We Analyzed 822 Recruitment Agency Calls - Here's What Tools They're Actually Using in 2026

We pointed Claude Code at 822 real call transcripts with recruitment agency owners, recruiters, and ops leaders across the UK, US, and DACH region. Total cost: $2.36. The result: 820 unique software tools, 15,743 total mentions, and the most honest picture of what recruitment agencies are actually running in 2026. No vendor bias. No sponsored rankings. Just data from real conversations.

How We Analyzed 822 Calls for $2.36 (And Why You Should Trust This Data)

Most "tech stack reports" are either vendor surveys where people tick boxes they barely remember, or listicles written by tools trying to rank themselves. We wanted something different. We wanted to know what agencies actually talk about when they're on a call with us - what they're using, what's frustrating them, and what they're asking about.

The dataset is 822 real sales and client calls with recruitment agency founders, team leads, and operations managers. These aren't survey responses. Nobody checked a box next to "Bullhorn" because it was listed as an option. These are tools mentioned naturally in conversation - unprompted, in context, with real opinions attached.

We used Claude Code with Claude Haiku 4.5 for LLM-based extraction across 10.2 million input tokens. The total API cost was $2.36 - that's not a typo. The kind of analysis that would have taken a research team weeks and cost tens of thousands of dollars, done in hours for less than a coffee.

Here's what the process actually looked like: we fed each call transcript through Claude Haiku with a prompt asking it to extract every software tool, platform, and service mentioned. The LLM returned structured data - tool name, category, and the context in which it was mentioned. We then aggregated all 822 results into a single dataset: 15,743 individual tool mentions across 10 categories (ATS, CRM, Sourcing, Outreach, Data Enrichment, Automation, AI Tools, Communication, Job Boards, and Other).

What LLM extraction catches that keyword matching would miss: abbreviations, casual references, misspellings, and niche tools that don't appear on any "top 10" list. When an agency owner says "we use that thing, what's it called, the n8n workflow builder," the LLM captures that. Keyword search wouldn't.

But raw extraction isn't enough. We identified 820 tool names, then ran a full deduplication and verification pass. "Vincuri" is actually Vincere. "NNN" is n8n being spoken aloud and picked up phonetically by the transcript engine. "Hayreach" is HeyReach. "Epify" is Apify. "Monatel" is Manatal. We verified every tool in the top 100 against its actual product name, merged duplicates, and removed entries that turned out to be companies or services rather than software tools. Without this cleanup, you'd be reporting phantom tools that don't exist. This is why raw LLM extraction without human review produces garbage data - and why the $2.36 in API costs was really just the starting point. The real value was in the deduplication and analysis layer on top.

What this data does NOT capture: tools that never come up in conversation, internal-only systems, or tools that prospects haven't discovered yet. We're seeing what agencies know about and use, not the entire market. It also skews toward the UK, US, and DACH markets because that's where our client base is concentrated.

This is what AI-powered operations look like in practice. And it's exactly the kind of system we build for recruitment agencies - turning unstructured data into decisions. If you want to understand how tools like Claude Code work in a recruitment context, we wrote a full guide on AI agents for recruitment agencies.

What ATS and Recruitment CRM Are Agencies Actually Running?

This is the unbiased gold in the dataset. ATS and recruitment CRM platforms are tools agencies bring up themselves - they're not influenced by our recommendations. Here's what we found:

Loxo and Bullhorn are virtually tied at the top. That's significant. Bullhorn's brand dominance in the staffing world suggests it should be running away with this, but Loxo has quietly built a massive user base - particularly among agencies that want an all-in-one platform without the enterprise complexity.

Atlas at 100 mentions is a strong showing for a newer AI-native recruitment CRM. It signals where the market is heading - agencies want AI built into their core platform, not bolted on as an afterthought. The ATS/CRM convergence is real: Loxo, Atlas, RecruitCRM, and Crelate all serve as both applicant tracking and relationship management. The days of needing a separate ATS and CRM are ending for most agencies.

We also identified some lesser-known but significant platforms: Recruiterflow (15 mentions), Manatal (8), Thrive TRM (8), and Greenhouse (6). The long tail is real - agencies are running everything from enterprise platforms to tools most people in the industry have never heard of.

A note on Vincere (32 mentions): it's a respected platform, but we're hearing growing frustration from agencies about data export limitations and access restrictions. Users on review platforms report charges north of EUR 1,000 to export their own data, with exports capped at 5,000 records per batch. If data portability matters to your agency - and it should - this is worth investigating before you commit.

The pattern we see across almost every agency: build on top of what you already have. One of our clients, Cast UK, runs Bullhorn and didn't want to switch. So we built their automation system on top of it. That's the right approach for most agencies - your ATS is your foundation, not the thing you replace.

The General CRM Layer - Why HubSpot and Salesforce Keep Showing Up

HubSpot (219 mentions) and Salesforce (170 mentions) show up heavily in our data, and it surprised us at first. These aren't recruitment-specific tools. But many agencies run a general CRM alongside their recruitment platform - HubSpot for marketing and client BD, Salesforce for managing the sales pipeline.

Pipedrive (32 mentions), Attio (28), and Close (26) also appear. GoHighLevel (11 mentions) is an interesting newcomer - agencies exploring all-in-one marketing + CRM plays that sit outside the traditional recruitment tech world.

The real question isn't "which CRM" but "do you need two?" For most agencies under 20 people, the answer is no. A modern recruitment CRM like RecruitCRM or Atlas handles both candidate relationships and client pipeline. Running HubSpot on top of Bullhorn doubles your data entry and halves your data quality.

Where Are Agencies Sourcing Candidates?

LinkedIn's dominance is absolute and shows no sign of weakening:

Job boards still matter more than the "everything is LinkedIn now" narrative suggests. Indeed (185 mentions), Google for Jobs (175), and StepStone (73) generate significant candidate flow, especially for volume roles.

The new sourcing tools on the block: Juicebox/PeopleGPT (24 mentions) and PIN (8 mentions) are AI-native sourcing platforms that bypass LinkedIn's limitations entirely. They're pulling from databases of hundreds of millions of profiles and using AI to match candidates against your requirements in seconds. Instead of building Boolean strings and scrolling through LinkedIn Recruiter results, you describe the candidate you need in plain English and the AI returns a ranked list. Still early in adoption, but the mention frequency is growing fast - these tools barely existed 18 months ago.

The DACH market operates in a completely different tool universe. XING (12 mentions), StepStone (73), Bundesagentur fur Arbeit (29), and niche boards like Praktischarzt (medical) and jobs.kliniken.de (hospitals) show that agencies in Germany, Switzerland, and Austria need sourcing strategies that go well beyond LinkedIn. GitHub (9 mentions) appears for tech sourcing because developers aren't on LinkedIn the way other professionals are.

The Outreach Stack - What Agencies Use to Reach Candidates and Clients

One of the biggest patterns in our data: a large number of recruitment agencies still don't use a dedicated outreach tool at all. They send candidate and client emails directly from their ATS or CRM - doing mailshots from their primary domain. This destroys deliverability and lands their emails in spam folders. If you're running outreach from your recruitment CRM, you're burning your main domain's reputation. We recorded a full free cold email infrastructure course that breaks down how to set up proper sending infrastructure - this is the single most common mistake we see agencies make.

The other thing the data doesn't fully capture: cold calling is still huge. Many agencies run heavy phone-first outreach, especially in the UK and DACH markets. Phone-based BD and candidate outreach doesn't show up as a "tool" in the same way - it's baked into their ATS or done from personal phones. The shift to multi-channel is real, but the phone isn't going anywhere.

Among agencies that do use dedicated outreach tools, here's what we see:

Instantly and Lemlist dominate. Dripify (35 mentions) is popular for LinkedIn connection campaigns and drip sequences - quite a few agencies are running it alongside their email outreach.

SourceWhale at just 12 mentions is a surprise. It's purpose-built for recruitment outreach, but it's being outpaced by general outreach tools that recruiters are adapting to their workflow. The market is telling us something: recruiters prefer flexible tools they can customize over recruitment-specific platforms that constrain them.

The clear trend: agencies are moving from single-channel (just email or just LinkedIn) to multi-channel stacks. If you're still sending outreach from one channel, you're leaving replies on the table. We break down the exact tools and costs in our guide on the best recruitment tech stack for small agencies.

Data Enrichment - 111 Tools and the Fragmentation Problem

111 unique enrichment tools across 822 calls. By far the most fragmented category in our entire dataset. To put that in perspective: the ATS category had 49 unique tools. Outreach had 61. Enrichment had 111. The fragmentation is nearly double any other category.

The reason is simple: no single tool covers all data types (emails, phone numbers, company data, intent signals, technographics) for all markets. European data is different from US data. Healthcare contacts are different from tech contacts. And the quality of any single provider's data degrades over time as people change jobs and phone numbers.

Apollo.io leads at 355 mentions because it doubles as a B2B database and an outreach platform - agencies use it as a one-stop shop for finding and contacting prospects. ZoomInfo (127) remains the enterprise standard for company intelligence, though its price point puts it out of reach for most small agencies. ContactOut (92) is the go-to for finding personal email addresses and direct phone numbers from LinkedIn profiles.

The result is waterfall enrichment - agencies stack three to four tools to maximize coverage. BetterContact and Clay both enable this by routing a single lookup through multiple providers automatically, using the first provider that returns a verified result. If you're still relying on one enrichment source, you're missing 30-40% of available contact data. The agencies with the highest outreach performance in our client base all run waterfall enrichment.

Which Automation Platforms Are Agencies Adopting?

Clay is mentioned more than any other tool in the entire dataset at 1,908 mentions. It bridges data enrichment, automation, and outreach in a single platform, which is why it dominates. You can build an entire prospecting workflow in Clay - scrape job postings, enrich the hiring manager's contact details, personalize an outreach message, and push it to your email tool - without writing a single line of code.

n8n at 576 mentions has pulled significantly ahead of Zapier (224). That's a big shift. Two years ago, Zapier was the default. Now agencies are gravitating toward n8n because it handles complex, multi-step workflows that Zapier struggles with - conditional logic, loops, error handling, and custom API calls. The open-source, self-hosted option also resonates with agencies that want control over their data. Make (126 mentions) is the solid third option with particular strength in the European market, offering a visual interface that sits between Zapier's simplicity and n8n's power.

Apify (30 mentions) appears as the go-to web scraping tool - the glue behind custom job board scrapers, LinkedIn data extraction, and any workflow that needs to pull data from a website that doesn't offer an API.

We need to be honest here: Clay and n8n are tools we actively recommend and build with. Their mention counts are inflated because we discuss them in almost every call. But the trend of agencies moving from Zapier toward n8n and Make for more complex, multi-step workflows is real and confirmed by what we see in implementations.

If you're thinking about building your first automation stack, we mapped out the exact sequence in our 30-day playbook.

Are Recruiters Actually Using AI? Here's What the Data Shows

ChatGPT leads, but Claude at 321 mentions is closer than most people expect. And Claude's share is growing - here's why. Claude Code and Claude's ability to build actual systems (not just chat) makes it the tool of choice for agencies that want to go beyond "write me a job description." When recruiters discover they can build automation workflows, custom research agents, and entire operating systems with Claude, they stick. It's the difference between a tool that answers questions and a tool that does work.

Perplexity (73 mentions) is gaining real traction as a research tool. Recruiters are using it to research companies before BD calls, understand niche industries, and build candidate briefs. It's faster than Google for getting synthesized answers, and the source citations give recruiters something to reference in their outreach.

ElevenLabs (14 mentions) signals the voice AI wave. Agencies are starting to explore AI-powered candidate outreach calls, voicemails, and even voice agents that can handle initial screening conversations. This is still early, but the agencies experimenting with voice AI are seeing engagement rates that traditional outreach can't match.

The reality check: Korn Ferry says 84% of talent leaders plan to use AI extensively in 2026. Gartner predicts that by 2027, 80% of recruiting technology vendors will have advanced AI capabilities built into their platforms. Our data confirms agencies ARE experimenting, but mostly with general-purpose AI tools like ChatGPT and Claude rather than recruitment-specific AI products. The gap between intention and implementation is narrowing, but it's still real. We wrote a full practical guide on how to actually use AI agents in recruitment.

820 Tools Across 822 Calls - What This Really Tells Us

The average call references 19.2 different tools. That number should make every agency founder pause. Industry reports suggest recruiters use 4-6 tools daily, but the total ecosystem they depend on is three to four times larger. Most agency owners don't even realize how many tools their team touches in a given week.

Airtable sits at 1,254 mentions - the third-most mentioned tool in our entire dataset, right behind Clay and LinkedIn. That's not because Airtable is a great recruitment tool. It's because agencies are duct-taping their systems together with spreadsheet-databases when their core tools don't connect properly. Airtable becomes the unofficial "glue layer" - tracking candidates that fell out of the ATS, managing BD pipelines that don't fit in the CRM, storing outreach templates, and running reporting that should be automated. Google Sheets (68 mentions), Notion (52), and ClickUp (58) play the same role. They're filling gaps that should be handled by proper integrations between purpose-built tools.

Fathom (93 mentions) is the most-mentioned meeting tool that nobody talks about as "recruitment tech." But it's in nearly every agency's stack, quietly recording and transcribing every client and candidate call. It's also a signal that agencies are starting to treat their call data as an asset - the same instinct that led us to analyze 822 transcripts for this report.

The real problem isn't which tools you pick. It's whether they talk to each other. Most agencies buy best-of-breed tools and then spend 20-30 hours per week on manual work because the systems don't connect. HRD Canada reports that recruiters spend up to 75% of their working hours on tasks that could be automated. That's not a tool problem - it's a system problem. And it's exactly what we mean when we say you should think about building a recruiting operating system instead of just buying more software.

Here's what the winning agencies in our client base have in common: they don't have the most tools. They have the most connected tools. Fewer platforms, tighter integrations, automation handling the data flow between them. That's the difference between an agency running 15 disconnected tools and an agency running 8 tools that operate as one system.

“It's not your people, it's your system. The agencies winning in 2026 are the ones that connected their tools instead of adding more.”

If you're curious about the actual financial impact of connecting your tools, we broke down the numbers in our recruitment automation ROI analysis.

A note on this data: This report reflects what comes up in our conversations with recruitment agency owners, recruiters, and ops leaders - both prospects and existing clients. Tools we recommend (like Clay, n8n, Instantly) are naturally over-represented because we discuss them frequently. Tools agencies use silently in the background (like their phone system or accounting software) are under-represented because they rarely come up on a call. This is not a statistically controlled survey - it's a real-world signal from 822 conversations. Treat the recruitment-specific tools (ATS, CRM, sourcing, job boards) as the most unbiased data points, and the automation/outreach/AI tools as directional trends influenced by our own recommendations. We believe that transparency makes data more valuable, not less.