Sales talent: AEs, SDRs and the people who close

Go-to-market hiring gets bundled with engineering hiring because both are "recruiting," and the two markets share almost nothing that matters. Engineers turn over at about 9% a year; sales and design at about 13% (SignalFire). Engineering candidates are hard to assess and hard to poach; sales candidates are easy to assess — quota, attainment, ACV, cycle length — and easy to poach, because they move for money by professional habit. A CV tells you almost nothing about either, but for sales the missing information is knowable from references and numbers rather than from a work sample.

The consequence shows up in the fee structure. RevPilots publishes the clearest public fee ladder for GTM roles (RevPilots):

Read the top and the bottom of that table as two different businesses. A CRO search at $50–92K is retained-search work in contingency clothing. An SDR placement at $9–14K is a $10K fee against a role with 30–50%+ annual attrition and a 90-day guarantee sitting against a roughly 9-month median tenure. The mismatch is why the low end of GTM is the worst segment in contingency recruiting, and why the marketplaces in Contingency recruiting marketplaces gravitate to engineering instead.

#Whose budget

The buyer is a funded B2B startup with a revenue target, and this is the one place GTM hiring is better than engineering hiring: the pain is immediate and arithmetic. An unfilled engineering seat slips a roadmap; an unfilled AE seat misses a quarter, and the founder can compute the miss. That shortens the sale and raises willingness to pay at the closer end.

It is still a displacement sale. The buyer's comparison is always a salary. The in-house benchmark for a US SDR is $48.6K base, ~$63.1K OTE, plus >30% benefits, $6–10K of recruiting cost per hire, ~$6K of first-year training and ~$120K/yr of manager overhead (memoryBlue). Every price you quote is measured against that number, and against the outsourced alternative next to it — see Outbound and GTM-as-a-service.

Budget scores a 3: urgent and immediately felt, but small per transaction at the volume end and priced against a salary at every level.

#Can you get the supply

Easily, which is the problem. This is the weakest supply side in the atlas.

Sales candidates are abundant, self-identifying and already organised into communities that predate any marketplace — Bravado, Pavilion, RepVue. Sales-specialist recruiting shops (RevPilots, HireDNA, SalesGig) are numerous and small. Nothing here is scarce in the sense Which side you build first means: both sides can be assembled with a budget and a quarter.

What is scarce is trustworthy performance data. Everyone claims 120% of quota; the quota is unverifiable and the number that matters — attainment relative to a team median, in a comparable motion — is not on the CV. RepVue raised $5M in 2022 to attack exactly this from the employer side, building crowdsourced ratings of sales organisations, and monetises through employer branding and job listings rather than placement fees (PR Newswire). That is a revealing choice: the company holding the best sales-performance dataset in the category decided not to be a marketplace.

Supply scores a 2. Hard supply is the only thing a competitor cannot copy in a weekend, and this supply is copyable in a weekend.

speed: 4, for the same reason. Supply that assembles in a weekend is supply you can be selling the week after, the fee ladder is public so there is no pricing to discover, and the buyer is the most motivated in this atlas — a founder who can compute the quarter they miss while the seat stays empty signs a contingency agreement in an afternoon, because it costs nothing until someone starts. What holds it off a 5 is that the invoice is triggered by a fill rather than by an engagement, and a fill takes about two months before any guarantee period runs — the same lag that caps Contingency recruiting marketplaces at 4. A quarter from a standing start to money, and no GTM-specific fill-rate data exists to tighten that estimate.

#What the spread looks like

Same shape as Contingency recruiting marketplaces, one notch worse at every level. If a marketplace ran the 70/30 split Paraform is reported to run (HeroHunt), a platform's take on an SDR placement is roughly $2.7–4.3K and on a VP Sales placement $12–19K. [UNVERIFIED] — derived from the RevPilots ladder and the Paraform split, not observed anywhere in GTM.

The volume end therefore cannot support a marketplace's fixed costs, and the executive end is where retained search already lives and wins, because what a board buys in a VP Sales search is legible accountability for a $500K decision, not candidate flow. Retained runs 25–33% of first-year total comp with $80–100K minimums, paid in thirds (Pin). The middle — Mid-Market and Enterprise AE at $16–36K a placement — is the only band where a marketplace is arithmetically interesting, and it is also the band where the buyer most credibly says "I will just ask my last VP Sales who they would bring."

spread: 3. The fraction is the same as Contingency recruiting marketplaces — roughly 25–30% of the fee, on the only split anyone has reported — and the fraction is what this score measures, so it lands on the same digit. The absolute prize does not: $2.7–4.3K on an SDR placement against $12–19K on a VP Sales one, which is why the level of the score flatters the business. The reason it is not a 2 is the top of the ladder, where CRO and VP searches at 25–33% of first-year comp carry a real fee; the reason it is not a 4 is that the fee percentage is set by a public, shoppable ladder that RevPilots and every competitor publish, and a number the buyer can look up is a number the buyer negotiates — 3–7 percentage points is achievable on ordinary terms.

See What a rake can actually be for the general form of this argument and GMV is not revenue for why the fee, not the salary, is the denominator.

#Can you hold it

Barely. Every leakage condition from Getting cut out applies, and one more on top: sales leaders hire from their own networks as a matter of professional identity. A VP Sales arrives with a list of people they have hired before and intends to hire them again. That relationship is the competitor, and it is free.

The buying is also episodic in a worse way than engineering. A startup hires an AE pod, then does not hire again for two quarters, then hires four at once when the round closes. There is no habit to build on and no retention flywheel — the same defect that killed Hired and Vettery, described in Triplebyte and Hired.

hold: 2, the same digit as Contingency recruiting marketplaces with less room underneath it. Both carry the full Getting cut out set — large single-deal values, an identity-revealing introduction, episodic buying — and this page adds a fourth condition the engineering side lacks: the buyer's cheapest alternative is the phone of the sales leader they just hired. Free, better-informed and culturally expected. What keeps it off a 1 is that 13% annual attrition manufactures repeat demand whether or not a habit forms, which is a grim reason to score a point.

#What AI does to it

This is the dimension where sales talent scores worse than engineering talent, and it is the reason for the stance on this page.

The AI application flood argues for human curation everywhere — 244 applications per open role in 2025, up from 116 in 2022, and time-to-fill up 37% to 59.7 days (TheHireHub). Sales roles attract that flood harder than engineering roles do, because the barrier to applying is lower.

But the job you are placing at the volume end is itself being automated. Clay-based GTM agencies bill $3.5–10K/month mid-market, with a practitioner survey putting the median retainer at $5–8K/month (GTME Pulse), explicitly benchmarked against the ~$100K all-in cost of one US SDR. Belkins publishes "from $5,000/mo"; SalesRoads quotes $9,950 per 4 weeks for one SDR (RevenueFlow). If a startup can buy the outcome for the price of the salary, it does not hire the SDR, and there is no fee.

ai: 2, a point below Contingency recruiting marketplaces and the clearest reason these two pages carry different stances. In engineering hiring models attack the recruiter's labour and leave the hire intact — sourcing gets cheaper, the fee is unchanged. Here they attack the hire. An SDR seat a startup can replace with a $5–8K/month retainer is a seat that never opens, and a seat that never opens generates no fee at any take rate. That removes the volume end of the ladder rather than repricing it, and the volume end is where the placement count lives. It is not a 1 only because the closing half of the funnel is so far untouched. See What better models do to each layer.

#What would kill it

#Who is already there

Thinly, and mostly not as marketplaces.

• Bravado raised $26M led by Tiger Global and acquired Compgauge in the same announcement, to build a "B2B sales hiring ecosystem"; its main competitor is Pavilion (PR Newswire; CB Insights). Its Rise layer is a community-to-marketplace conversion play — the right sequence, since the community is the hard part.
• RepVue: $5M, crowdsourced sales-org ratings, monetised through branding and listings.
• RevPilots: 20–25% for full-time sales hires with replacement guarantees; the published ladder above is the clearest public GTM benchmark.
• HireDNA, SalesGig: small assessment-led sales recruiting shops.
• CloseFactor raised $4.5M led by Sequoia in 2022 (CloseFactor) and is routinely miscategorised in this vertical. It is sales intelligence — account research — not a talent marketplace. It does not belong on this page except as a warning about category lists.

#Where the record is thin