Iygge Sampaio
Message on WhatsApp

A full pipeline is not commercial capacity

A full pipeline creates commercial capacity only when the team can actively own the buyer work inside it. If opportunities keep aging without valid evidence, a due next action, or a clear owner, the pipeline is not capacity. It is a queue.

The failure mode is opportunity hoarding. Every plausible deal remains active, each seller carries more open threads, and the CRM reports abundance. In practice, attention fragments. Follow up becomes reactive. Blocked work hides behind stage labels. New demand enters faster than the commercial system can resolve what it already accepted.

The repair is not an arbitrary purge. It is a commercial WIP ceiling: an explicit boundary on how much accepted buyer work the current system can hold open while still giving each item honest ownership.

Definition

Definition: A commercial WIP ceiling is a controlled limit on accepted opportunities between a defined start and finish, supported by explicit entry, exit, defer, and disqualify rules.

WIP means work in progress. The Kanban Guide defines it as work between explicit started and finished points. It also calls for controlling that work, watching item age, and unblocking blocked items. Kanban is not a sales methodology, but its flow logic exposes a useful GTM mistake: the CRM can store unlimited records while the team cannot actively progress unlimited buyer decisions.

Commercial capacity is therefore not the count or nominal value of everything open. It is the ability to move valid buyer work toward a resolved outcome without hidden reconstruction, unmanaged waiting, or repeated founder rescue.

First decide what counts as active work

Most pipeline arguments start too late. Teams debate coverage, conversion, or hiring before agreeing on which records consume active commercial attention.

An opportunity should enter active work only when the team can identify:

  • a buyer signal worth pursuing
  • an accountable owner
  • the evidence that supports the current state
  • a next action the seller or buyer has accepted
  • the condition that will finish, defer, or disqualify the work

A lead in nurture may be commercially relevant without being active work. A proposal waiting on a buyer decision may remain active if the waiting condition, owner, and review point are explicit. A deal with no next action is not automatically active merely because its stage is open.

Salesforce describes a B2B pipeline as both buyer progress and the actions sellers need to take. Its guidance also emphasizes exit criteria. That distinction matters. A stage name describes location. An exit contract tells the team what must become true before the work moves.

Without the contract, sellers can advance records to preserve optimism or leave them open to avoid a loss. Both choices inflate inventory without increasing capacity.

Build the commercial WIP ceiling

The ceiling is not a universal number imported from another company. It is an operating policy learned from the current path.

Start with one path, such as qualified opportunity to resolved decision. Mark every record inside it as one of four states.

Progressing: Evidence supports the current stage, the next action is explicit, and someone owns it.

Waiting by design: The buyer or team has an explicit dependency, a review condition, and an owner who will restart the work.

Blocked: The work cannot move because evidence, authority, routing, or an internal decision is missing.

Unowned: No role is responsible for the next action or the record has no valid reason to remain active.

Now inspect age within each state. Age is not an automatic verdict. A complex decision can take longer than a simple one. The question is whether time is accumulating under an explicit policy or under neglect.

HubSpot documentation separates open deals from stalled deals, records with no scheduled activity, and records with stale close dates. Its product thresholds are not universal benchmarks. The useful idea is that these are different operating states, not one generic bucket called pipeline.

Set the first ceiling only after seeing the current work. It should be low enough that blocked and aging items become visible, but not so low that the team games entry. Treat it as a temporary policy. When the ceiling is reached, the team has four legitimate moves: finish, disqualify, defer with a return condition, or unblock. Adding another active item is not the default fifth move.

Use the stack to locate the restriction

A ceiling makes overload visible. It does not explain the cause. The Truth, Playbook, Architecture, Operator stack does.

Truth: Is the buyer state supported by inspectable evidence, or is the seller reconstructing intent from memory?

Playbook: Are entry, exit, defer, and disqualify decisions explicit enough for two competent sellers to make the same normal choice?

Architecture: Does the system preserve owner, next action, due condition, blocked reason, and age so work can survive a handoff?

Operator: Does someone review aging work, resolve exceptions, and change the policy when repeated cases expose a bad rule?

The Theory of Constraints uses a related sequence: identify the constraint, use it well, align upstream work to it, and only then test added capacity. Applied carefully to GTM, the point is not that sales behaves like a factory. The point is that feeding more opportunities into an unresolved restriction can create more work without creating more progress.

This is also why automation can make the queue worse. Automatic enrichment, routing, reminders, and outreach may increase activity around every open record. If active work has no honest boundary, automation gives opportunity hoarding more speed.

Decision rule

Use this rule before adding demand, automation, or headcount to a full pipeline:

If the oldest active opportunities have valid evidence, explicit next actions, accountable owners, and no recurring process block, test more execution capacity. If any of those conditions repeatedly fail, repair the first failed GTM layer before increasing inflow.

A capacity test should be bounded. Choose one active path, protect its entry policy, add the smallest credible execution change, and observe whether more buyer work reaches a resolved outcome without increasing neglect elsewhere.

Headcount can be the right answer. The ceiling is not an argument against hiring. It prevents the business from asking a new person to absorb ambiguity that the current team cannot explain.

Checklist

Run this audit on one pipeline path this week:

  • Define the exact point where a record becomes active work.
  • Define what finishes, defers, and disqualifies that work.
  • List every active opportunity in the path.
  • Mark each one progressing, waiting by design, blocked, or unowned.
  • Record owner, next action, due condition, blocked reason, and age.
  • Start with the oldest blocked or unowned item.
  • Name the first failed layer: Truth, Playbook, Architecture, or Operator.
  • Finish, disqualify, defer, or unblock before accepting more work at the ceiling.
  • Test added capacity only after the recurring ambiguity is removed.
  • Review whether the ceiling exposes problems or merely encourages stage gaming.

The output is not a cleaner dashboard. It is a smaller, honest set of buyer work that the commercial system can actually own.

What this is not

This is not a command to reduce pipeline coverage. Demand that has not entered active work can remain visible in nurture or another explicit state.

It is not a fixed opportunity quota. Different motions, deal complexity, skills, and buyer dependencies change what a team can responsibly hold open.

It is not a promise that lower WIP will improve revenue. The ceiling is a diagnostic control. It exposes where work waits so the founder can choose the right intervention.

FAQ

Should a sales pipeline have a fixed opportunity limit?

Not one permanent limit for the whole business. Define a temporary ceiling for a specific active path, observe what it reveals, and adjust only with evidence. The purpose is controlled learning, not numerical theater.

Does deferred work disappear from the forecast?

Not necessarily. Deferred work should remain visible with a reason, return condition, owner, and appropriate forecast treatment. It stops consuming active execution capacity until the agreed condition occurs.

When does more headcount become the right answer?

When accepted work has reliable evidence, stable decisions, working handoffs, explicit ownership, and still exceeds available execution. At that point, a bounded capacity test is more rational than another process repair.

If your pipeline looks abundant but the team cannot name what it can actively own, a Lorde GTM diagnosis can trace the restriction across Truth, Playbook, Architecture, and Operator ownership.

Lorde

Message on WhatsApp