The way a first remote hire fails is almost always the same, and it is rarely dramatic. Nobody is dismissed

in week two. The arrangement thins out instead: the queue of work gets shorter, the weekly check-in is

moved and then cancelled, and after a couple of months somebody says the person was not proactive

enough and stops assigning. Because the ending is so undramatic, the diagnosis goes unexamined, and

the business concludes that remote hiring does not work for them. Occasionally that is true. Far more

often something specific went wrong, and it was visible early.

The most common cause is that the hire arrived into a week nobody had defined. Everyone agreed there

was too much work, which is not the same as agreeing what the first month should produce. In the

absence of a defined week, the assistant does what any competent person does in an ambiguous role:

they wait to be asked. Meanwhile the person who hired them is waiting to be impressed. Both are being reasonable, and the placement is already drifting, because nothing exists that either of them can point to

as progress by Friday.

Consider a six-person marketing agency bringing in its first assistant to take reporting off the account

team. In week one they are given last month's client pack as an example, a login to the dashboard, and the sentence "just get familiar with things". Nobody says which reports are due when, who checks the

first draft, or what the account team will stop doing once the handover happens. By week three the

assistant has produced two packs, both returned with heavy edits and no explanation of the pattern

behind them. By week six the account team has quietly resumed doing the reports themselves, because

editing felt slower than writing. The assistant is not failing; the arrangement is. But it is the assistant

who will be described as not working out.

The second cause is the missing named person. Questions in a new role arrive daily and are mostly

small: which template, which client, does this number look right. Where there is one named person who

answers the same day, the questions taper off within a month as the answers become knowledge.

Where the answer is "just ask anyone", every question becomes a social calculation about whom to

interrupt, and the safest option is always to guess silently. Silent guessing is what gets labelled, months

later, as carelessness.

The third sits deeper: the work handed over was never actually delegable in the state it was in. It lived

in one person's head, full of unwritten exceptions, and the handover consisted of a screen recording and optimism. The new hire executes the steps, reality departs from the steps, and the work comes back — with interest, because now there is a handover cost too. This failure gets blamed on the hire's initiative, but

no amount of initiative can recover judgements that were never written down.

What predicts success is the mirror image, and none of it is glamorous. A first month defined in writing

before day one. One named person for questions. Work that repeats often enough to be worth learning.

A weekly check-in that survives busy weeks, because placements drift in exactly the weeks the

check-in is skipped. And a client-side willingness to explain the pattern behind an edit once, rather

than making the same correction silently forever.

None of this requires an exceptional candidate. It requires an arrangement that a competent person

can succeed inside. The uncomfortable, useful conclusion is that most first placements are

decided before the assistant ever logs on — and that the businesses which get this right tend to get

it right on every hire that follows.