Follow up until they answer, then stop automatically

The Wave 1 cadence builder does two kinds of step and does them properly: an email step, and a task step assigned to the recruiter who owns the record. That combination covers the work that actually converts. Email carries the message; the task step is where a call, a LinkedIn look or a note to the client belongs, because those are things a person does and a sequence should only remind them to do.

Steps are spaced in business days rather than calendar days, so a Friday step does not land on a Sunday morning. Sending windows are set per cadence and respect the candidate's time zone where it is known. A cadence can branch on a simple condition, such as whether the candidate opened but did not reply, without turning into a flowchart nobody on the desk can debug.

Mail sends from your own connected mailbox through Gmail or Microsoft 365, not from an anonymous system address. That matters for two reasons. Replies arrive where the recruiter already works, and the message carries the reputation of a real mailbox rather than a shared sending pool.

Sync runs in both directions. A reply the candidate sends to your Outlook inbox is read back into the CRM, attached to the candidate record and counted as a reply for cadence purposes, even if the recruiter never opens Surhires that day. A message the recruiter writes by hand from their mail client also lands on the record, which is what stops the CRM history from being a partial account of the relationship.

A follow-up that arrives as a brand-new email with a new subject line reads as a mailshot. A follow-up that appears underneath the original, in the same thread, reads as a person who has not forgotten. Surhires threads each step onto the first message using the same conversation reference the mail clients use, so the candidate sees one conversation and can scroll back to what you originally asked.

Threading also protects the recruiter. When the candidate finally replies on touch four, the whole exchange is in front of them rather than split across four subject lines, and the record shows what was promised at each stage.

The single most damaging failure in outbound recruiting is chasing somebody who already answered. A cadence halts on a genuine reply from the candidate, and the detection is deliberately conservative about what counts. Automatic out-of-office responses, delivery receipts and mailing-list bounces do not stop a sequence; a human reply does, including a reply that arrives on a different channel or from a different address belonging to the same record.

Stopping is not the only outcome. A reply raises a task for the owning recruiter, moves the record where the stage machine expects it to move, and records the time to first reply for the reporting. A bounce suppresses the address and flags the record for correction rather than repeating the same failure three more times.

Outbound email fails quietly. Nothing errors; the messages simply stop reaching inboxes, and a desk can spend a month blaming the market. Setup includes verifying that the sending domain publishes SPF, DKIM and DMARC records that actually align, because a sending domain without them is treated as suspicious by every major provider.

After setup the product watches the signals that predict trouble: bounce rate by domain, complaint rate, sudden volume increases from a new mailbox, and reply rate falling while send volume holds. A new sending domain is warmed by starting at low daily volume rather than sending three thousand messages on day one. None of this is glamorous and all of it is the difference between a cadence that works and a cadence that is technically running.

Cadences do not fix a bad first message. If the approach does not say what the role is, what it pays and why you thought of this person, sending it four times makes the problem four times larger. The reporting will show you that, which is uncomfortable and useful.

There is also a hard ceiling on volume. Gmail and Microsoft 365 enforce daily sending limits per mailbox and we do not attempt to route around them. A team that wants to send more than those limits allow needs more mailboxes or a dedicated sending domain, and we will say so during onboarding rather than after a suspension. SMS and LinkedIn steps are not part of Wave 1; the reasons for that are set out on the multi-channel page rather than glossed over here.