Lesson 4 of 7 · 9 min

Pipeline stages: lead, estimate, sold, done

Design a pipeline that follows the job: lead, estimate, sold, done.

Read the lesson

Goal: Map the client's sales steps onto ServiceAgent's pipeline stages and read a weighted forecast from it.

You need: Settings → Contacts & Pipeline → Pipeline stages. The workshop sheet. Lessons 1 to 3.

Why this matters

A pipeline answers two questions the owner asks every week. What is out there, and what will close? Without it, the answers live in the owner's head or a spreadsheet on one laptop.

A pipeline also drives follow-up. Workflows react when a deal is created or its status changes. So a clean pipeline is not just a report. It is the switch that starts estimate follow-up, task creation and owner alerts.

Set the size line with the owner. Write it on the build sheet, such as "a deal for any job quoted over $2,000". Without a line, one office person makes deals for everything and another makes them for nothing. Then the board means nothing.

Not every job needs a deal. A $150 drain clearing booked on the phone goes straight to the calendar. A $12,000 roof, a new furnace or a panel upgrade needs a deal. Those take days or weeks, and they slip away without follow-up.

Fig 1

Does this job need a deal?

  1. A job is quoted or asked about
  2. Is it over the agreed size, such as $2,000?YesCreate a deal in NewNoBook it straight on the calendarEnds here
  3. The deal moves through the stages to Won or Lost
Agree the size line with the owner and write it on the build sheet.

The default stages, and how a service sale maps onto them

ServiceAgent's default pipeline stages are New, Qualified, Proposal, Negotiation, Won and Lost. They fit a service sale well. Keep the names. Workflows and Build with Emma prompts must use the exact stage names, so renaming adds risk.

StageWhat it means in a service businessWhat moves it on
NewA lead asked about bigger workSomeone talks to them
QualifiedA site visit or estimate appointment is bookedThe visit happens and an estimate is written
ProposalThe estimate is sentThe customer answers
NegotiationThe customer is deciding: options, financing, timingA yes or a no
WonThe customer said yes in writingThe job is booked
LostThe customer said no, or went silent past your limitNothing

"Done" is not a pipeline stage. Once a deal is Won, the job lives in Jobs and the money lives in Payments. A job is done when the appointment is completed and the invoice is paid. Do not add a "Done" stage after Won. It splits your win numbers.

Fig 2

After Won, the job leaves the pipeline

  1. PipelineDeal moves to Won
  2. JobsJob booked on the calendar
  3. JobsAppointment Completed
  4. PaymentsInvoice sent
  5. PaymentsInvoice paid: the job is done
A job is done when the appointment is completed and the invoice is paid.

Why Lost needs a reason

Lost is an ordinary stage. There is no built-in lost-reason field. Without a reason, you cannot learn why deals die.

So add a custom field on the Deal entity: lost_reason, a Dropdown. Good answers: Price, Chose a competitor, Timing, Financing, No response, Not a fit, Other. The office picks one every time a deal moves to Lost.

Fig 3

A lost reason the office can count

Add Field: lost_reason (Dropdown)

Keylost_reason
Apply to Entity TypesDeal
Help TextPick one: Price, Chose a competitor, Timing, Financing, No response, Not a fit, Other
Add Field
There is no built-in lost-reason field. One Dropdown on the Deal turns losses into lessons.

Probability and the weighted forecast

Each deal has a value and a probability. The Deals filter can even show deals above a Minimum Probability (%). The probability is your best estimate that the deal closes.

Each stage has a probability too. The app's defaults are New 10%, Qualified 25%, Proposal 50%, Negotiation 75% and Won 100%. A deal shows its own probability if one is set. If not, it shows its stage's probability.

Agree one probability per stage with the owner, and change each stage to their number. Leave the deal's own probability empty, so the deal follows its stage when it moves. Set a deal's own number only for a real exception.

The weighted forecast is the sum of each deal's value times its probability. It is a better guess of what will close than the raw total. It does not predict any single deal. It predicts the group.

Weighted forecast = sum of (deal value x stage probability).

Start with the owner's guesses for each stage. After three months, check them against Win Rate in Contacts → Reports (lesson 6). Adjust the rules to match what really closed.

Getting deals into the pipeline

Deals come in four ways:

  1. By hand. On a contact, use Create deal. Or use Convert to Deal from the contact row menu.
  2. By workflow. A Create Deal step can open a deal, for example when a form asks for a replacement quote.
  3. By stage changes. An Update Deal Stage step can move a deal. For example, it moves a deal to Qualified when an estimate visit is booked.
  4. From the voice agent's work. The agent saves leads and creates tasks. The office turns real sales leads into deals.

Workflows can start from the events Deal Created, Deal Updated and Deal Status Changed. That is how lesson 5 creates tasks at each stage.

Build it in ServiceAgent

  1. Go to Settings → Contacts & Pipeline → Pipeline stages.
  2. Check the stages: New, Qualified, Proposal, Negotiation, Won, Lost. Check the Order.
  3. Set a Color for each stage, so the board reads at a glance.
  4. Do not rename stages unless the client insists. If you rename, record the new names for every workflow.
  5. Go to Custom fields. Add lost_reason as a Dropdown on the Deal entity.
  6. Agree the probability per stage with the owner. Set it on each stage, and write it on the build sheet.
  7. Open Contacts → Deals. Create 3 test deals. Move each one through the stages.
  8. Check it works. Filter by Stage = Proposal. Check values and probabilities. Work out the weighted forecast by hand.

Running the weekly pipeline review

A pipeline only stays true if someone looks at it every week. Book 20 minutes with the owner or sales lead, on the same day each week.

Work the board from right to left. Start with Negotiation, because those deals are closest to money. For each deal, ask one question: "What is the next step, and when?" If nobody knows, that is the finding. Create a task with a person and a date.

Then look at Proposal. Any deal that has sat there more than 14 days without a touch gets a phone call this week. Then look at New and Qualified. Any lead with no visit booked after 3 days gets a call too.

Last, look at the deals that moved to Lost this week. Read their lost_reason values. One lost deal is noise. Three lost deals for the same reason in a month is a pattern worth fixing.

Write down the weighted total each week. A short table of weeks and totals shows the owner whether their sales pipeline is growing or shrinking. That one number often matters more to them than any chart.

Fig 4

The weekly review, right to left

  1. 1NegotiationWhat is the next step, and when?
  2. 2ProposalCall any deal untouched for 14 days
  3. 3New and QualifiedCall leads with no visit after 3 days
  4. 4LostRead lost_reason; 3 alike is a pattern
  5. 5Weighted totalWrite it down for the weekly table
20 minutes, same day each week. Every deal leaves with a next step and a date.

Size changes the pipeline

An owner-operator may need no pipeline at all. If most jobs are booked on the first call, the calendar is enough. Add a pipeline only for jobs over a set size, such as $2,000.

A fleet with a sales team needs the full pipeline. Each estimator works their own deals. The weekly sales meeting runs from the board.

A multi-location business can share one pipeline. Tag each deal's contact with its location, so each manager can see their own deals.

Worked example: Summit Roofing

Summit Roofing agrees these probabilities with the owner: New 10%, Qualified 20%, Proposal 30%, Negotiation 70%.

Summit has 3 estimators. Start with one estimator's open deals this week:

  • Proposal: 6 deals at $12,000 each.
  • Negotiation: 3 deals at $11,000 each.

Raw total:

  • Proposal: 6 x $12,000 = $72,000.
  • Negotiation: 3 x $11,000 = $33,000.
  • Total: $72,000 + $33,000 = $105,000.

Weighted forecast:

  • Proposal: $72,000 x 30% = $21,600.
  • Negotiation: $33,000 x 70% = $23,100.
  • Total: $21,600 + $23,100 = $44,700.

The owner had been planning that estimator's crew around $105,000. The weighted number, $44,700, is closer to what will close. They plan crews around the weighted number, not the raw one.

Fig 5Example numbers

One Summit estimator's weighted forecast

Proposal: 6 × $12,000 = $72,000 × 30%$21,600
+Negotiation: 3 × $11,000 = $33,000 × 70%$23,100
Weighted forecast$44,700
The raw total is $105,000. The weighted total is a better guess of what will close.

After three months, Win Rate shows Proposal deals close at 25%, not 30%. You lower the Proposal rule to 25%. With the new rule, the same 6 Proposal deals weigh $72,000 x 25% = $18,000. The forecast becomes $18,000 + $23,100 = $41,100.

Fig 6Example numbers

One estimator's raw total vs weighted forecast

Raw total of open deals$105,000
Weighted forecast$44,700
Weighted, Proposal at 25%$41,100
After three months, Win Rate lowered Proposal to 25%. The forecast moved to $41,100.

Common mistakes

  • A deal for every small job. The board fills with noise. Fix: deals only above an agreed size.
  • A "Done" stage after Won. Fix: track done in Jobs and Payments.
  • Lost with no reason. Fix: a required lost_reason Dropdown on the Deal.
  • Renaming stages after workflows exist. Steps point at names that no longer exist. Fix: keep defaults, or update every workflow.
  • Probabilities nobody checks. Fix: compare them with Win Rate every quarter.

Do this now

Write the stage map and probability rule for one real or practice client. Create 6 test deals across the stages. Work out the raw total and the weighted forecast by hand. Check them twice.

Practice the conversation

  1. Operator What's out there right now? In quotes, I mean.
  2. Owner My estimator Dan has about a hundred grand in roof quotes.
  3. Operator How much of that will close?
  4. Owner Hard to say. Half, if I'm lucky.
  5. Operator The pipeline will say it with numbers. This week, Dan's weighted total is about $44,700.
  6. Owner That's less than I thought.
  7. Operator It's what usually closes from those stages. We'll check it against real wins each quarter.
  8. Owner And the ones we lose?
  9. Operator Your office picks a reason every time. In three months, you'll know why you lose.

Quick check

  1. A roofer has 8 deals in Proposal at $10,000 each, weighted 30%. It has 4 in Negotiation at $9,000 each, weighted 90%. What is the weighted forecast?
    • $24,000
    • $32,400
    • $56,400
    • $116,000
  2. The owner wants a "Done" stage after Won. What do you advise?
    • Add it, so finished jobs show on the board
    • Rename Won to Done
    • Add Done and remove Lost
    • Keep the pipeline at Won; track done in `Jobs` and `Payments`
  3. Where do you record why a deal was lost?
    • In a Dropdown custom field on the Deal, such as `lost_reason`
    • In the Lost stage's built-in reason field
    • In the contact's `Address`
    • Nowhere; Lost is enough

ServiceAgent UniversityLearn to run a service business on autopilot. Free courses, Blueprints and certification for Service Operators.

Learn

Get certified

Community