The Source-Company Archetype Reference for Sourcing

You are building the list of companies to mine for candidates, and before you spend hours inside any one of them you need to know what type of company it is and what the people inside are worth to your search. This is a lookup document for in-house recruiters, sourcers, and founders doing their own hiring. Jump to the archetype that matches a row on your list, read what it proves about caliber and reachability, and read the specific way it will fool you.

The premise is simple: a company's type predicts its people. A direct competitor and a freshly-acquired startup both put engineers on your list, but they hold different calibers, open at different moments, and mislead you in different directions. Naming the archetype turns a flat list of logos into a ranked, timed sourcing plan.

Why archetype comes before mining

Classifying a source company first tells you the caliber inside, the moment it is reachable, and its built-in trap, so you spend your hours on the right rows. Sourcing pays off before you touch a single logo: sourced candidates are 5x more likely to be hired than inbound applicants, drawn from a benchmark of over 140 million applications. But that leverage evaporates if you mine the wrong companies in the wrong order.

The cost of getting the list wrong is compounding. US median time-to-fill for nonexecutive roles now sits at 44 days, the slowest in SHRM's tracking history, and interviews per hire have climbed to about 20, up from 14, a 42% jump. Every hour spent mining a prestige logo whose best people already left is an hour that pushes those numbers the wrong way. Archetype classification is the cheapest filter you have: 30 to 60 minutes across a whole list, done once, before any deep mining.

The six source-company archetypes

There are six archetypes worth naming, and each carries a distinct caliber profile, a reachability window, and a characteristic lie. Read the row that matches your company, then verify with the trigger-event and pool-depth sections below.

The two archetypes people most often get wrong are the prestige feeder and the post-mandate outflow. The prestige feeder is where the logo does your thinking for you. The post-mandate outflow is where the movement is real but too slow to notice without a date on it. Both are handled in the failure-modes section, because both cost real hours.

Direct competitor

A direct competitor holds people already doing your exact role in your exact market, which makes it the highest-precision archetype on the list. The skills match by definition, and the candidates understand the space without onboarding. The reachability is steady and jumps whenever the competitor hits public turbulence: a reorg, a missed round, a return-to-office mandate. The trap here is legal, not qualitative, and it is covered in its own section: mining a rival is lawful, agreeing with a rival not to mine each other is what became criminal.

Prestige feeder

A prestige feeder is a famous logo you put on the list because the name signals quality. The name proves the person once cleared a hard bar, nothing about their current caliber or whether they are still there. This archetype is reachable only when its attrition trend says so, which means the logo alone is never enough. You verify a prestige feeder by its hire-to-departure ratio, not its reputation.

Freshly-acquired startup

A freshly-acquired startup holds high-caliber builders who just lost the culture they signed up for, and it runs on two clocks. Individual contributors react fast: within the first year of an acquisition, 33% of acquired workers leave, against 12% of comparable regular hires. Managers wait out earn-outs and bleed slower: acquired firms lose 4 out of 10 managers within the first 24 months, three times the non-merger rate. So you mine engineers in the first year and keep leadership on the list for two full years.

Declining incumbent

A declining incumbent is shrinking headcount, which concentrates deep experience among people whose skills may not all be current. It grows more reachable as the shrink deepens, but the shrink direction alone does not tell you who is leaving. You need the attrition detail behind the trajectory, because a company can shed headcount by freezing hiring while its best people stay put.

Post-mandate outflow

A post-mandate outflow is a company that issued a return-to-office mandate and is now leaking talent on a slow schedule. RTO mandates raise turnover by 13% and time-to-fill by 23%, and the effect runs for months, not weeks. The highest-option employees leave first, which means this archetype yields disproportionately senior and skilled candidates. It misleads by being slow: over any single week the movement looks like noise, so you have to read it against the mandate date.

Adjacent-industry pool

An adjacent-industry pool holds people with transferable skills outside your obvious competitor set, and it is systematically undercounted by profile-based data. This is the archetype the base-rate math loves and the tooling hides, because self-reported profiles skew toward tech and white-collar roles. It is reachable at a steady rate and does not depend on trigger events, so it is the stable backbone of a source list when the triggered companies dry up.

Trigger events: what precedes departures and for how long

Four events reliably precede departures, and each opens a dated reachability window you can sequence outreach against. This is what turns an archetype tag into a calendar. The window is the difference between reaching someone while they are open and reaching them a year late.

The WARN filing is the cleanest signal on the list because it is a legally required, dated primary source. Federal WARN obliges employers with 100 or more employees to give 60 days' written notice before a covered mass layoff or plant closing, so a filing hands you a roughly 60-day lead before people actually separate. That is a window where they know the end date and you know it too.

The acquisition signal is the one most people mistime. The 33% first-year figure is an average across individual contributors, and it does not describe managers at all. Read the announcement date against the role level: engineers are a first-year play, leadership stays on your list for two years.

Sizing the pool: rank by skill depth, not company size

Rank source companies by the depth of the skill-filtered pool they hold, not by their headcount, because scarce skills cluster in a few firms that raw size will bury. This is the single correction that changes a source list most. A giant with a shallow pool of your skill is a worse feeder than a small shop that concentrates it.

The numbers make the point. In Refolk's index of professional profiles, US Software Engineers listing Kubernetes number 13,596, while those listing Rust number 614, a gap of about 22x. The Rust engineers cluster in a handful of firms - Google, Oxide Computer, Meta, Citadel - so ranking by company headcount would push the small, dense sources to the bottom of your list. Pool depth, not company size, predicts feeder value.

Geography multiplies the base rate on top of skill. The same title holds vastly different pool depth by country, which means the "prestige feeder" archetype is always geography-relative. A mid-tier US firm can out-supply a top-tier German one purely on base rate.

This is where knowing an archetype is not enough on its own: you also need real counts filtered by the skill and geography that matter, so a small dense feeder rises above a large shallow one.

Once you can size pools by skill and geography, the archetype tag and the trigger window combine into a ranked, dated source list. Refolk supplies the raw counts by title, skill, and geography that make step six of the procedure a 15-minute query rather than a manual census.

The procedure: from raw list to sequenced outreach

Work the source list in eight steps, from a flat list of names to an outreach calendar aligned to each company's reachability window. Two ordering notes are worth stating up front: some practitioners rank feeders before classifying archetypes, and the legal check is placed last here for workflow reasons even though compliance guidance implies it should gate the entire program. Move the legal check earlier if your risk posture calls for it.

The verification step relies on how talent-flow data is built. Refreshed daily, growth metrics compare the current employee count to the count 12 full months prior, and the talent-flow view shows hires and departures over the last 6, 12, or 24 months plus the current month, with a hire-to-departure ratio. That ratio is what you rank on in step four.

How the archetypes fool you

Every archetype has a characteristic false positive, and the failure modes below are the most valuable part of this reference because they are where hours get wasted. Read the one that matches the archetype you just tagged.

Passive-share inflation. The clean "70% passive" figure hides a documented spread from roughly 63% to 90% across surveys, years, and geographies. LinkedIn's 2014 study of 18,000 professionals found about 75% passive, and between 78% and 90% "approachable" per market; the active share rose from 25% in 2013 to 36% in 2016; Workable's Great Discontent survey found 37.3% passively open. Using one number as universal misstates reachability. Check the specific survey, year, and geography before you plan volume.

Prestige-feeder halo. A famous logo proves brand, not current caliber, and the best people may have left after an RTO or an acquisition. This is the prestige-feeder archetype's built-in lie. Check the hire-to-departure ratio and the attrition trend, not the name on the door.

Acquisition timing miss. The 33% first-year figure is an average, and the false positive is assuming everyone is reachable immediately at close when the 24-month manager window runs much longer. Check the announcement date against the role level before you decide someone is in play.

WARN false negative. Federal WARN exempts firms under 100 employees and has state variations, so the absence of a filing does not mean no layoff. New York requires 90 days' notice; California's threshold sits at 75 employees. Check state mini-WARN registries before you conclude a declining incumbent is stable.

Talent Insights single-source bias. Profile-based data covers only the LinkedIn member population, which is not a full representation of the labor market, especially in niche industries or non-traditional hiring. Independent researchers found a 57-million-member gap between the product and LinkedIn Recruiter for India alone. This is why the adjacent-industry pool looks thinner than it is. Check against a second count before ranking feeders.

No-poach self-sabotage. Treating a source company as "off-limits" through an informal understanding with a peer employer can itself be the illegal per se agreement. This inverts intuition: mining a competitor unilaterally is safe, agreeing with a competitor not to mine each other is the danger.

Skill-pool scarcity misread. A 22x gap between Kubernetes and Rust means a small niche company may still be your best feeder for the scarce skill. Ranking by raw headcount buries scarce-skill sources. Check skill-filtered counts, not company size.

The legal boundary: what is safe and what is not

Poaching individuals from a competitor is lawful; agreements between employers not to poach are the criminal risk, and the enforced danger is the opposite of what most people fear. The 2016 DOJ/FTC Antitrust Guidance for Human Resource Professionals warned that agreements between competitors to fix wages or refrain from soliciting each other's employees could bring criminal prosecution. These are treated as per se illegal - automatically prohibited, with no chance to argue a business justification.

The exposure is severe on paper. Corporate antitrust crimes carry a maximum penalty of up to $100 million or twice the gain or loss caused, whichever is greater, and the 2010 Silicon Valley "no cold-call" case involving Adobe, Apple, Google, Intel, Pixar, and Lucasfilm settled for more than $400 million. Enforcement of criminal cases, though, has been weak: DOJ's first criminal no-poach indictment was voluntarily dismissed with prejudice, and the Department has secured zero convictions in contested cases.

The practical read for a sourcer: the direct-competitor archetype is safe to mine unilaterally and dangerous only if you negotiate a mutual restraint around it. Keep any off-limits list your own unilateral policy, documented as such, and never as a handshake with a peer employer.

Keep the source list current

A source list is a live document, because archetypes change tag as trigger events fire and attrition trends turn. Re-run classification whenever a company on your list issues a return-to-office mandate, gets acquired, files a WARN notice, or misses a funding round, since any of these can move a steady competitor into a timed, high-yield window. Set a standing check against state mini-WARN registries and acquisition news for your top-ranked feeders.

Re-verify the counts too. Talent-flow data refreshes daily and its growth metric always compares against the count 12 full months prior, so the hire-to-departure ratio you ranked on last quarter drifts. Pull skill-filtered pool sizes again before a new search, and always confirm profile-based rankings against a second count so the adjacent-industry and scarce-skill feeders are not silently undercounted.

Worked in this order, the reference does one job well: it lets you look at any company on your list, name its type, and know in a glance what the people inside are worth, when they open, and how the row in front of you is trying to fool you.