Estimates with win and loss outcomes: why option-based proposals matter

What estimate win and loss data contains#

Estimate win and loss data is the full history of proposals your team presented, each one closed out with an outcome. An estimate sitting in open status for years is a draft; an estimate marked sold, lost, declined or expired, with a date and a reason, is a decision record.

Most field service platforms, including ServiceTitan, Housecall Pro, Jobber and FieldEdge, store estimates as their own object tied to a customer, a location and often a job. The fields are usually there. What tends to be missing is the discipline of closing each estimate with an honest outcome.

• 1Estimate ID, the visit or job it came from, and the person who presented it.
• 2Each option presented, with its equipment, scope, warranty terms and price-book items.
• 3Financing or membership discounts offered alongside each option.
• 4The chosen option, or a lost or declined status if none was chosen.
• 5A reason code and a short note explaining the outcome.
• 6Dates for presentation, follow-ups and the final decision.
• 7Links to the resulting job, invoice and any later callback or warranty claim.

Why option-based proposals carry more signal than single quotes#

Option-based proposals carry more signal because the customer compares several priced alternatives for the same problem and picks one. A single-price quote only tells you yes or no. A good-better-best proposal tells you which trade-off the customer made: repair or replace, upfront cost or efficiency, a basic warranty or an extended one.

Over hundreds of proposals, those choices describe how your customers actually decide. You can see which tier sells for older equipment, whether financing changes the chosen option, and which comfort advisors present all options rather than leading with one.

Illustrative: one estimate record from start to outcome#

Illustrative: Ridgeline Comfort, a fictional residential HVAC company, runs replacement sales through its field service platform. After a no-heat call, a technician flags a cracked heat exchanger on an aging furnace and a comfort advisor builds a three-option proposal the same week. The record below shows the fields the company decided to require on every estimate.

Ridgeline also reviewed a lost estimate from the same month. The customer chose a competitor, and the reason code showed the competitor offered a shorter install date, not a lower price. The owner decided to require reason codes and the decision date before any estimate could be closed, and to stop bulk-closing old estimates without an outcome. A year later, sales reviews used the closed records instead of memory.

Reason codes: the field most contractors skip#

Reason codes are the short labels that say why an estimate was won or lost, and they are the field most contractors leave blank. Free-text notes help, but they are written differently by every advisor, so they are hard to compare across years.

Keep the list short enough that advisors use it on a phone in a driveway, and make the field required when an estimate is closed. Watch for an overused other code; when it grows, the list needs a new value. The codes below cover most residential situations.

• 1Price or upfront cost
• 2Financing declined or not approved
• 3Chose a competitor
• 4Timing, deferred to a later season
• 5Repaired instead of replaced
• 6Decision-maker not present
• 7Disagreed with the scope or diagnosis
• 8No response after follow-up

How to recover years of unclosed estimates#

Years of unclosed estimates can often be recovered by matching them to the work that followed. An estimate whose customer later booked an install at the same location usually converted; one with no later job and no contact probably lapsed.

Do the matching on exported copies, not in the live system, and record how each outcome was inferred. A back-filled outcome is useful, but it should be labeled differently from an outcome an advisor entered at the time. Never purge old estimates in a cleanup; deleted unsold estimates cannot be rebuilt from invoices, because no invoice exists for a job you did not win.

• 1Export every estimate with its status, options, dates, customer ID and location ID.
• 2Export jobs and invoices for the same period with the same IDs.
• 3Match each open estimate to any later job at the same location for the same equipment or scope.
• 4Mark matched estimates as converted and unmatched ones as lapsed, with an inferred flag on both.
• 5Leave the original status untouched in the source system and keep the matching rules in a short note.

Why AI teams value option-level outcomes#

AI teams value option-level outcomes because they show real decisions with the context that drove them, which is what pricing assistants, proposal tools and sales coaching models need to learn from. A record that pairs a technician's finding with the options offered and the one chosen is a labeled example a developer cannot easily create on their own.

Your price book is commercially sensitive, so price fields are a decision for you. Options can be described by tier, equipment class and warranty without exact prices, or prices can be grouped into bands you approve. Any license grants defined use of the records; the company keeps ownership, and customer identities are removed during preparation.

How SourceX looks at estimate histories#

SourceX assesses estimate histories with the SourceX Enterprise Data Value Framework. The drivers that matter most here are human-generated signal and domain expertise in the options and reason codes, data cleanliness in the links from estimates to jobs and invoices, scale and recency of the proposal history, and rights under your customer terms; preparation cost and privacy burden reduce net value when identities and price fields must be removed. The first fit check collects metadata only, such as the system name, years of history and which fields are filled.

If a package proceeds, the SourceX five-step transaction of Supply, Rights, Preparation, Approval and Delivery applies, and the SourceX Evidence Packet records which estimate fields were included and how price fields were handled. You approve each step.