Published cartoon ·
What is the ETA for an ETA?
The date field must be filled, even if the answer is another date.
Why this one bites
Calendars and clocks loop around an employee asked to predict when a prediction will be available. The demand for certainty has moved one level outward, without adding any of the information required to estimate the work.
There can be a sensible deadline for investigation. Here, however, the caption says the content of the answer hardly matters as long as it looks like a date. The reporting requirement has detached itself from the planning question it was meant to answer.
Accessible summary
- A manager points at a calendar while a seated employee writes. Arrows connect progressively smaller calendars and clocks in a loop.
- The caption says the speaker has been tasked with obtaining an ETA and will accept an answer provided it is in date format.
Reading notes
Sometimes a team can estimate the investigation before it can estimate the implementation. That is a legitimate distinction. It may need to inspect a dependency, reproduce an error or learn whether an assumption is true. The absurdity arrives when a provisional date for that learning is treated as a way to manufacture certainty about the final result. An estimate of when knowledge may improve is still not the knowledge itself.
This produces a nested commitment problem. The employee must first defend how long it will take to determine how long the work will take. If the investigation finds something unexpected, both dates can be portrayed as failures. The incentive is to supply a reassuring answer quickly, even though the question exists because a responsible answer is not yet available.
A more useful conversation asks what is unknown and which step will reduce that uncertainty. The outcome of the step may be an estimate, a range, or a finding that the original approach is unsuitable. All three are progress if they prevent a false promise. The joke in asking for the ETA of an ETA is that the organization has begun managing the appearance of predictability before it has allowed anyone to investigate what would make a prediction credible.