Only available on sites with the Outreach feature enabled (admin: Cross-border trade → Outreach). If you cannot see it, this site does not have it.
Sender identity and warm-up (on the Outreach page)
- Sender address: the one set for outreach (fromEmail in update_outreach_settings); if none is set, the Resend sender address from your credentials.
- Is the domain right: use a separate cold-email domain (for example getyourbrand.com). Using your website domain shows a warning (spam flags on cold email would hurt your order emails); a platform domain must be replaced with your own.
- DNS: click Check DNS to see whether DKIM, SPF, bounce MX and DMARC are in place for the sender domain. Add any missing record with the values from your Resend dashboard; for DMARC, v=DMARC1; p=none is a fine start.
- Warm-up curve: a new sending setup sends at most 20 a day for the first 7 days, then 20% more each day up to your daily cap. The chart shows which day you are on and when you reach the cap.
- Left today: sent today, cap today and how many are left.
Ask your AI: get_outreach_sender_identity (add checkDns to check DNS); change the sender or the daily cap with update_outreach_settings.
Subject A/B tests
Write two subject lines for the same outreach email: contacts with an even enrollment number get A, odd numbers get B (fixed groups, half and half). A can be left empty, meaning the original subject of that email.
- Each group first needs enough sends (30 by default), then a waiting time (48 hours by default, so replies can come in).
- The winner is picked by reply rate; if equal, by open rate; if still equal, A stays.
- After that, this email always goes out with the winning subject.
Open rate is only indicative: Apple Mail preloads images automatically and clients that block images are not counted, so reply rate comes first.
You decide: on the Subject A/B card click Use A or Use B to pick directly, or Back to automatic pick.
Ask your AI:
- Create or change: set_outreach_subject_ab (sequence, which email, subjectB; optional subjectA, minSample per group, waitHours, locale to include only contacts of one language).
- Results: get_outreach_subject_ab (sends, opens and replies per group, the winner, what is still missing).
- Decide or stop: winner a or b picks directly, auto goes back to automatic; stop true ends the test (the email goes back to its original subject, stats are kept).
- Changing a subject after sends have gone out keeps the old round on file and starts a new round from zero.
Only write true things in subject lines; variables such as {{company}} work, so fill them in when enrolling contacts.