Message sequences

For each persona, Wrenwire generates a message sequence: a series of emails, or a mix of LinkedIn actions and emails, timed to reach a prospect over several touches.

Sequence steps

A typical sequence moves through steps such as:

• Intro: the first message.
• Follow-up and Follow-up B: later touches, usually from a different angle.
• New thread: a fresh subject line rather than a reply.
• Breakup: a closing message that lets the prospect step away gracefully.

“A different angle” is the part worth reading before you approve. A follow-up that repeats the first message more loudly is the most common way a sequence fails: the recipient already declined that argument by not answering. The generated follow-ups therefore change what they lead with, usually moving from the outcome to the specific situation, or the other way round.

Email steps use merge variables such as first name and company, so each message is personalized per prospect without being rewritten by hand.

LinkedIn sequences

LinkedIn sequences combine connect requests carrying a note, messages sent once the connection is accepted, and, on paid LinkedIn plans, InMail for reaching people outside your network.

If a connect request is not accepted within a number of days you set, the sequence can fall back to a backup email, so the prospect is still reached through another channel rather than being silently dropped.

Sending on LinkedIn is not automated. A person sends those messages, which is a deliberate constraint rather than a missing feature: it is what keeps the channel inside the platform’s terms.

Quality review

Before a sequence can be activated, the quality review scores it and returns specific feedback. You can ask for a flagged message to be improved, or edit it yourself, and a sequence stays in “awaiting review” until this has run and the issues are addressed.

The review is a filter, not a formality. Weak versions are dropped before they reach you, so what you read is what survived rather than the first attempt.

Edits are versioned. An earlier draft is never lost, and you can see what changed between two versions, which matters when a rewrite improves one thing and quietly breaks another.

What the product adds on the way out

Two blocks are added to messages rather than written by the model, and they behave differently on purpose.

The Article 14 GDPR notice, telling the recipient where their data came from, is added into the body of the first message when the sequence is saved. It has to be part of the stored text because messages leave through external tools and by hand, so a placeholder would ship literally.

The sender’s legal details for the German, Austrian and Swiss markets are resolved at send time instead, so that changing your company address does not leave stale details inside series generated months earlier. A campaign can deliberately waive that block; doing so does not remove the Article 14 notice, which is a separate obligation. See Compliance.

After you send

Every generation and review run is logged with what it cost, visible in the AI journal alongside your credit usage.

Replies, unsubscribes and bounces feed the suppression list, so somebody who opts out of one campaign is not contacted again by another. That list is organization-wide, not per campaign, and the filter runs before every hand-off of contacts to sending.