wiki › Validation

Engine stall

An engine stall is the case where neither side gets anything: users aren't getting closer to what they came for, and the provider isn't earning revenue from them. Users may still be showing up and spending effort, but that effort isn't turning into progress for them or into payment for the business, which leaves nothing on the user's side for a pricing or packaging change to collect against.

Definition

Engine stall is the case in the double-value problem's four-case matrix where neither value stream is produced: the user makes no progress toward their sought state, and the provider collects no revenue.

Neither stream produced

The matrix sorts a population by which of the two streams — user value and business value — actually gets produced:

The engine is the arrangement that turns user energy into those two streams. The user supplies energy, attention, and intent in real time; the provider supplies pre-recorded work that makes that energy go further toward the state the user is pursuing. Stall is the condition where no energy is moving through that arrangement. The user isn't traversing, so there's no route efficacy to improve, and no margin-on-efficacy for the provider to collect against.

The cell describes a population, not an offering. Any offering of reasonable size has users sitting in all four cells simultaneously; the question is what share sits where.

Two ways an engine stalls

  1. Direct stall. The user arrives, the offering never becomes route-efficacious for what they came to do, and they leave. Neither stream was ever produced. This is the ordinary case, and at activation it accounts for most of the loss.
  2. Collapsed parasitism. A parasitic offering produces business value by drawing on user energy without improving route efficacy. That draw isn't replenished by value receipt, so the user-side supply depletes and the business-value stream stops with it. Parasitism is the transitional condition; stall is where it lands at scale.

Stall and coupling failure in one cohort

Both cells contain non-paying users, which is why they get read together as "users who didn't convert." Separating them changes what the intervention should be. In coupling failure, the user is producing value the provider hasn't tied to a commercial outcome — the fix operates on where and when payment is asked for. In stall, the user never got anywhere — the fix operates on the route itself.

Value milestone instrumentation is what separates the two in the data. A non-converting user who cleared milestones is coupling failure; a non-converting user who cleared none is stall.

Example

A signup cohort for an invoicing tool. One user signs up, gets stuck connecting a bank account, never sends an invoice, and lets the trial expire. No invoice went out, so nothing improved in their billing; no payment came in, so nothing accrued to the provider. That user is stalled, and they exit inside starting churn, before enough has happened to establish whether the offering would have served them.

Another user in the same cohort sends invoices weekly for two years on the free tier and never upgrades. Same absence of first paid conversion, different cell — that user is producing user value that hasn't been coupled to revenue.

Both land in missed conversions when the cohort is counted at the commercial outcome alone, which is what makes stall easy to overcount and hard to act on without milestone-level readings.

See also