Partial Allocation and Why It Fails
When half a claim produces half the value and when it produces none, the threshold test that distinguishes them, and the cost of fragmenting people across claims.
Giving every claim a share of a short capacity feels fair and is frequently the worst available outcome. Some work scales with the resource given to it; some does not, and giving the second kind half of what it needs produces nothing at all while consuming the capacity.
The allocation decision in “Partial Allocation and Why It Fails” should be made from visible constraints rather than an assumed headcount. Used carefully for does microsoft teams track your activity, does microsoft teams track your activity for planning teams can contribute recent workload context without replacing the documented priority rule or the manager accountable for the trade-off.
The distinction is knowable in advance, by asking, and almost no book of claims records it.
For an outside reference relevant to “Partial Allocation and Why It Fails”, consult Kellogg operations research. Compare the general guidance with the organisation’s own unit, period and decision rights.
Quantity claims and threshold claims
A quantity claim scales: half the people, half the output, twice the elapsed time. Maintenance, support, throughput work and most ongoing service are like this.
A threshold claim does not: below a certain allocation nothing useful emerges. Anything with a fixed scope and a fixed date, anything requiring a critical mass of people to make progress, anything where partial delivery is worthless.
The question that sorts them
"If we could only give you half of this, what would you get?" The answers divide cleanly between "half as much, or twice as slow" and "nothing useful".
One question per claim, recorded as one word in the book. It is the single most valuable field after the state.
What partial allocation does to a threshold claim
Consumes the capacity, produces no result, and is usually discovered at the end of the period when the deliverable does not exist.
It also damages the claimant's trust in the process, because they accepted half expecting a reduced outcome and received none, which was predictable at the point of the decision.
The fragmentation cost
Partial allocation usually means splitting people across claims, and a person split across three claims does not deliver a third of each. Context switching, three sets of meetings and three sets of relationships consume a visible share of the time.
Where the account is kept in hours, that cost can be estimated and should be: a split person is worth noticeably less than the sum of their fractions.
The rule of thumb worth having
Fewer, fuller allocations beat more, thinner ones. Where capacity is short, finishing three claims completely beats making progress on six.
That is both arithmetically true in most cases and strongly resisted, because starting six looks like serving everybody and finishing three looks like refusing three people.
When partial allocation is right
Quantity claims with no deadline: support, maintenance, improvement work, anything where a slower rate is acceptable.
Also pilots and investigations, where a small allocation is genuinely informative. Those claims should be recorded as deliberately partial with a stated expectation, rather than as a full claim that got less.
The sequencing alternative
Instead of giving everybody half for a quarter, give one claim everything for six weeks and the other everything for the following six.
Elapsed time is the same, both finish, and neither suffers the fragmentation cost. Sequencing is usually available and rarely proposed, because the conversation is harder in week one and much easier in week twelve.
Explaining it to claimants
The sequencing conversation goes better with the arithmetic in front of everybody: "split, both of you finish in fourteen weeks; sequenced, one finishes in six and the other in twelve."
Most claimants choose sequencing when they can see the numbers, and almost none propose it unprompted.
Recording the decision
Which claims were partially allocated, which were sequenced, and what each was expected to produce at the reduced level.
That record is what makes the variance analysis meaningful: a claim that underdelivered on a partial allocation did not fail, and without the record it will look as though it did.
The three ways to give less
- Less capacity, same scope, longer elapsed time — works for quantity claims.
- Same capacity, smaller scope, same date — works where scope is divisible.
- Full capacity, later start — works for almost everything and is least used.
Naming which of the three is being offered, rather than simply allocating a smaller number, is what lets the claimant tell you it will not work before the quarter rather than after it.
What fragmentation costs in hours
Estimates vary and the direction does not: a person on three simultaneous claims loses a visible share of their week to switching, and the loss is larger when the claims are unalike.
Whatever figure an organisation adopts, it should be a line in the account rather than an unstated discount, because otherwise the plan assumes three half-people equal one and a half, which they do not.
The failure to avoid
A threshold claim given a partial allocation, with no conversation, recorded in the book at its full figure. It consumes the capacity, produces nothing, and shows in the variance as a delivery failure belonging to whoever was working on it.
That sequence is common and every step of it is avoidable with one field in the book.