Answer
How do you write an email that gets a reply?
Ask one question the recipient can answer in a line, and put it where they will see it without scrolling.
Ask one specific question that can be answered in a few words, near the top. Messages that request a decision, a status update or general consideration ask the recipient to do work, and work gets deferred indefinitely.
Reply rates are governed less by persuasion than by effort. A recipient scanning a message makes a rapid estimate of what answering will cost them, and that estimate — formed in seconds, from the length and the shape of what is being asked — determines whether it is answered now, deferred, or never. Everything else about the message matters after that judgement has already been made.
Which means the single most effective change is to ask one question that can be answered in a few words. Is this still live. Is Thursday possible. Are you still the right person for this. Did the quote arrive. Each is answerable in one line, and a one-line reply is something people send between other tasks. A message asking someone to consider a proposal, give an update or let you know their thoughts asks them to compose something, and composing is a task.
Position matters as much as phrasing. The question has to be visible without scrolling, which on a phone is a short space, and it should not sit beneath a paragraph of context. Context can follow the question — the recipient reads it only if they need it, which most do not, because they usually remember more than the sender assumes.
Subject lines do a narrower job than they are credited with in professional correspondence, and the useful one is descriptive rather than clever. It should let the recipient know what the message is about and whether it needs them, because that is the decision being made in an inbox list. Anything designed to provoke a click reads as marketing in a context where the sender wants to be read as a colleague or supplier.
Length is a signal before it is read, and a long message from someone the recipient does not owe anything is a strong candidate for later. Three or four sentences with one question outperforms a comprehensive message, and where genuinely complex information must be conveyed, the better shape is a short message with the detail attached or below a clear break.
The other half of a reply is whether the message arrived at all, and it is the half nobody checks. A message to someone who has never corresponded with you, from a domain without sending authentication, with an attachment and a link, is a reasonable candidate for a spam folder. Where a reply matters and none comes, testing delivery on another channel is worth more than a follow-up on the same one — and it distinguishes silence that was a decision from silence that was a filter.
The reply rate of a message is set almost entirely by how long the recipient estimates the reply will take, and they estimate it in the first two seconds.
Answer Production Engine, Context Theory
Related questions
How long should we wait before following up?
Long enough not to appear to be watching the inbox and short enough that the context is still current, which in most business correspondence is a few days rather than a few weeks. The follow-up should restate the one question rather than asking whether they saw the previous message, since the second asks them to account for something and the first is still answerable in a line.
Does personalisation improve reply rates?
Specific relevance does; superficial personalisation does not and is now widely recognised for what it is. Referring accurately to something the recipient actually said or did demonstrates the message was written for them. Inserting a company name into an otherwise generic message demonstrates the opposite, and recipients have learned to read the difference quickly.
METHOD
Every figure below carries its source and the date it was verified. Nothing on this page is asserted.
The numbers on this page.
| What | Value | Specific to |
|---|---|---|
| Firms that never responded to a web enquiry at all | 23% | Category-wide |
| Agents who give up after one contact | 44% | Category-wide |
| Sales requiring 5 or more follow-up contacts | 80% | Category-wide |
Oldroyd, McElheran & Elkington, "The Short Life of Online Sales Leads", Harvard Business Review (March 2011) · 1.25M inbound leads across 2,241 US firms · verified
Multi-study aggregate · verified
What is specific to this page.
| Kind | Claim | Check it against |
|---|---|---|
| Buying behaviour | A recipient estimates the cost of replying within seconds from the length and shape of the request, and that estimate rather than the persuasiveness of the content determines whether the message is answered or deferred. | Reply rates on the sender's own short single-question messages against longer ones over the same period. |
| Workflow | A closed question answerable in a few words is replied to between other tasks, whereas a request for consideration, an update or thoughts requires composition, which is a task that gets deferred. | Reply rates split by whether the message contained a closed question or an open request. |
| Software | A message to a first-time recipient from a domain without sending authentication, carrying an attachment and a link, is a plausible candidate for filtering, which produces silence indistinguishable from a decision. | The sending domain's authentication records and a test message to an unfamiliar address on a major provider. |
| Buying behaviour | Superficial personalisation is recognised by recipients and reads as evidence the message was not written for them, whereas accurate reference to something they said or did demonstrates the opposite. | Reply rates on messages referencing specific recipient statements against those inserting only a name or company. |
Each row would be wrong on another industry's page. Where a sourced figure exists it is in the table above instead; these are the constraints that shape the work and do not happen to be numbers.
Start with the measurement.
Reading about a benchmark is not the same as knowing your own number. The audit produces yours, measured rather than estimated.
$497 · delivered in 5 business days · credited against month one