The Candidate Openness Score: Message Now, Nurture, or Skip

You have 200 profiles that all fit the role and no way to tell whom to message first. This guide is for in-house recruiters, sourcers, and founders doing their own hiring who need to sort a found list by one thing: how open each person is to a move right now, judged only from signals you can see from outside. It gives you a weighted rubric, act/nurture/skip thresholds, and, for every signal, a note on how it lies.

Most passive-sourcing writing stops at listing the signals. That list is not a decision. What follows turns the signals into weights anchored to real response rates, an overlay for company events that move whole cohorts at once, and a procedure you can run over a list in an afternoon.

What "openness" means here, and what you can actually see

Openness is the probability that a specific candidate will engage with outreach right now, estimated only from publicly observable signals. It is not the same as fit, and it is not the manager's private read on whether someone is about to quit.

That distinction sets the whole rubric. Behavioral tells - delayed email responses, low productivity, emotional outbursts - are visible almost only to the current manager, not to an external sourcer. You cannot score what you cannot see, and pretending otherwise inflates false confidence. So the rubric counts only four categories of signal that reach the open web:

• The Open-to-Work flag, in either its private recruiters-only mode or its public frame.
• Tenure measured against the candidate's own job history.
• A recent company event: acquisition, layoff, or stale funding.
• A recent profile edit.

The single most reliable openness signal is invisible to the candidate's own employer, which is exactly why it works. LinkedIn's Open-to-Work has a recruiters-only mode that is invisible to everyone except premium Recruiter license holders, while the profile looks normal to all other viewers. Because the candidate can raise their hand without risk to their current job, the flag carries real intent. In one practitioner's A/B data, profiles with it returned a 14.5% positive response rate against 4.6% for those without - roughly triple.

The weighted rubric: how each signal scores

Each signal earns a weight anchored to how much it moves response rates, not to how often the signal appears in blog posts. High-weight signals carry real intent or act on whole cohorts; low-weight ones are easily faked by routine behavior.

Here is the proposed weighting. It is derived from the response-rate and attrition ratios in the dossier, not an external standard, so treat the weights as a starting point and recalibrate against your own reply data.

To turn weights into numbers, use a simple point scale: High signal = 3 points, Medium = 2, Low = 1. Sum the points a candidate earns. A private Open-to-Work flag plus a recent acquisition is a 6; a lone profile refresh is a 1. The thresholds later in this guide read off that total.

Why tenure gets a medium weight, not a high one

Tenure predicts a move, but only in a window. Reachability clusters at roughly 18 months to 3 years of tenure, and again just past a 4-year vesting completion. The mechanism is equity: standard tech equity vests over 4 years with a 1-year cliff, 25% at month 12 and the rest monthly over the next 3 years, and compensation designers note the drop-off in vesting value in year 5 effectively signals to employees that this is the time to leave. The signal is strongest when the candidate's own history shows a 2-to-3-year cadence, because a pattern that has repeated tends to continue. It is medium, not high, because tenure alone tells you nothing about the person's current state of mind.

Company events: the overlay that scores a whole list at once

A company event beats individual behavior as a predictor because it acts on an entire cohort, so you can score hundreds of employees with one lookup. Acquisition is the strongest documented trigger.

The overlay: flag any candidate whose employer had a layoff, an acquisition, or a funding round that closed 18 to 24 months ago with nothing since. Add the high-weight points and record the window, because the window is when the candidate is most reachable.

Two figures do the heavy lifting. Acquired staff leave at 33% in year one against a 12% baseline for comparable regular hires, a 2.75x lift. And layoff victims are 65% more likely to quit the job that follows a layoff, so survivors and recent victims stay elevated well past the event. Funding staleness works as a proxy: signal-first recruiters treat a last round that closed 18 to 24 months ago as a reach-out trigger, on the logic that the company is either doing very well or is running low, and if the latter it is time to reach out.

Refolk lets you build the overlay as a query rather than a manual audit, so the cohort arrives already tagged with the event.

The channel decision: rate versus math

Choose the channel by expected reply rate, then reconcile it against how many contacts the channel can actually supply. Warm intros win on rate and lose on math.

Warm introductions reply best at 21 to 34%, but the average rep can secure only 5 to 8 quality warm introductions per month. Against a 200-profile list that covers under 5% of it. So route intros to your very top act candidates and send everyone else a signal-triggered, personalized InMail at 18 to 25%. Reserve generic cold email for last; at 1 to 5% it barely clears the noise, and roughly 16.9% of cold emails never reach the inbox at all.

Warm intros win on rate but lose on math, so most of a 200-person list has to move on signal-triggered InMail.

The procedure: from raw list to sorted buckets

Run these seven steps in order. A sourcer can complete steps one and two on 200 profiles in roughly half an hour plus a minute or two per profile; a lead handles the scoring and bucketing.

Practitioners disagree on order. Signal-first sourcers score the signal before pulling the list at all, building the search around a trigger such as a stale funding round, rather than scoring a list after the fact. Both work; the signal-first path saves the de-duplication grind when your tool can query the trigger directly.

Reading the thresholds

• Act (score 4 or more, or any high-weight signal): message now. These carry direct intent or sit inside a live event window.
• Nurture (score 2 to 3): park with a re-check date tied to a vesting anniversary or funding milestone. The signal is real but early.
• Skip (score 0 to 1, or only a lone profile refresh): do not spend a personalized touch. Revisit only if a new event lands.

Name | Employer | Tenure | Open-to-Work (private/public/none) | Company event + window | Profile edit (headline change/cosmetic/none) | Score | Bucket | Top 2 reasons | Channel | Re-check date

How this goes wrong: false positives and where each signal lies

Every signal in the rubric can mislead, and the failures cluster into predictable patterns. This is the part of the standard worth reading twice, because a wrong score sends a personalized touch to someone who will never move and buries someone who would.

Stale Open-to-Work banner. LinkedIn automatically turns off the status if the person does not respond to recruiter messages within 30 days, so a banner that is still showing may sit on a disengaged profile. Before scoring it high, check for recent activity. A banner switched on out of momentary anxiety is a false positive too.

Counter-offer leverage. A candidate may be interviewing only to extract a raise, using the process and any offer to force a counter from their current employer. They look open and will not move. A very recent internal promotion or comp cycle is the tell.

Profile refresh mistaken for intent. Guidance everywhere tells people to keep their profile current, so a refresh can be routine maintenance. Check whether the headline or summary actually changed versus a cosmetic edit before you read job-search prep into it.

Short tenure read as openness. A short stint may be a genuine cadence or simply a red flag the next employer dislikes, not evidence of current intent. Short tenure should not automatically be treated as a signal either way. Read the full history for a repeating pattern.

Data-lag illusion. People update their profiles 2 weeks to 2 months after starting a new role, and some never do. The freshest-looking profile is often the least open, and an old-looking employer may already reflect a move. Cross-check against company news and mutual connections.

Company-event over-weighting. Not everyone at an acquired firm leaves; retention packages lock in key staff. And attrition is uneven by role: recruiters ran a 26% post-acquisition attrition rate and acquired senior management near 20% at 18 months, while engineering tends to hold better. Weight the event by role type.

Before you send: the pre-outreach checklist

Run this before any candidate leaves the act bucket for an actual message. It catches the false positives that survive the score.

Keeping the score current

An openness score decays, because the signals underneath it decay. The Open-to-Work flag auto-expires after 30 days of no reply; the acquisition window closes as retention packages take hold; vesting milestones pass. Treat the score as a snapshot with a shelf life, not a fixed label.

Two habits keep it honest. First, re-check nurture candidates on their trigger date, not on a calendar cadence, so you re-score exactly when a signal is likely to have changed - a vesting anniversary, a funding milestone, or the next round of company news. Second, recalibrate the weights against your own reply data every few months. The dossier's ratios anchor the starting weights, but your reply rates by signal are the real ground truth. In Refolk's index of professional profiles there are 613,118 US professionals listing Python as a skill, which is a reminder that the constraint is never supply; it is knowing which of them to message first. If a signal in your data stops correlating with replies, drop its weight. A rubric you never recheck is just an old blog post with numbers on it.

number: 613,118
label: US professionals in Refolk's index listing Python as a skill
note: Supply is rarely the bottleneck; the openness score is what turns a large pool into a ranked first-message list.