Read a complete audit before you buy one.

The business, as described in the intake

Two people: the owner on the tools full-time, plus one worker four days a week. Around 70 recurring maintenance clients on two-, three- or four-week cycles, plus one-off jobs — clean-ups, hedges, small landscaping — worth about $450 on average. Recurring visits average $180.

The stack is the common one: a personal mobile, Google Calendar, quotes agreed by text or on-site, invoices typed into Xero at night, and a whiteboard for the week's jobs. No CRM, no job-management software, no shared record of enquiries. It mostly works — until the phone rings while both people are on mowers.

Stated assumptions used throughout (from the intake questionnaire): ~35 inbound calls a week; about 60% get answered live; roughly 5 of the missed calls each week are genuine new-work enquiries; ~4 hours a week on quoting admin; ~3 hours a week on invoicing; a full rain day forces about 8 bookings to move.

01

Workflow pain map

How work actually moves through the business today, with the friction marked.

1. Enquiry arrives — phone call to the owner's mobile, or word of mouth. Friction: ~14 of 35 weekly calls go to voicemail while the owner is on the tools. Voicemail greeting is the phone-company default; most callers hang up without leaving a message and there is no record they ever called.
2. Quote — owner visits or prices from a photo, replies by text when there's time, usually that night. Friction: no template, no follow-up. A quote that gets no reply is never chased. Nobody knows how many quotes are outstanding.
3. Booking — agreed by text, entered into Google Calendar by the owner. Friction: single point of entry. The worker can't see changes until the owner updates the calendar, which happens at night.
4. The work — crew does the job. Minimal friction. This part runs well and needs no technology.
5. Weather reschedules — rain day: the owner calls or texts each affected client between jobs or at night. Friction: ~8 bookings to move, one at a time, about 6 minutes each — roughly 48 minutes of admin per rain day, and the occasional job that falls through the crack entirely.
6. Invoice & payment — typed into Xero at night, sent by email; follow-up when the owner remembers. Friction: ~3 hours a week, done tired, sometimes days after the job — which is also when payment disputes start.

The pattern worth naming: nothing in this business is broken during the work. Every leak is in the capture and follow-through around the work — missed enquiries, unchased quotes, slow rescheduling, late invoices. That is good news, because capture and follow-through are cheap to fix.

02

Top three improvement opportunities

1. Missed-enquiry capture

   About 5 genuine new-work enquiries a week currently reach voicemail. Scenario arithmetic, not a promise: if a proper missed-call routine recovers even 3 of the 5, the 2 still lost represent 2 × $450 × 40% win rate = $360 a week in expected work — about $18,720 a year — still on the table. Before the fix, all 5 are at risk. This is the single biggest number in the business and it costs nothing to address (Section 03).

2. Quote templates and a 48-hour follow-up

   Quoting is ~4 hours a week of night admin with no consistency and no chase. A standard quote format (three sizes: mow-and-tidy, clean-up, landscaping) plus one follow-up text at 48 hours turns "sent and forgotten" into a managed pipeline. The follow-up is one saved template, sent from a five-minute Friday review of the outstanding-quotes list.

3. Weather-day batch rescheduling

   Rain days are predictable chaos: ~48 minutes of one-at-a-time calls, done while wet. A calendar tag for weather-flexible jobs plus one saved broadcast message brings a rain-day reshuffle down to about 10 minutes, and the affected clients hear earlier, which they read as professionalism rather than cancellation.

03

The recommended quick win: a missed-call routine that costs $0

The obvious pitch here is an AI receptionist. It is not the first move, and it's worth being clear about why: at this call volume, a managed voice agent costs more than the problem once a free routine is in place. The honest first fix uses the phone already in the owner's pocket.

1. Rewrite the voicemail greeting so a missed call still captures the job: ask for suburb, the kind of work, and a photo by text. Callers who text a photo have self-qualified — they're serious.
2. Set up the one-tap declined-call reply (both iPhone and Android support responding to a declined call with a saved text) using the template below, so a call missed on the mower still gets a response within seconds.
3. Keep a shared callback list — a pinned note or spreadsheet both people can see — and protect 15 minutes at 4:30pm to work through it. Same time every day, so callers can be told exactly when to expect the return call.

When a voice agent becomes worth it: a managed Standard-tier agent runs $549 a month — about $127 a week. If, after four weeks of keeping the callback list, the log shows more than roughly 5 new-work enquiries a week still being lost (not just missed — lost), the arithmetic starts to favour the agent and it's worth a conversation. Below that, keep the $127 and the routine. The callback list is also the evidence either way — that's why it comes first.

04

Reusable templates and prompts

05

Implementation roadmap

• Week 1: Missed-call routine — greeting, one-tap reply, shared callback list, 4:30 slot. One evening of setup.
• Weeks 2–3: Quote templates and the Friday outstanding-quotes review with the 48-hour follow-up text.
• Week 4: Weather tags in the calendar and the rain-day broadcast message. Test on the next wet day.
• Month 2: Same-day invoicing using the voice-note prompt on the drive home (passenger or parked) so Xero entry becomes a 20-minute batch, not a 3-hour night.
• Month 2 checkpoint: Review the callback log. Decide on job-management software (see Section 06) and whether the voice-agent arithmetic in Section 03 has changed.

What stays manual, deliberately

• Pricing. Every quote number is the owner's. No tool sets a price.
• The first conversation with a new client. That call is where trust and scope are set — it should be a person for as long as volume allows.
• Complaints and anything that went wrong. Automation ends the moment a client is unhappy.
• Quality checks on site. Obviously.

06

Tool recommendations

Start with what's already paid for. Lumomatics has no affiliate or reseller arrangements with any tool named here.

• Keep: Xero (invoicing works; the problem is when invoices get typed, not where), Google Calendar (fix is shared visibility and tags, not replacement), the existing mobile.
• Add now: nothing paid. The quick win and the first month of the roadmap are deliberately $0 so the callback log can generate real numbers first.
• Decide at the Month-2 checkpoint: a job-management layer, once the enquiry volume is visible. For a crew this size the realistic candidates are a trades-specific tool or a configurable base like Airtable — the trade-offs are covered in the ServiceM8 vs Airtable comparison, and the callback log from Week 1 is exactly the data that makes that choice obvious.
• Skip for now: a CRM migration (nothing to migrate yet), a website chatbot (enquiries are phone-first), an AI email assistant (email volume doesn't justify it), and a managed voice agent (revisit only if the Section 03 threshold is met).