys your product, pitch that hiring manager while the req is live, because they are building the team and its tooling now. If the posted role is the buying seat itself, wait for the new hire, who will make the call. No published outcome data compares the two approaches.
Does a first hire in a department mean the company will buy?
A first hire in a department makes a purchase more likely but does not guarantee one. A first-in-function hire creates a seat with no inherited stack, so there is often a greenfield decision to make. In one study, VP hires came with 28% more software purchases. Check for title inflation, though: at a small company, a 'Head of' may be the only person in the function and may have little budget.
Try it on the search you came here for
Stop building boolean strings. Just describe the person.
Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- No boolean, no filters, no seat to buy. One box.
- Read at search time, so a profile updated yesterday counts today.
- Every step visible as it runs, every name with its reason.
500 free credits on sign-up. No card, no demo call. See real searches.