LSAs can absolutely print money for plumbers when the setup's dialled in - tight targeting, right service categories, fast call answering, solid reviews, and a booking process that actually converts. One bad £109 lead doesn't kill a channel.
Honestly, this screams bad lead qualification, wrong service area, weak profile optimisation, low review count, slow response time, or a mismatch between the jobs you want and what Google's algorithm is feeding you. And "free estimate" isn't a useless signal - every home services provider does it. It's table stakes, not a competitive weakness. The real money comes from backend work, referrals, emergency call-outs, water heater swaps.
Thumbtack and Nextdoor can work, but the bidding model inevitably pushes you into a price war. That gets ugly fast.
Before writing off Google LSA, I'd fix the offer, targeting, reviews, call handling, landing pages, follow-up, and booking rate. The bottleneck is almost always conversion math, not the source of traffic.