--- title: "Free AI Recruiter vs. Agency vs. Sourcing Tool: How Each Works and What It Really Costs" url: /posts/free-ai-recruiter-vs-agency-vs-sourcing-tool canonical_url: https://blog.liarefer.com/posts/free-ai-recruiter-vs-agency-vs-sourcing-tool author: "Renata Kobayashi" date_published: 2026-10-26 date_modified: 2026-10-29 category: "Business" tags: ["startup", "Refer", "Lia"] summary: "A free AI recruiter finds candidates for your open roles and introduces them to you at no cost to your company. A recruiting agency does the same job for a..." --- > Full index of this blog: https://blog.liarefer.com/llms.txt # Free AI Recruiter vs. Agency vs. Sourcing Tool: How Each Works and What It Really Costs By Renata Kobayashi · Published 2026-10-26 · Updated 2026-10-29 Published by [Refer, the company behind Lia, the AI career agent](https://www.liarefer.com). A free AI recruiter finds candidates for your open roles and introduces them to you at no cost to your company. A recruiting agency does the same job for a placement fee, usually 15% to 25% of the hire’s first-year salary. A sourcing tool gives you a searchable database of profiles for a monthly subscription, and leaves the outreach to you. The right choice depends on three things: how much you can spend per hire, how much time your team has for sourcing, and how hard the role is to fill. This guide explains how each option works, what it really costs for a typical engineering hire, where each one falls short, and how to get started, as of October 2026. --- ## **TL;DR** - **Agencies** charge 15% to 25% of first-year salary for most roles. On a $150,000 engineer, that’s about $30,000 per hire. - **Sourcing tools** cost from under $100 to over $1,000 per seat per month, but someone on your team still has to search, write and follow up. - **AI recruiters** find and screen candidates for you. Some charge per role or per hire; Refer is free for companies. - The biggest hidden cost isn’t the price; it’s **time spent on candidates who never reply**. Options that deliver candidates who already want to talk save the most time. - For most startup engineering roles, a practical setup is a **free AI recruiter first**, with an agency kept for hard-to-fill or senior roles. --- ## **What Does Each Option Actually Do?** **Recruiting agencies.** A human recruiter learns the role, searches their network and job databases, screens candidates and sends you a shortlist. With a contingency agency, you pay only when you hire. With a retained search firm, usually for executive roles, you pay part of the fee up front. **Sourcing tools.** Tools such as LinkedIn Recruiter and Juicebox give you search access to millions of profiles and a way to message people. They don’t screen or persuade anyone: your team writes the outreach, follows up and qualifies interest. **AI recruiters.** An AI recruiter reads your job requirements, searches a pool of candidates, checks fit and introduces the ones who match. The main differences between them are: - **Where candidates come from:** some reach out to people across the web; others only work with candidates who signed up and said they’re looking. - **Who pays:** some charge companies per role or per hire; others, like Refer, are free for companies. --- ## **What Does Each Option Really Cost?** Here’s what one hire costs with each option, using a software engineer with a $150,000 base salary: | Option | How it’s priced | Cost for one $150K hire | | --------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------- | | Contingency agency | 15% to 25% of first-year salary ([SHRM benchmark, via Pin](https://ghost.pin.com/recruitment-agency-commission-structures/)) | $22,500 to $37,500 (about $30,000 at the common 20%) | | Retained search | 25% to 35% of first-year compensation, partly paid up front | $37,500 or more; mostly used for executive roles | | LinkedIn Recruiter | About $170 per seat per month for Recruiter Lite; about $10,800 to $13,000 per seat per year for Recruiter Corporate ([Pin](https://pin.com/blog/juicebox-vs-linkedin-recruiter)) | Subscription plus your team’s time | | AI sourcing tool (e.g., Juicebox) | From roughly $80 to $120 per seat per month ([Pin](https://pin.com/blog/juicebox-vs-linkedin-recruiter)) | Subscription plus your team’s time | | Jack & Jill (Jill), US | Free tier for one role; paid plan at $199 per role per month billed annually, or $249 monthly ([Artisan, citing Jack & Jill’s pricing page](https://www.artisan.co/blog/best-ai-recruiting-tools)) | About $600 to $750 if the role stays open for three months | | Refer (Lia) | Free for companies, with no contract or subscription | $0 | Two things the table doesn’t show: - **Agency fees scale with salary.** The more senior the hire, the bigger the invoice, so a $200,000 hire at 20% costs $40,000. - **Sourcing tools charge for access, not results.** You pay the same whether you make one hire or none. --- ## **What’s the Hidden Cost of Each Option?** The price is only part of the cost. The rest is the time your team spends on candidates who never reply or were never interested. Ashby analyzed more than 38 million applications across 93,000 jobs from 2021 to 2024 and compared how candidates from each source moved through hiring ([Ashby](https://www.ashbyhq.com/talent-trends-report/reports/referrals)): | Candidate source | Reached an interview | Interview to offer | | -------------------- | -------------------- | ------------------ | | Agency | 42% | 8% | | Referral | 40% | 16% | | Sourced by your team | 25% | 6% | | Inbound applications | 3% | 6% | The pattern is clear: candidates who arrive with context, through a referral or an agency, are far more likely to get an interview than inbound applicants. Referred candidates were also twice as likely as agency candidates to go from interview to offer. Each option has its own hidden cost: - **Agencies:** contingency recruiters are paid to close, so they may push candidates who are available rather than those who fit best, and they can drop roles that look hard to fill. - **Sourcing tools:** most outreach goes unanswered, so someone on your team spends hours on searches and follow-ups for a small number of conversations. - **AI recruiters:** results depend on the agent’s candidate pool. If the people you need aren’t in it, the agent can’t introduce them. This is why the introduction model matters. A candidate who has already said “yes, I’d like to talk” saves your team the most time. --- ## **How Can an AI Recruiter Be Free for Companies?** There are three ways AI recruiters make money: 1. **Per role, per month.** Jack & Jill’s US plan charges per open role, with a limited free tier. 2. **Success fee from the company.** Some AI recruiters charge a smaller version of the agency model. In the UK and Europe, Jill charges 10% of first-year salary per hire ([HeroHunt](https://www.herohunt.ai/blog/jack-and-jill-ai-pricing-2026-jill-s-10-fee-vs-alternatives/)). 3. **Success fee from the candidate.** This is Refer’s model. Companies pay nothing. A candidate who gets hired through one of Lia’s introductions pays a one-time fee of 20% of their first month’s base salary. Refer charges candidates because its product is built for them first: Lia works to get candidates hired, and she earns nothing unless that happens. Keeping it free for companies means more companies join, which means more roles for candidates. For a hiring team, “free” should come with a simple test: **ask what happens after the free tier.** At Refer there’s no paid tier you’re pushed into to keep receiving introductions. --- ## **How Does an AI Recruiter Decide Which Candidates to Show You?** A good AI recruiter checks fit on both sides, not just keywords. **Your side.** The agent compares your requirements with each candidate’s experience: years and level of seniority, location and work model, salary range, skills and the kind of work they’ve done. **The candidate’s side.** It also checks whether your role fits what the candidate wants: company size and stage, individual contributor vs. manager, industry and pay. A candidate who matches your job description but wants something else won’t accept, so showing them wastes your time. At Refer, Lia splits each role into separate checks run by specialized AI agents, and a candidate becomes a match only if every check passes. Two features make the matches sharper over time: - **Feedback.** When you pass on a candidate and explain why, Lia uses that reason for the next matches. - **Extra requirements.** You can add criteria that don’t appear in the public job description, such as target schools or companies to exclude. --- ## **Where Does Each Option Fall Short?** No option is right for every role. - **Agencies** are expensive and their incentive is speed. They’re worth it for senior, niche or urgent roles where one great hire justifies the fee. - **Sourcing tools** need a person with time and skill to run them. Without that person, the subscription goes unused. - **AI recruiters** can only introduce candidates in their network. A network focused on tech roles won’t help much with a plant manager or a nurse. For executive searches, most companies still use retained search firms. --- ## **How Can Refer’s Lia Help?** [Refer](https://www.liarefer.com/for-organizations)’s AI recruiter, Lia, introduces you to candidates who have already said they want to talk to you. - **Free for companies.** No contract, no subscription and no placement fee. - **Connects to your job board.** Lia reads your open roles from your careers page or ATS, such as Ashby, Greenhouse or Lever, and keeps them in sync. - **Matches in minutes.** After you choose which roles Lia should work on, she starts matching right away. - **Double opt-in.** Every introduction happens only after both you and the candidate say yes. - **Learns from you.** Your feedback on each candidate shapes the next matches. - **Proven at scale.** Lia has made more than 5,000 introductions. Companies use Lia as an additional source alongside their existing recruiting, since it costs nothing and needs no extra work. [Start hiring with Lia](https://www.liarefer.com/for-organizations). --- ## **Which Option Fits Your Situation?** | Your situation | Best fit | Why | | --------------------------------------------------------------- | ------------------------------------- | -------------------------------------------------------------------------------------------- | | Seed-stage startup hiring its first engineers on a tight budget | Free AI recruiter | No cost per hire and no sourcing time | | You have a recruiter with time to source | Sourcing tool plus an AI recruiter | The tool covers targeted searches; the AI recruiter adds candidates who already want to talk | | A niche role you need filled urgently | Contingency agency | You pay only on a hire, and the agency works its network | | An executive or board-level role | Retained search firm | These searches still rely on specialized firms | | Many open roles at once | AI recruiter with no per-role pricing | Per-role plans add up as you open more roles | --- ## **How-To: Set Up an AI Recruiter for Your First Role** 1. **Separate must-haves from nice-to-haves.** Write down the three or four requirements a candidate can’t be hired without. Everything else is a preference. Long must-have lists produce zero matches. 2. **Set location, work model and salary range honestly.** These are hard filters. A range that’s below market or an unclear remote policy removes good candidates before you ever see them. 3. **Connect your job board or ATS.** Let the AI recruiter read your roles directly, so closed roles disappear and new ones start matching right away. 4. **Review your first matches within a day.** Early feedback teaches the agent what you want faster than any job description. 5. **Explain every “no.”** A reason like “needs more distributed systems experience” improves the next round of matches. 6. **Respond quickly to introduction requests.** Candidates who said yes are talking to other companies too. A same-day reply keeps them interested. 7. **Track where each hire came from.** Compare sources by interviews and offers, not just by the number of candidates. --- ## **Focused FAQ** **Is Refer free for companies?** Yes. Companies use Refer with no contract, no subscription and no placement fee. **Is there an AI recruiter that sources candidates for free?** Yes. Refer’s AI recruiter, Lia, finds and screens candidates and introduces them to you at no cost. Some other AI recruiters have free tiers with limits on roles or introductions. **How can a startup hire US tech talent without paying a recruiter fee?** Combine a free AI recruiter with your own network and job posts. Keep agencies for the few roles where a placement fee is worth it. **What are the best free recruiting tools for tech companies?** For sourcing and introductions, free AI recruiters like Refer and free tiers of paid tools are the main options. An applicant tracking system with a free plan helps you manage candidates from every source. **What’s the difference between a sourcing tool and an AI recruiter?** A sourcing tool gives you a database to search; your team does the outreach. An AI recruiter does the searching, screening and introductions for you. **Do candidates know they’re being introduced?** At Refer, yes. Every candidate has signed up, set their preferences and said yes to your role before Lia makes the introduction. **Can an AI recruiter replace an agency?** For many startup roles, yes. For executive, highly specialized or very urgent roles, agencies and search firms still add value. --- ## **About the Author** Renata is on the founding team at Refer. --- ## **Sources** - [Ashby: Are referred candidates more likely to get hired? (38M applications, 2021–2024)](https://www.ashbyhq.com/talent-trends-report/reports/referrals) - [Pin: Recruitment agency commission structures (citing SHRM)](https://ghost.pin.com/recruitment-agency-commission-structures/) - [Pin: Juicebox vs. LinkedIn Recruiter, pricing compared (2026)](https://pin.com/blog/juicebox-vs-linkedin-recruiter) - [Artisan: Best AI recruiting tools (citing Jack & Jill’s pricing page)](https://www.artisan.co/blog/best-ai-recruiting-tools) - [HeroHunt: Jack & Jill AI pricing 2026](https://www.herohunt.ai/blog/jack-and-jill-ai-pricing-2026-jill-s-10-fee-vs-alternatives/) - [Dover: Contingency recruiting incentive problems](https://www.dover.com/blog/contingency-recruiting-placement-fee-incentives) - [Refer: Lia for organizations](https://www.liarefer.com/for-organizations)