01/04/2026
💷 How AutoSkills turned a "dormant" database into 15 bookings a day?
AutoSkills UK is one of the leading automotive recruitment agencies in the UK, founded in 1993.
Owner - Graham, came to us with aN idea: unleash voice AI on their dormant candidate database - tens of thousands of technicians, panel beaters, HGV mechanics, service advisors, and sales execs.
His human consultants simply didn't have the bandwidth to re-engage them all.
But Graham knew those were real, quality people sitting in the system.
So we built a lean, cost-effective, and fast system using only Google Sheets, Cal.com, and Make.com - plus the telephony and email tools they already had.
Shortly after, Beth was born - a natural UK-voiced AI agent that calls every business day from 8:30 AM to 4 PM and books callbacks with their recruiters.
Here's how Beth works:
1. Every 10 minutes, she scans the database for leads that haven't been called yet.
2. She picks 10 in order and initiates 10 simultaneous calls.
3. Every outcome is logged into a Google Sheet that acts as a lightweight CRM for daily tracking.
4. There are 6 possible outcomes: conversation & booking made / conversation & no booking / warm transfer / referral / not connected / voicemail.
5. Based on the outcome, Beth automatically sends a tailored follow-up email using one of five pre-built templates.
6. Beth gets two call attempts per day — if the first goes unanswered, she calls back later the same afternoon.
7. If the candidate is interested, Beth books a 10-minute slot via Cal.com up to 3 days out — which triggers a scheduled recruiter callback, not a meeting.
8. A confirmation goes to the candidate, the recruiter, and the team manager automatically.
Results after two months:
✅ 6,300+ leads processed.
✅ 1,330 connected conversations.
✅ 518 callbacks booked.
✅ 38.9% booking rate from connected calls.
Hundreds of additional bookings driven by follow-up emails.
15 new bookings every single day, without fail, for two months straight.
Only 0.3% of candidates asked if they were speaking to AI.
Sentiment stayed positive on ~80% of calls.
Their recruiters now have a steady pipeline of warm candidates - without burning out on high-volume cold calling.
Huge thanks to Graham, Andy, Neil, Lisa, and Kelsey for their patience during the build and for helping us fine-tune Beth to get the most out of every call.
Brilliant results all round 🔥
If you're in recruitment and sitting on a database of old leads, this lean approach might give you some ideas.
Feel free to reach out.
