How Macro Biz AI Works

You don't need to automate everything.Start with what's getting missed.

Some parts of your business work well. Some rely on someone remembering. Some opportunities simply disappear.

Macro Biz AI starts by finding those gaps. Then we work out where AI and automation genuinely make sense.

We don't start with the AI.We start with how your business actually works.

Before recommending anything, we need to understand how customers contact you, what they usually want, what happens next, what your team handles manually, where things get delayed, what gets forgotten, what software you're already using, and which decisions require a person.

Only then does it make sense to decide where AI or automation belongs.

8The Implementation Methodology
DWG.08 · Problem First · Human Judgement · Trade Tested
01

Your Business

How work currently flows

02

Find Gaps

Where calls & leads drop

03

AI Scope

Predictable, repeatable tasks

04

Keep Human

Preserve trade judgement

05

Connect

Work with existing systems

06

Build & Test

Stress-test the edge cases

07

Start Small

Solve one painful problem

08

Review

Refine on real customer calls

Technology comes after the problem. Not before it.Macro Biz AI Implementation Rule
Step 01

We start with your business.Not the software.

We look at how work currently moves through the business: phone calls, website enquiries, SMS, bookings, quotes, follow-up, customer records, job management, staff handovers and existing systems. The aim isn't to redesign everything. It's to understand what's already happening.

Questions search

Questions We Might Ask First

To build an accurate picture of your operations, we unpack everyday customer scenarios:

?What happens when the phone rings while everyone is busy?
?What happens if nobody answers in time?
?Where do website and online form enquiries go?
?Who follows them up and how quickly?
?How are bookings and diary slots confirmed?
?What happens after a quote is sent out?
?What repetitive questions does your team answer daily?
?What information do customers have to provide every time?
?Where is customer information and history stored?
?What parts of your current process already work well?
Core Standard

"We don't want to replace something simply because it isn't AI."

If your booking diary or whiteboard works, we respect it. If your reception process runs smoothly when staff are present, we don't interfere. We only focus on the drop-offs.

Step 02

Then we find the gaps.The places where customer communication breaks down.

In almost every trade or service business, customers slip away at familiar moments. These are the gaps we're looking for:

Phone Call Icon

The Call That Wasn't Answered

Customer calls. Nobody is available on the tools.

Nothing happens until someone remembers to call back.

Identifiable GapSolvable
Opportunity stalled clock emblem

The Enquiry That Sat There

Customer submits a website form or sends an SMS.

Someone sees it hours later. By then, the customer has booked a competitor.

Identifiable GapSolvable
Quote notification icon

The Quote Nobody Followed Up

Detailed quote sent. No immediate reply.

Everyone assumes the customer wasn't interested. No one checks in.

Identifiable GapSolvable
Calendar Icon

The Booking That Takes Three Calls

The customer wants to book, but vital details are missing.

Multiple phone tag messages back and forth before anything is confirmed.

Identifiable GapSolvable
Question Mark Icon

The Same Question For The 20th Time

Trading hours, service area, accepted waste, base pricing.

Your team keeps stopping productive work to answer predictable basics.

Identifiable GapSolvable
Past Customer Reactivation Icon

The Customer Who Never Came Back

Years of past customer history sitting in your software.

Nobody has looked at whether there's a sensible lifecycle reason to reconnect.

Identifiable GapSolvable

These are the gaps we're looking for.

Step 03

Just because it can be automated…doesn't mean it should be.

Once we've found a gap, the next question is: Which part should technology handle, and which part still needs a person? We enforce strict boundaries between predictable tasks and human trade expertise.

AI Receptionist Icon

AI / Automation Can Be Useful For

Predictable, rules-based tasks

  • Routine customer questions and service scope
  • Information gathering (name, vehicle, address, waste type)
  • Automated acknowledgements and booking confirmations
  • Appointment and drop-off reminders
  • Predictable, structured follow-up on cold quotes
  • Direct booking requests into verified diary windows
  • Straightforward workflow steps and customer routing
  • Repetitive administration and basic qualification
Person Icon

People Are Better For

Trade experience and human judgement

  • Trade judgement and real-world mechanical diagnosis
  • Technical advice, safety and compliance recommendations
  • Unusual situations or non-standard site requirements
  • Complex, high-value and custom job quotations
  • Customer complaints and delicate dispute handling
  • Sensitive conversations requiring empathy and care
  • Safety, hazardous waste or council permit exceptions
  • Relationship-critical commercial client conversations

"AI handles the predictable. Your team handles the exceptions."

Step 04

No ripping everything out just to add AI.Keep what works. Add what's missing.

Your existing software may already do an important job well: customers, vehicles, jobs, quotes, bookings, dispatch, calendars, invoices and service history. We don't automatically replace it. Instead, we look at whether AI and automation can work around it or with it.

Incoming Interactions

Customers

Calls · SMS · Website Enquiries · Booking Requests

↓
Customer Layer

Macro Biz AI

Answer · Gather Details · Follow Up · Book · Route Exceptions

↓
Operating Core

Your Business

Your Team + Your Existing Software (Running Operations)

The Practical Integration Reality

Some systems are easier to connect with than others. Some have open APIs. Some support webhooks or direct connectors. Some may require a different approach. And occasionally, the right answer is: “This system can't reliably do what we want it to do.” That's useful information too.

Verified / ProtectedOur Integration Standard

"We'd rather tell you an integration isn't ready than pretend two systems work together when they don't."

Step 05

Then we build the workflow.Around your specific trade rules, not generic templates.

Once we know the gap, the software and where AI should stop, we build the actual sequence. We define: what questions to ask, what information is essential, what answers change the next step, where information needs to go, what triggers follow-up, when automation stops, and when your team takes over.

Example: Workshop CallMotor Mechanics
01Customer Calls
Inbound call answered by AI
02New or Existing?
Checks caller name or phone in database
03Which Vehicle?
Rego / Make / Model verified cleanly
04Service · Fault · Other?
Identifies routine logbook vs breakdown
05Preferred Day
Checks diary availability or acceptable window
06Book · Request · Handover
Book direct or forward full context to service advisor
Example: Skip Bin EnquirySkip Bins & Waste
01Customer Enquires
Call, website form, or SMS message
02Where Is The Site?
Suburb verified against your delivery zone
03What Waste Type?
General, green, heavy, cleanfill or prohibited
04What Size Bin?
Guidance on 2m³ to 10m³ based on project scale
05Where Will It Go?
Private property vs council land (permit rules)
06When Do You Need It?
Drop-off date, duration, and pickup window

Illustrative workflows. Actual capabilities depend on your business rules, available software and supported integrations.

Same underlying technology. Different trade business rules.

Step 06

Then we try to break it.Before your customers do.

A workflow isn't ready just because the obvious example works. We need to know what happens when the conversation doesn't follow the script.

Edge cases alert

12 Real-World Edge Cases We Test

Every workflow is stress-tested against these exact scenarios:

•Customer gives an unexpected or ambiguous answer
•Crucial information (rego, address, waste type) is missing
•Customer changes their mind mid-conversation
•The requested booking date or bay is completely full
•Caller is already an existing account with special notes
•Multiple matching customer records exist in software
•Customer cannot be found in existing database records
•Customer asks technical questions the AI does not know
•Customer explicitly asks to speak to a real person
•Customer makes an urgent or unhappy complaint
•Request falls outside normal operating or delivery rules
•One of the connected software systems goes offline
The QA Cycle
BUILD
↓
TEST
↓
FIND THE EDGE CASES
↓
ADJUST
↓
TEST AGAIN

"The exceptions are where you find out whether the workflow actually works."

Step 07

Don't automate the whole business on day one.Start somewhere useful.

Where possible, we prefer to start with a clearly defined problem. That gives us a chance to see how customers respond, how the team uses it, what gets handed over, what needs adjusting, whether the workflow is actually saving work, and where the next opportunity might be.

Phase 01

START SMALL

One high-friction pain point

Phase 02

USE IT

Live in your real day-to-day trade

Phase 03

REVIEW

Look at actual call & lead logs

Phase 04

ADJUST

Tighten answers and boundaries

Phase 05

EXPAND

Add more workflows where sensible

Build around reality, not the perfect demo.

Different businesses have different gaps.Here are three common places to start.

You don't need all three. Most businesses know immediately which one hurts most:

Starting Point 01

"We're missing calls while we're on the tools."

Start with: AI Receptionist

Answer common calls, collect vehicle/job details and keep enquiries moving so you never lose a job simply because everyone was busy.

Starting Point 02

"Quotes and enquiries go quiet."

Start with: Lead Follow-Up & Bookings

Keep enquiries moving automatically with polite, timely follow-up before the customer drifts to someone else.

Starting Point 03

"We have thousands of past customers sitting there."

Start with: Past Customer Reactivation

Look for useful customer groups and test whether there's a sensible lifecycle reason to reconnect without spamming your database.

What Happens Next

You don't need to know which AI solution you need.Tell us what's happening in the business.

Get Started begins with a few practical questions about your business, what's getting missed, how things work now, your existing software, and the volume you're dealing with. Then we review where the most sensible starting point might be.

Step 01

You Tell Us What's Happening

A short, guided diagnostic about your trade and software.

Step 02

We Review The Gap

We look at where calls, quotes, or diary slots fall through.

Step 03

Sensible Starting Point

We identify one specific area worth testing first.

Step 04

Recommendation & Proposal

Personalised video or written proposal with clear pricing.

Step 05

Controlled Rollout

Build, rigorous testing, and gradual implementation.

No high-pressure sales calls required.

You won't be trapped into booking a discovery call before getting an answer. If your situation is clear, we provide a personalised recommendation or video walkthrough first. A conversation is always available when the setup genuinely calls for it.

Not sure where AI fits?That's what we work out first.

You don't need to map the automation yourself. Tell us what's getting missed, what you're already using and where the pressure is showing up. We'll start there.

A few questions first. No need to book a sales call just to find out whether there's a fit.