The best way to automate lead follow-up for landscapers

If you run a landscaping crew, your inbox is a leaky bucket. Somebody emails asking for a quote on a fall cleanup. You see it at the end of the day, say “I’ll get to it tomorrow,” and then Tuesday is on fire, and Wednesday two jobs overran, and by Friday they’ve already hired the other guy.

This is not a motivation problem. It’s a cadence problem. And it’s fixable.

The cadence that works

After working with a dozen landscapers, the cadence that consistently recovers the most lost leads looks like this:

• Hour 0 — acknowledge. Short. “Got it. I can swing by to quote next week — does Tuesday or Thursday work better for you?”
• Hour 24 — if they didn’t reply, nudge with the specific ask. Not “just checking in.” Something like: “Are mornings or afternoons easier for the walkthrough?”
• Hour 48 — one more try, different angle. Maybe social proof (“we just did your neighbors on Oak St”) or urgency (“only two slots left for spring cleanup”).
• Day 5 — final email. “Happy to keep you in mind for next season — just reply and I’ll put you on the list.” Then let go.

Four emails. Each one under 80 words. Each one asks a specific next step.

Most landscapers send one email and pray. That’s why they close 20% of quote requests instead of 40%.

Why templates alone don’t work

The obvious fix is a template file. But templates break down the moment a real email doesn’t fit the mold:

• “I need someone to cut down a tree AND do pavers” — two service types
• “What about organic fertilizer?” — a question, not a quote request
• “My address is 123 Main but the job is at 456 Oak” — two addresses

A pure template wants to substitute {{name}} and {{service}} and go. It doesn’t know what to do with the messy middle 60% of your inbox.

This is where an LLM pass — just one, narrow, bounded pass — earns its keep. Not to write the email. To read the email and pull the structured facts (service type, estimated size, address, urgency) before the template fills in.

The implementation, in five boxes

[Inbox] → [Classifier] → [CRM row] → [Drafter] → [Drafts folder]

1. Inbox watches one email address.
2. Classifier asks the LLM: is this a lead? What service? What data is missing?
3. CRM row writes the structured record to Google Sheets or Airtable.
4. Drafter picks the right template based on service type and writes a draft.
5. Drafts folder — not Sent. You read it. You click send.

That fifth step is the whole ballgame. The first two weeks, you’ll tweak the prompts, catch the classifier’s misses, and build trust. Then you start sending drafts two at a time without reading closely. Then you flip the follow-up workflow from draft-only to auto-send for your warm replies, because it’s been right 50 days in a row.

What to measure

Track two numbers:

• Reply latency — hours from inbound email to outbound response. Target: under 4 hours during business days.
• Quote-to-close rate — of the leads that get a quote, how many book. Most landscapers are at 40-50%. Cutting latency alone can move this 5-8 points.

Don’t bother measuring “number of emails drafted.” It’s a vanity metric.

The short version

If you want the whole kit — the classifier prompt, the four templates, the n8n workflow JSON, the lead schema — we ship it as Inbox Autopilot for Landscapers. Drop-in for Gmail or Outlook, Google Sheets or Airtable. Under 30 minutes to install.