CGP Group Explains How a 200-Person Recruitment Firm Puts AI at the Core with Spott
TL;DR
CGP Group is a 200-person recruitment firm with 12 offices, running executive search, contract staffing and employer of record services across APAC, the Middle East and the US.
It had accumulated three applicant tracking systems in parallel (Mercury, Bullhorn and Vincere) plus a patchwork of point tools for sequencing, sourcing and dialing.
CEO Chris Watkins ran a company-wide digital transformation in which AI at the core, not consolidation, was the actual goal. Replacing the ATS layer was the starting point, not the destination.
Three requirements cut the vendor field down fast: AI-native architecture, real API agility, and a vendor that moves at their speed. Spott, the AI-native ATS and CRM the firm selected, is how each of those got answered.
The rollout is bringing 200 users across 12 countries onto Spott, with weekly training across APAC time zones and a standing weekly engineering sync between CGP's developers and Spott's product team.
The transferable lesson: if you are betting an AI transformation on a platform, what you are really buying is how fast the team behind it moves. Not the feature list on the day you sign.
The situation: three ATSs, one firm
Most agencies replace one applicant tracking system. CGP Group is replacing three at once.
The firm had grown across offices in Asia, the Middle East and the US, and different parts of the business had landed on different platforms: Mercury in one place, Bullhorn in another, Vincere in a third. Around them sat the usual accumulation of point tools for sequencing, sourcing and dialing, each solving a real problem and none of them talking to the others.
This is an extremely common pattern in multi-office recruitment firms, and it is rarely the result of bad decisions. It is good local decisions made at different times, in different markets, by teams who needed something to work that quarter. The cost only shows up later: no single view of the network, no consistent process, and a data layer too fragmented to put anything intelligent on top of.
"By automating and centralizing everything, it makes life easier and delivers real impact for us at scale." Chris Watkins, CEO, CGP Group
The goal was AI, not tidiness
CGP did not set out to consolidate its ATS estate. It set out to put AI at the core of how the firm recruits, and consolidating the estate turned out to be the prerequisite.
That ordering changes the specification. If the goal is tidiness, you pick whichever incumbent has the most users and migrate the others onto it. If the goal is AI doing real work, matching candidates, drafting outreach, keeping profiles current without
recruiters retyping information, then the question becomes which platform is actually capable of that. That is the question that led CGP to Spott.
What "AI at the core" looks like in Spott on an ordinary Tuesday, shown here with demo data: the recruiter describes the role in a sentence and the platform searches the firm's own database against it.
image file: images/cgp-customer-story/asset-1-ai-search.png
What a 200-person firm insisted on, and how Spott answered
Chris and his team did the evaluation themselves rather than outsourcing it: dozens of platforms compared, including the legacy systems they already ran and the newer AI-first generation alongside them. Three requirements narrowed the field, and each is worth reading as a test Spott had to pass.
1. AI-native, not AI bolted on
CGP had watched established platforms relaunch with an "AI strategy" layered over data structures designed fifteen years ago. Their read was that this produces AI features that demo well and disappoint in daily use, because the underlying data model was never built to support them.
Spott cleared that bar because AI is the architecture, not a module: candidate matching, automated campaigns and self-updating profiles are what the platform is built around, reading the notes, call summaries and correspondence it captures itself.
The useful question for any HR or talent leader running a similar evaluation is not "do you have AI features," which everyone now answers yes to, but "was the data model designed for this, or retrofitted?" Ask the vendor to show you what the AI reads, and how it got there.
2. Real API agility
A firm at CGP's scale lives on integrations. Their requirement was genuine APIs, webhooks and custom data access, not a closed system with a connector marketplace attached. Spott's open API and webhooks are what CGP's own developers now build against directly, pushing data into the finance systems, reporting layers and internal tooling a 12-office firm runs on.
Recruitment tech evaluations tend to skip this, and it is the gap that hurts most at scale. If the platform's answer to every integration question is "that is on the roadmap," the firm ends up rebuilding the workarounds it just paid to escape.
3. A vendor that moves at their speed
CGP wanted a direct line to the people building the product: a working relationship with the team that sets the roadmap, not a ticket in a support queue. With Spott, that is literally what they got. Chris, a founder himself, framed the relationship with Spott's founders as a two-way street, a partnership in which each side helps the other "build each other's businesses."
Conventional procurement logic says a 200-person firm de-risks by buying from the largest vendor available. CGP inverted that: when you are betting a transformation program on a platform, a direct line to the people deciding what ships next is worth more than the reassurance of a big logo.
What the rollout actually looks like
CGP is bringing 200 users onto Spott across 12 countries. The database is migrated as it stands first, then fine-tuned with the firm in a validation environment before anyone works in it. Two cadences define how the rollout runs.
Weekly training, in the right time zones. Sessions run every week across APAC time zones rather than being compressed into a single go-live event. Most ATS rollouts fail on adoption, and the pattern is always the same: one training session, delivered once, at an inconvenient hour, to people who then go back to their old habits.
A standing weekly engineering sync. CGP's leadership and engineering team meet Spott's product and engineering side every week, and this is where the honest part of the story lives. Spott's API did not have everything CGP's developers needed on day one. Specific webhooks, endpoints and filters were missing, and CGP asked for them by name. Those requests came back shipped, often within the week and sometimes the same day: on a recent call, the previous week's list was read back as a checklist, done, done, shipped in today's release.
That is also the honest state of the project overall: the rollout is still in progress, training is still running week by week, and nobody on either side is calling it finished. Chris went into it with clear worries, and measured against those, his verdict so far: "I think we've nailed the two biggest things I was most concerned about."
Four lessons from CGP's move to Spott
1. Decide what the AI is for before you shortlist vendors. CGP's requirements were sharp because the goal was specific. "We want AI" produces a demo tour. "We want matching, outreach and data upkeep handled by the platform," the specification Spott was scored against, is something you can actually test.
2. Ask where the AI's data comes from. A matching engine that only sees structured fields is a search box. Spott's reads the notes, call summaries and correspondence it captures itself, which is what made requirement one testable rather than a marketing claim.
3. Treat integration access as a hard requirement. APIs and webhooks determine whether you can ever connect the platform to the rest of the business. CGP's developers were building against Spott's API during the evaluation-to-rollout window, not after some future marketplace release.
4. Buy the team's velocity. Every platform you evaluate will look different in a year. What CGP effectively bought is the weekly cadence at which Spott ships what they ask for, and that access to the people setting priorities is the durable thing a buyer chooses.
See what CGP saw
The evaluation CGP ran is one any firm can repeat in a single meeting: bring a real role, a real integration question and your hardest data-quality problem, and see how the platform holds up. Book a 30-minute Spott demo and run that test yourself.
and run that test yourself.About Spott
Spott is an AI-native ATS and CRM for recruitment and staffing firms, built around matching, outreach and automatic data upkeep rather than manual record keeping. It is Y Combinator backed, ISO 27001 certified and GDPR compliant, and it is used by agencies on every continent, from boutique executive search teams to multi-country groups like CGP.
