Salesforce opportunity stages: the defaults, and a set that forecasts better

By Giorgio Bonan. Updated October 22, 2026, published October 22, 2026.

Salesforce opportunity stages are the values of the Stage picklist on an opportunity. Each stage carries a default probability, a forecast category and a flag for closed and won. Salesforce ships ten of them, from Prospecting at 10% to Closed Won at 100%. They work as a starting point, but stages named after the seller's activity rather than the buyer's commitment are the main reason a pipeline looks bigger than it is.

The ten default stages

We read the stage settings of a Salesforce Developer Edition org through the API on 22 October 2026 (the OpportunityStage object). This is what an org gets before anyone edits the picklist:

Two details stand out. Eight open stages is a lot: in practice reps skip some and park deals in others. And in that org every open stage maps to the Pipeline forecast category, so nothing reaches Best Case or Commit unless a rep sets the category by hand or an admin remaps the stages.

How stages feed the forecast

Salesforce's Pipeline Forecasting Implementation Guide (Winter '27) states it plainly: forecast categories map to opportunity stage values, and the stage determines how an opportunity contributes to a forecast. The mapping lives in Object Manager, Opportunity, Fields, Stage, in the Opportunity Stages Picklist Values list.

Probability plays a different role. The forecast grid sums Amount by category; probability feeds the Expected Revenue field and any weighted pipeline report. So a stage can be wrong in two ways at once: mapped to the wrong category, which distorts the forecast, and given the wrong probability, which distorts weighted pipeline. Our guide to Salesforce forecasting shows both effects on one worked month.

A six-stage set built on exit criteria

A stage should describe what the buyer has done, not what the seller did. "Proposal sent" is a seller action; "buyer confirmed budget and timeline" is a buyer commitment you can check. Here is a set we recommend as a starting point, with the criterion a deal must meet to leave each stage:

The probabilities above are placeholders. Replace them with your own: after two quarters, compute for each stage the share of deals that reached that stage and were eventually won. The win rate calculator does the arithmetic, and the stage durations feed the sales velocity calculator.

If several teams sell differently, keep one Stage picklist and create a Sales Process per record type, each selecting its own stages. One picklist means one mapping to forecast categories, so every team's forecast adds up the same way.

Find deals stuck in a stage

The Opportunity object has a LastStageChangeDate field: the date the stage last changed. It is empty for a deal whose stage never changed since creation. This query lists open deals that have not moved stage in 45 days, by owner, oldest first:

SELECT Owner.Name, Name, StageName, Amount, CloseDate, LastStageChangeDate, CreatedDate
FROM Opportunity
WHERE IsClosed = false
  AND (LastStageChangeDate < LAST_N_DAYS:45
       OR (LastStageChangeDate = null AND CreatedDate < LAST_N_DAYS:45))
ORDER BY Owner.Name, LastStageChangeDate ASC NULLS FIRST

A deal stuck in a stage is not always dead; a long legal review can sit in Commit for weeks. What separates the two is activity. A stuck deal with meetings and emails logged this week is slow; a stuck deal with no logged activity for a month is a candidate for Closed Lost. Pair this query with Last Activity Date, and with Pipeline Inspection for what changed.

Five rules that keep stages honest

1. One stage, one verifiable criterion. If two reps would put the same deal in different stages, the criterion is too vague.
2. Map late stages to forecast categories on purpose. Do not rely on reps overriding the category deal by deal.
3. Use Path to show the criteria on the record. Salesforce's Path component displays guidance per stage at the top of the opportunity, where reps update it.
4. Recompute probabilities from history every two quarters. A probability nobody measured is a guess presented as a number.
5. Close what is dead. Deals that stay open forever inflate every report; our CRM hygiene checklist gives a cadence.

TruthPipe applies the last rule every night: it reads the stage, the close date and the last activity of every open deal, lists the ones that stopped moving, and scores each rep's pipeline, for $99 per month per org.