steadystate · Tools

Next Drop Check

Five dimensions, twenty statements, scored zero to three. The output is a profile of your next drop, not a grade. Nothing is stored; your answers live in the URL.

The next drop is where product and delivery meet. Most of what goes wrong with it was decided before anyone wrote a line of it.

Score the drop you are about to commit a team to: a season, a DLC, a feature, an update. Each statement is scored against four anchors describing what that score looks like in practice, because without them everyone gives themselves a two. Answer from what happened on the last drop rather than from what is supposed to happen. The result names the two weakest dimensions and what usually sits underneath them.

How to score this honestly

Score the drop as the team behaved on the last one under pressure, not as the plan says this one will go. If a practice exists but was dropped the last time a date slipped, it scores a one. The anchors are written so that the flattering answer and the true answer are different sentences. If you find yourself reading past an anchor to reach a higher number, that is the finding.

01 · The market reason

Who the drop is for, which titles it is measured against, and whether anyone outside the team agrees it will sell.

The drop has a named audience inside the current player base, with a size.

What it costs: A drop aimed at everyone is aimed at the players who already spend, and the most expensive content a live team makes is the kind that moves nobody new.

The comparable set includes titles that shipped the same kind of content and did not move.

What it costs: A comparable set made of hits teaches survivorship. The misses are where the two or three weeks of rework hide, and they are found before or after the money is spent.

A number the drop must move, and by when, was written down before the first build.

What it costs: Without a target set in advance, the result is read as a success by the people who built it and the next drop inherits the same reasoning.

Someone outside the team read the market reason before the drop was approved.

What it costs: The cheapest correction a drop ever gets is the one made before the first sprint, and only an outsider is placed to make it.

02 · The engagement evidence

Whether the assumption about how players will respond comes from the title's own telemetry or from what the model needed.

The engagement assumption in the business case comes from the title's own telemetry, with the query named.

What it costs: An engagement assumption chosen to make the case close is the most common reason a drop that shipped on time still loses money.

The last drop's actual engagement was compared with its assumption, and the difference was written down.

What it costs: A team that never measures its own forecast error repeats it every quarter, and the error is usually in one direction.

The people who approve drops look at the telemetry that would tell them a drop failed.

What it costs: A failed drop that nobody with authority sees for a month is a second drop built on the same assumption.

The drop's first-session and retention funnel is instrumented before it ships, not after.

What it costs: The most valuable cohort a drop will ever have is its first week, and it is unmeasured for every drop instrumented afterward.

03 · The money

Fully loaded cost, the return against the next best use of the same people, and who decided.

The cost of the drop includes the team's time at fully loaded rates, not only external spend.

What it costs: A drop costed without its people is the reason a quarter of content can show a positive return while the studio loses money.

The return is compared with the next best use of the same people, not with doing nothing.

What it costs: Almost everything beats doing nothing. The money lost is the drop that the same four people would otherwise have shipped.

There is a return threshold, and the last time it was waived somebody signed the exception.

What it costs: A threshold waived twice without a name is not a threshold, and the third waiver is where the quarter's margin goes.

The drop has kill criteria that were agreed while everyone was still optimistic.

What it costs: It is impossible to agree kill criteria fairly once the drop is struggling and people are defending their own work, so the money is spent to the end.

04 · The date

What the date depends on, in weeks, and who outside the planning room has read the list.

What the date depends on is written down, each item with a number of weeks against it.

What it costs: A dependency discovered inside two weeks costs the marketing spend already committed against the date, the largest avoidable write-off a drop produces.

Someone who was not in the planning room has read the dependency list.

What it costs: The dependencies a team cannot see are the ones it takes for granted, and only an outsider asks about those.

A cut line exists, and cutting to it does not need an escalation.

What it costs: A drop with no cut line ships late or ships broken. Both cost more than the feature that would have been cut.

The last three drops shipped on the date first announced internally, or moved once with a stated reason more than a month out.

What it costs: A team whose dates do not hold is managed by the publisher in detail, which costs the publisher's time and the team's authority at once.

05 · The risk

Whether every risk has an owner and a date, what is at stake if the drop slips, and whether the status report could go to the publisher's board unchanged.

Every risk on the drop has an owner and a date by which it needs a decision.

What it costs: A risk without an owner is carried by the schedule, and the schedule pays for it in the last two weeks at the highest rate.

The revenue at stake if the drop slips is stated, in the title's own numbers.

What it costs: Until a slip is a number, it competes for attention with everything else and loses to whatever is loudest that week.

Status is reported as red or green, with the reason for every red, and no amber.

What it costs: Amber is where a slip spends the weeks in which it could still have been cheap to fix.

The current status report could be put in front of the publisher's board without being rewritten.

What it costs: Every rewrite between the team and the funder is a place where risk is rounded down, and the rounding is paid for when the funder finds out.