Speed wins early jobs, and it can be scheduled

Cover the few hours when enquiries cluster, reply within minutes inside them, and switch off outside them. Speed does matter: in a Management Science study of more than 11 million Fiverr exchanges, Einav Hart and colleagues found a one-hour delay nearly halved a freelancer's chance of being hired.

Response time is the one competitive advantage available to someone with no record, because it is the only thing a buyer can observe before they have committed anything. It is also the advantage most likely to eat your week, and it does not have to, because demand is not spread evenly across the day.

Demand clusters, and the clusters are findable

Enquiries in this trade arrive disproportionately in evenings and at weekends, in the buyer's local time, not yours. That is the general shape. The specific shape - which evenings, how late, whether your Sunday is dead or your best day - is different for every earner and depends entirely on where your buyers are.

You do not have to guess it. Log the arrival time of every enquiry for three weeks, in one column of a spreadsheet. Thirty data points is enough to see a pattern that will not change much afterwards, and it is the cheapest research available in this trade.

If most of your buyers turn out to sit in one timezone that is not yours, that is a scheduling fact rather than a problem, and it is worth knowing before you build a routine around your own evenings. The wider question of where your buyers are and what that does to your working day is covered in the piece on geography and who your buyers are.

What "fast" actually has to mean

Fast does not mean instant, and the returns fall off a cliff rather than declining smoothly.

The relevant comparison is not against a clock, it is against the other three people the buyer messaged at the same time. Somebody shopping five profiles at once will usually book whoever gives them a competent answer first, and after roughly an hour the decision has often already been made by someone else. The same research team reported that a reply after a full day did about as well as no reply at all.

Which produces a simple rule. Inside your declared window, reply in minutes. Outside it, reply when the window next opens, and say in your listing when that is.

A stated window that you hit is worth more than an unstated availability that you mostly hit, because the first is a promise kept and the second is a lottery the buyer did not know they were entering.

Saved replies that do not read as templates

You will answer the same four or five enquiries forever, and typing them fresh each time is the reason people end up chained to a phone.

The four that cover most of it:

• A thin brief. Acknowledge, state what you would produce, ask exactly one clarifying question. One, not three: an interrogation loses the booking.
• A request outside what you do. One sentence, no apology paragraph, and a note of what you do instead.
• A price question. The number, what it includes, the turnaround. Never a range, never "depends".
• A yes. Confirm scope and date in two sentences, then start.

The thing that makes a saved reply read as a template is that it does not mention anything the buyer wrote. So build each one with a deliberate gap in the first sentence, and fill it with their own noun, their own timeline, their own detail. Ten seconds of specificity in sentence one buys you three sentences of boilerplate afterwards without anybody noticing.

Keep them short. A long reply to a short enquiry reads as eagerness, and eagerness reads as no work on.

Windows you can actually keep

Pick fewer hours than you think and defend them.

A realistic starting structure for someone part-time is two windows a day - one at lunch, one in the evening - of about forty minutes each, plus a longer weekend block if your log says weekends are where your buyers are. That is under two hours a day of being reachable, and it covers the overwhelming majority of enquiries within an hour of arrival.

Then turn the notifications off outside the windows, properly, not aspirationally. The cost of being permanently reachable is not the minutes, it is that no other block of your day is ever fully yours, and the longer account of what always-on availability actually costs is the case against solving this problem with willpower.

Two practical notes. Set an away message with a real return time rather than an apology, because a buyer who knows they will hear at 7pm usually waits. And do not let the windows creep - a window that expands to cover every enquiry is just being online with extra steps.

The part that is not about speed

Speed only converts if the reply is any good, and a fast bad reply loses the job faster.

What buyers are reading for in a first reply is risk: whether you understood the brief, whether you will disappear, whether the price is the price. The buyer-side account of how a commission conversation is expected to open is short and corrects most of what new earners assume about what a good first message contains.

It is also worth knowing what the buyer did before messaging you. Most have already run something automated and are messaging a person because that output did not answer their actual question - what an automated assessment produces tells you what they have already seen, and your reply is more persuasive if it does not repeat it. And where an enquiry is really asking about numbers and how they were arrived at, the data side of that question belongs to someone else and one sentence pointing there is a better answer than a paragraph of your own.

Log the enquiries that did not convert alongside the ones that did, with the reply time attached. After a month that column will tell you whether your window is in the wrong place, which is a different problem from being too slow, and the fix for it is a calendar change rather than an effort increase. For the platform-side mechanics of enquiry handling and notification settings, Rate Cock's judges page is the current source.