A typical mid-sized construction or trades business receives 50–150 inquiries a month: quote requests through the website, e-mails, WhatsApp messages, calls noted on paper. The owner knows roughly what is in the pipeline. Nobody knows exactly.
This is the most common starting point we see — and the most rewarding to automate, because the chaos is in the process, not in the people.
What to automate first
Not everything at once. The highest-leverage first step is intake: the moment an inquiry arrives until it exists as a structured record with an owner. That is where leads silently die today.
Concretely: every channel (web form, e-mail inbox, WhatsApp export) feeds one workflow. AI reads each message, classifies it — new project, maintenance request, supplier offer, spam — scores its value and urgency, and writes a clean record into the CRM with a summary the owner can read in ten seconds.
The blueprint, in phases
Phase 1 — Capture: connect the web form and the shared inbox to the workflow. From day one, nothing arrives unlogged.
Phase 2 — Classify and score: AI separates a €200k project inquiry from a €200 repair, and both from spam. High-value leads notify the owner immediately.
Phase 3 — Follow up: if a quote has been sent and the customer is silent for five days, the system creates a follow-up task — or drafts the follow-up e-mail for approval.
Each phase works on its own. You stop where the value stops, and most companies see the difference within the first two weeks.
What stays human
Site visits, estimates, pricing, negotiation — everything that wins the job stays with your team. The system's job is narrower: make sure every inquiry is seen, sorted and followed up, so the team spends its time on the leads that matter.
Realistic outcome for an 80-inquiry month: response time under one hour instead of days, zero lost inquiries, and a pipeline the owner can read on one screen.