Complete Refrigeration & Air
Process MapStage 1 · current
2026-05-05
Process Deep Dive · Quoting

How quoting actually works at CRAWA today.

A complete map of the as-is quoting process, including every person involved, every handoff between systems and channels, and every place where rework can send work back upstream. This is the Stage 1 baseline — built from the AroFlo data we have plus what Richard has described. Items that are confirmed vs. inferred are clearly labelled. The next step is staff verification before we move to Stage 2 (the optimised redesign).

Stages 1 · Map current state 2 · Design optimised version

§1Who's involved

Five distinct actors, each with a different surface (in-field, office, off-shore, customer, AroFlo itself).

Field tech

Onsite assessor

Diagnoses the issue, captures notes/photos, sends info back to the office. ~19 active tradesmen at CRAWA. Examples: Jayden Hosking, Rob Rumble, Michael Brennen.

Office expert

Validator & scoper

Reads what the tech sent, validates the diagnosis, decides what to quote on. Mostly Dave Katsaros (~21% of quotes); also Kurtis Dixon, Joel Baldini, Richard Barnes for escalations.

VA · Indonesia

Quote preparer

Anggaa Adisastra. Looks up parts, supplier prices, customer rate cards; assembles line items, drafts the narrative description, applies markup, formats. Named on ~70% of quotes in the AroFlo "estimator" field.

Customer contact

Decision maker

Reviews quote, accepts / rejects / asks for changes. Manager-access contact at the customer end. Example: Vance at Brownes Dairy.

Systems

AroFlo · Outlook · Teams

AroFlo holds Tasks, Quotes, Assets, Notes. Outlook handles customer comms. Teams / WhatsApp / phone handle internal handoffs that aren't formally tracked.

Field tech Office expert VA Customer Systems Rework loop

§2Process flow — the messy network

How the actual paths run between actors. Solid arrows are AroFlo / system flow; dashed arrows are off-system handoffs (phone, Teams, email, paper). Diamonds are decision points where the path can loop back. Three terminal states show how quotes really end (accept / reject / lapse).

flowchart TB classDef tech fill:#cffafe,stroke:#0e7490,color:#0e7490,stroke-width:1.5px classDef office fill:#ede9fe,stroke:#6d28d9,color:#6d28d9,stroke-width:1.5px classDef va fill:#ffedd5,stroke:#c2410c,color:#c2410c,stroke-width:1.5px classDef customer fill:#dcfce7,stroke:#166534,color:#166534,stroke-width:1.5px classDef system fill:#e2e8f0,stroke:#475569,color:#475569,stroke-width:1.5px classDef sched fill:#dbeafe,stroke:#1d4ed8,color:#1d4ed8,stroke-width:1.5px classDef decision fill:#fef3c7,stroke:#b45309,color:#78350f,stroke-width:1.5px classDef good fill:#dcfce7,stroke:#15803d,color:#15803d,stroke-width:2px classDef bad fill:#fee2e2,stroke:#b91c1c,color:#b91c1c,stroke-width:2px classDef neutral fill:#fef3c7,stroke:#b45309,color:#78350f,stroke-width:2px classDef inflight fill:#e0e7ff,stroke:#4338ca,color:#3730a3,stroke-width:2px Start([Customer call/email]):::system S0[Office schedules
tech site visit]:::sched T1[Tech onsite
assess + diagnose]:::tech T2[Tech notes
into AroFlo]:::tech O1[Office expert
reads notes]:::office D1{Diagnosis
clear?}:::decision O2[Office expert
validates + scopes]:::office O3[Hand off scope
to VA]:::office V1[VA looks up
parts / prices / rates]:::va D2{All info
known?}:::decision V2[VA assembles
line items + narrative]:::va O4[Office expert
reviews draft]:::office D3{Quote OK?}:::decision O5[Office expert
sends quote]:::office C1[Customer
reviews]:::customer D4{Decision?}:::decision S1[Office schedules
the work]:::sched T3[Tech assigned
to scheduled job]:::tech E1([Approved · ~38%
Job in progress]):::good E2([Rejected · ~8%]):::bad E3([Lapses · ~54%
open > weeks/months]):::neutral E4([In progress · 20% of all quotes
not yet completed]):::inflight Start --> S0 S0 --> T1 T1 -.->|"phone for urgent"| O1 T1 --> T2 T2 --> O1 O1 --> D1 D1 -->|"unclear
↺ rework"| T1 D1 -->|OK| O2 O2 --> O3 O3 -.->|"Teams /
email"| V1 O3 --> V1 V1 --> D2 D2 -->|"unknown SKU /
unclear scope
↺ rework"| O2 D2 -->|good| V2 V2 --> O4 O4 --> D3 D3 -->|"needs revision
↺ rework"| V2 D3 -->|OK| O5 O5 -.->|email| C1 C1 --> D4 D4 -->|"wants changes
↺ rework"| V2 D4 -->|"accepts"| S1 S1 --> T3 T3 --> E1 D4 -->|"rejects"| E2 D4 -->|"silence"| E3 V2 -.->|"awaiting
review"| E4
Five rework loops visible (D1 unclear-diagnosis · D2 missing-info · D3 quote-needs-revision · D4 customer-changes · plus silent lapse).

Outcome percentages are rebased so the three completed states sum to 100%, derived from the AroFlo status sample of n=105 quotes:
· Approved 38%  (30 of 79 completed)
· Rejected 8%  (6 of 79 completed)
· Lapsed 54%  (43 of 79 completed — currently labelled "Pending Approval" in AroFlo; in practice many never get formal closure)

A further 20% of all quotes are currently In Progress (drafted internally but not yet sent to the customer) and so haven't completed yet — they will eventually settle into one of the three states above.

§3The flow, end to end

Read top to bottom. Each row is one step in time. Coloured blocks show who's doing what; arrows show handoffs; dashed amber blocks are rework loops that sometimes happen.

Step · time
Field tech
Office expert
VA (Anggaa)
Customer
Systems
1Onsite10–30 min
Assesses, diagnoses, photographs
Often poor signal · types onto phone or paper
AroFlo Task already exists
Created by office at intake
2Hands offhours – days
Notes typed into AroFlo
Often back at office or end-of-day · informal phone/Teams to office expert about complex cases
tasknotes updated
Free-text · variable quality
3Reads & validates5–20 min
Reads tech's notes; validates diagnosis
Mostly Dave Katsaros · sometimes mental check, sometimes deep re-diagnosis
Tech queryhours – days
Office calls/messages tech for clarification — tech may need to revisit notes or even site
Office expert can't tell what was found; pings tech
4Scope decision5–15 min
Decides what to quote on
Customer-aware (rate card, history, tone) · sometimes adjusts diagnosis
5Hand off to VAhours
Assigns quote to VA
Likely via Teams / email + AroFlo task assignment · scope written somewhere
Task assigned to VA in AroFlo
Time-zone gap (Indonesia)
6Quote preparation30–90 min
Looks up parts, supplier prices, contract rates
May be working hours behind AU TZ · cross-checks unfamiliar SKUs
VA queryhours
Answers VA's questions — adds further latency
VA hits unknown SKU / unclear scope; pings office
7Quote drafted30–60 min
Assembles line items + narrative + markup
Drafts the "Dear ___" preamble, formats per CRAWA template
Quote saved · status: In Progress
8Office review5–15 min
Reviews VA's draft
Sense-checks pricing, scope, narrative tone · edits if needed
VA revision15–60 min
Sends back to VA with corrections
Adjusts scope/pricing/narrative; resubmits
9Send2 min
Sends quote to customer
Email or AroFlo Quote URL · narrative + line items + total
Quote → Pending Approval
Email sent via Outlook
10Customer reviewsdays – weeks
Receives quote · reviews · responds (or doesn't)
Often slow · 43% of all open quotes sit at this step today
Customer changesdays
Receives questions/changes; routes to VA for revised quote
Reworks quote per customer feedback
Asks for alternatives, partial scope, different pricing
11Outcomedays – weeks
Accepts / Rejects / Lapses
Accept → AroFlo Task scheduled · Reject → status changed · Lapse → no formal close
Quote → Approved · Rejected · or stuck
Total typical lead time: 3–7 business days from site visit to customer receiving the quote, with 4 humans in the chain. Add another 5–14 days to customer response (often longer). Rework loops can extend any of these substantially.

§4Step-by-step detail

Each step expanded — actor, action, system used, typical duration, and any rework triggers.

# Actor Action System / channel Typical duration Rework triggers
1 Field tech Onsite assessment + diagnosis In-person · phone notes · paper 10–30 min Re-visit if missed info
2 Field tech Notes into AroFlo (often delayed) AroFlo Field UI · informal phone/Teams to office for complex cases 5–15 min
(latency: hours–days)
Office can't read tech's notes
3 Office expert Reads notes; validates diagnosis AroFlo desktop 5–20 min Diagnosis unclear or implausible
3a Field tech Clarifies / re-checks site Phone call · Teams · re-visit + hours – days
4 Office expert Decides what to quote on AroFlo + tacit customer knowledge 5–15 min
5 Office expert Hands off scope to VA AroFlo task · Teams / email 5 min
(latency: hours · timezone)
6 VA · Anggaa Looks up parts / prices / rate cards AroFlo · supplier portals · spreadsheets 30–90 min Unknown SKU · unclear scope · rate card stale
6a Office expert Answers VA's questions Teams · email + hours (timezone)
7 VA · Anggaa Drafts quote (lines + narrative + markup) AroFlo Quote · CRAWA template 30–60 min Markup rule unclear · narrative needs polish
8 Office expert Reviews VA's draft AroFlo Quote 5–15 min Pricing / scope / tone needs adjustment
8a VA · Anggaa Revises and resubmits AroFlo Quote + 15–60 min
9 Office expert Sends quote to customer Outlook · AroFlo Quote URL 2 min
10 Customer Receives, reviews, responds Email · phone days – weeks Has questions · wants alternatives · cost too high
10a Office expert Routes feedback for revised quote Email · Teams · AroFlo + days
11 Customer Accepts / Rejects / Lapses Email reply · phone · silence days – weeks Lapsed quotes never formally closed

§5What's not in AroFlo (the shadow process)

Plenty happens during this process that AroFlo never sees. These are the bits the API can't tell us about — only staff conversations can.

Channels carrying decisions
  • Phone calls — tech ↔ office for diagnosis clarification, urgent dispatches
  • Teams chat — office ↔ VA for scope handoff, supplier queries, customer-aware nuance
  • WhatsApp / SMS — possibly, especially for tech-side urgent comms
  • In-person — office staff turning around to ask each other directly
  • Email — customer-side comms, occasionally internal
  • Paper — tech notes that haven't yet made it into AroFlo
Tacit knowledge that never gets logged
  • "This customer always wants fixed-price, never T&M"
  • "That supplier's prices went up 12% last month"
  • "Brownes' Vance won't accept anything without an alternative quote"
  • "The drain heater on this model class fails about 1 in 8"
  • "Anggaa's been struggling with the Bromic-EU pricing structure"
  • Office expert's mental list of "this tech tends to under-scope; that one tends to over-quote"
Information loss at handoffs: every transfer between actors is a potential point where context is dropped or distorted. Tech → office expert → VA → office expert → customer is four handoffs. By the time the quote reaches the customer, it has been re-summarised three times. If this is happening over Teams/email/phone, the audit trail is fragmented.

§6What we know vs. what we're guessing

Honest accounting before staff verification.

✓ Confirmed by data

  • 5 actors involved in the chain (Richard confirmed VA exists; data confirms estimator distribution)
  • VA Anggaa Adisastra is named on ~70% of quote estimator fields (n=105 sampled)
  • Dave Katsaros is the most-named office expert on quotes (~21%)
  • Quote statuses: In Progress 20% / Pending Approval 43% / Approved 30% / Rejected 6%
  • 43% of open quotes sit in Pending Approval — customer-side latency is a real problem
  • Quote → Task linkage is universal (every quote has task.taskid)
  • Most CRAWA quotes are type: Simple — no multi-stage quoting
  • Customer rate cards visible on Tasks (e.g. "Brownes contract rate WAB-RF-01")

⚠ Inferred — needs staff verification

  • Tech-to-office handoff is informal (phone / Teams) — not visible in AroFlo
  • Office-to-VA handoff is informal — Teams / email assumed
  • VA spends 30–90 min per quote — not measured directly
  • Office expert spends 5–20 min validating + 5–15 min reviewing VA's draft
  • Rework rate at each step (3a, 6a, 8a, 10a) — not measured
  • Customer-side response latency (days–weeks) — confirmed for Pending Approval, but median time-to-response not yet measured
  • Whether Anggaa does any work beyond mechanical assembly (supplier negotiation, narrative writing, customer relationships)
  • Whether techs ever do their own quote drafting for routine work, bypassing office/VA entirely
  • What happens to quotes that lapse without being formally rejected

§7Open questions for staff verification

Take these into the team conversations. Answers shape Stage 2 directly.

  1. Is the actor list complete? Are there other people in the chain we haven't named — supplier reps, tender managers, accounts, anyone who reviews quotes before they go out?
  2. How does the tech actually hand off to the office expert? Is it AroFlo notes only, phone, Teams, or all of the above? Which channel carries the most context?
  3. How does the office expert hand off to Anggaa? Is the scope written in AroFlo, or sent over Teams/email separately? How does Anggaa know which job is next?
  4. How often does Dave (or whoever the expert is) override the tech's diagnosis? Is it a routine sense-check or do real corrections happen?
  5. How often does the office expert send a quote back to Anggaa for revisions? What's typically wrong — pricing, scope, narrative tone?
  6. How long does Anggaa typically spend per quote? Routine vs complex.
  7. What does Anggaa do beyond mechanical assembly? Does he query suppliers? Negotiate? Build customer relationships?
  8. How often do customers come back with questions or change requests? What % land first time vs. need a rev2?
  9. What happens to quotes that lapse? Does anyone follow up? Or do they just sit there?
  10. Are there quote types where the tech can quote directly onsite (no office/VA)? Standard service calls, like-for-like replacements?
  11. Are there quote types where the office expert always escalates further? High-value, government, unusual scope — what's the threshold?
  12. Where is the most painful step today? If you could remove one bit of friction, which would have the biggest impact?

§8Where this goes next (Stage 2 preview)

Once Stage 1 is verified, the optimised redesign collapses the chain.

Today: 4 humans, 3–7 days
  • Field tech (capture)
  • Office expert (validate + scope + handoff)
  • VA Anggaa (assemble quote)
  • Office expert (review + send)
Stage 2: 1 human, minutes
  • Field tech (AI-assisted in-field capture)
  • Agent (assembles draft, ranks diagnoses, applies rates/markup, drafts narrative)
  • Office expert (single review + approve)
  • Anggaa redeployed to higher-leverage queue work
Stage 2 will be drawn the same way as this page — same actor swim-lane format, same step-by-step table, with explicit annotation showing which steps were collapsed, which moved to the agent, and which human gates remain.