The Worked-Prospect Standard: When to Retire an Unresponsive Contact

This standard answers one question with a repeatable verdict: has a single unresponsive prospect been worked enough to retire from active outreach, or would stopping now be premature? It is written for founders selling their own product, account executives, SDR leads, and partnerships teams who need a manager and a rep to grade the same activity record identically. It gives you the criteria - valid touches, distinct channels, spacing, deliverability, and elapsed time - so a person is retired on evidence, not on the rep's patience running out.

Public advice answers "how many touches" with contradictory single numbers and never says when a touch actually counts. This document replaces that with a definition of done you can adopt as team policy. Read it, apply the checklist, and premature abandonment becomes auditable.

Why "how many touches" is the wrong question

The touch-count debate is unresolvable without a definition of a valid touch. Sources range from 5 to 12, but they count differently, so the number alone tells you nothing about whether a prospect has been worked.

Here is the spread from independent research, all pulled into one place so you can see how little the raw figures agree.

RAIN Group's buyer research puts the average at 8 touchpoints to an initial meeting, with top performers needing only 5. Highspot cites 8 to 12 before an early-stage prospect responds. These are not the same measurement dressed differently; one counts touchpoints across calls, email, and social, another counts attempts on a single channel. A "12-touch" record where the emails bounced is not worked at all. A "6-touch" record spanning three channels may be more worked than a "10-touch" email-only record.

So the leverage is not in picking a number. It is in defining what makes a touch count, and then counting only those. That is what the rest of this standard does.

A bounced email is not a touch, and twelve of them is not persistence.

The persistence data does argue against stopping too early. Eighty percent of successful sales require five or more follow-ups, yet 44 percent of reps give up after one, and only 8 percent make it to the fifth attempt. That is a real bias to correct. But "never give up" is equally wrong: follow-ups beyond the reply-capture window yield marginal or negative returns. Both premature retirement and over-persistence are measurable errors, and this standard is built to catch both.

What makes a touch count

A touch counts only if it was delivered, reached the right person, used a distinct channel, and was spaced from its neighbours. Miss any one of those and the touch is noise in your count, not evidence of work.

Break the definition into four conditions:

• Delivered. The message reached an inbox or a phone, not a bounce log. Emails sent does not guarantee emails delivered.
• Right person. The contact still held the targeted role when the touch went out. If they left, a non-response is a dead address, not disinterest.
• Distinct channel. The touch used one of the baseline channels - email, phone, or social - and adds channel coverage rather than repeating one.
• Spaced. The touch sat at least 48 to 72 hours from the prior touch on that contact, so it read as presence, not pressure.

Each condition proves something specific, and each has a tell when it lies:

The delivery condition is the one people skip, and it is the most consequential. A hard bounce means the address is invalid: the domain is valid but the mailbox is gone because the person left the company, the alias was deleted, or the address was never real. Because B2B email addresses go invalid at 2 to 3 percent per month, a 90-day-old unverified list carries an estimated 6 to 9 percent invalid rate. That decay is silent. Nobody tells you the address died; it just stops responding, and a rep counts the silence as disinterest.

The distinct-channel minimum

A fully worked cadence spans at least three distinct channels - email, phone, and social. A high touch count on one channel is not a worked cadence; it is single-channel padding that inflates the count while suppressing the response probability.

The reason is in the reply rates. Single-channel effort underperforms coordinated outreach by a wide margin, and the gap is large enough that it changes the verdict on a record.

Cold email replies around 4 to 5 percent. A LinkedIn message replies at roughly 10.3 percent. A well-built multi-channel sequence lands 12 to 18 percent, and multi-channel reply rates run consistently 2x to 3x single-channel email. That lift is why a 10-email record can be less worked than a 6-touch record that used email, phone, and social. The email-only record never gave the prospect the channel that would have converted them.

For scale, the Bridge Group's SDR report describes a median SDR running about 112 activities a day, split roughly 44 calls, 41 emails, 19 LinkedIn touches, and 8 text or other. A worked record should look like that split in miniature. If a record is all email and no calls or social, it has not been worked across the channels the response data rewards, no matter how high the email count climbs.

Spacing, elapsed time, and the break-up

Touches must be spread across the cadence window, not clustered, and the sequence must run its full documented span before you infer disinterest. Contacting a prospect twice on Monday and then going silent for two weeks loses momentum and does not count as two spaced touches.

Two guardrails govern this:

• Spacing. No more than one automated touch per 48 to 72 hours per contact. Clustered touches read as a machine and get muted.
• Elapsed window. A typical cold cadence runs 17 to 21 days, with a working range up to 30. Retiring before that window closes is premature.

The timing of replies is the strongest reason to hold the window and then let go. One analysis found a Day 0/3/10/17 cadence captures 93 percent of total replies by Day 10, with diminishing returns after. A break-up email typically lands around day 20 to 24 after 7 to 8 touches.

The break-up touch is not optional, and skipping it is a common way to lose a reply. A polite break-up often triggers a response because the pressure is off; the prospect knows you are stopping, so the cost of a quick reply drops. Send a final message that says you are giving up while offering one last chance to change their mind. Log it. A record retired without a logged break-up is not fully worked.

The retirement procedure

Grade one record against every condition, in order, then retire it or send it back to active. The procedure takes about 30 minutes per record the first few times and far less once it is habit. Do not skip a step to save time; the skipped step is where the wrong verdict hides.

Sources disagree on where the channel-escalation step sits and on whether you cap at 4 to 6 touches then pause 30 days, or run 7 to 8 touches then break up. Both are defensible. Pick one as policy and write it down, because the point of a standard is that everyone applies the same one. What is not negotiable is that delivery, channel, spacing, and elapsed time all pass before you retire.

Your sequencing tool should encode this. Salesloft's cadence settings, for example, route added, bounced, finished, and replied states to distinct stages - in the documented example, added goes to "Open," bounced to "Working," finished to "Completed," and replied to "Do Not Contact." The lesson is that "finished with no reply" and "still active" are different stages by design. Suppression rules should stop the sequence the moment a prospect replies, books, purchases, or unsubscribes.

Set the re-entry trigger before you retire

Retire records, never delete them, and attach a re-entry trigger so a retired prospect can legitimately return. A record with no trigger cannot come back on a job change, which is the single most valuable reconnect signal you have.

A job change resets buyer context, not just timing. The person arrives at a new company with a new mandate and often a budget to spend. The window matters: act within 48 hours of a verified move, because that separates a warm reconnect from a cold email. One published re-engagement program reportedly generated over $1.2 million in new pipeline and 8x ROI in its first year by treating movers as warm reconnects. That return is only available if the closed-lost record still exists with a trigger attached.

Set two kinds of re-entry condition on every retired record:

• Trigger-event watch. A job change, a hiring signal, or a change in the account. This is the high-value path.
• Elapsed-time timer. A 30-day-plus pause before restarting with a fresh angle, per the cap-at-4-to-6-then-pause pattern. This is the fallback when no event fires.

Finding movers by hand is the friction here: you would have to re-check every retired contact's employer against reality on a schedule. This is exactly the surface Refolk is built for. Instead of maintaining a stale list, you ask for the people who just changed roles and get them back.

The audience is deep enough that disciplined retirement pays off rather than starving your pipeline. In Refolk's index there are 21,430 current US SDR profiles and 240,748 current US Account Executive profiles - an 11.2x AE-to-SDR ratio - so the population your policy governs is large, and freeing rep time from truly exhausted records compounds instead of leaving reps with nothing to work.

The count column comes from Refolk's index; the ratio column is derived from those counts.

How this goes wrong

The retirement verdict fails in predictable ways, and every one of them is a specific check you can run. This is the most valuable part of the standard: a record can look worked and not be, or look abandoned and actually be dead. Grade against each failure mode before you sign off.

• Counting sends, not deliveries. A "12-touch" record where emails bounced is not worked. The false positive is a high touch count that never reached anyone. Check: reconcile the sent count against the bounce log before counting.
• Single-channel padding. Twelve emails is not a worked cadence when the baseline is email, phone, and social. The false positive is a high count on one channel. Check: count distinct channels, not touches.
• Clustered touches. Touches jammed into two days then silence. Contacting twice on Monday then waiting two weeks loses momentum. Check: spacing between touches meets the 48 to 72 hour guardrail.
• Wrong-person or stale record. The contact left the role, so the "no response" is a dead address. The mailbox is gone because the person left the company. Check: verify title and employer at the time of the last touch.
• Retiring during a deliverability failure. If bounce rate spiked, non-responses may be spam-foldered, not disinterested. A 6 percent bounce rate can drag open rates from 25 percent to under 10 percent in weeks. Check: was domain hard-bounce under 2 percent during the sequence?
• Break-up skipped. Retiring without the final touch loses the reply the break-up often triggers. Check: the break-up is logged before the stage change.
• Deleting instead of retiring. A record with no re-entry trigger cannot return on a job change. Check: retired records carry a trigger watch or timer.
• "Never give up" bias. Treating persistence as unlimited ignores that follow-ups beyond the reply-capture window yield marginal or negative returns. Check: the elapsed-time threshold is enforced, not open-ended.

The deliverability line deserves a hard number, because it is the one place the standard connects to your sending reputation. An acceptable cold-email bounce rate is under 2 percent for hard bounces and 3 to 5 percent for total bounces. For bulk senders, Gmail and Yahoo require a spam complaint rate below 0.1 percent and never above 0.3 percent. If a sequence ran while the sending domain breached those lines, its non-responses are contaminated, and you cannot cleanly retire on them.

The retirement checklist

Run this before moving any record out of active outreach. If every box is checked, retire the record on evidence. If any box fails, the record goes back to active, and you note which condition it failed so the fix is specific.

Keep the standard current

A retirement standard decays the same way a contact list does, so re-verify its inputs on a schedule rather than treating it as set once. The two things most likely to drift are your deliverability baseline and your re-entry pipeline.

Re-check deliverability monthly. Because addresses go invalid at 2 to 3 percent per month, the validity assumption behind your touch counts erodes continuously. Verify addresses before a sequence starts, and audit the domain bounce rate at the end of each sequence so the 2 percent line is a fact you measured, not one you hope for. When platform requirements change - the bulk-sender complaint thresholds are an example - update the deliverability condition to match rather than carrying an old number.

Re-check the re-entry pipeline continuously, because that is where retired records turn back into pipeline. A retired prospect with a job-change watch is only as good as your ability to detect the move inside the 48-hour window. Rather than manually re-scanning every archived contact's employer, ask for the movers directly and reconnect while the signal is warm. Retirement done this way is not a dead end; it is a queue that refills itself when the buyer's context resets.