File path · /Users/forgebot/forgeapps/.forgebot/skills/mailchimp/preview.html
ForgeFX SimulationsMailchimp / capability preview

Less campaign prep.
More human control.

A pretend AES campaign, from a Slack request to a draft ready for human review. Examples to react to—not a finished integration.

FICTIONAL DEMO DATA · All audience counts, campaign copy, metrics and conversations below are invented. No Mailchimp data is loaded. No buttons send, schedule or save anything to Mailchimp.
01 / Guided audience selection

Pick the right people—not the biggest list.

Example Slack exchange

You: “Who should receive our AES invitation?”

ForgeBot: I’d start with subscribed contacts tagged “AES interest”: 240 people in this example. It’s more relevant than the entire 1,800-person audience.

Do you want all 240, or the 90 also tagged “Training leader”?

No audience changes made. I’ll show the selection before attaching it to a draft.

Proposed selection

SubscribedAES interest
Recommended segment240 contacts
Narrower alternative90 contacts

Exclude unsubscribed and suppressed contacts. Event interest alone is not marketing consent.

Example tagging guidance

“I found 30 already-subscribed contacts in your selected event group without the AES interest tag. Review the matches and any automation triggered by this tag before applying it. This is a proposal, not a completed change.”

Discussion: should I recommend a segment first or always ask you to choose?
02 / Guided draft creation

Turn a short request into a usable draft.

You: “Draft a short AES invitation that asks for a meeting.”

ForgeBot: Here’s a first pass for the 240-contact segment. Before it becomes a Mailchimp draft, I need the confirmed event details and booking link.

Subject: At AES? Let’s talk hands-on training.

Preview text: Explore where simulation could help your team practice.

Status: Proposed copy · not saved · not approved

Discussion: draft first with visible gaps, or collect all details up front?
03 / Existing-draft editing

See what changed. Keep what works.

You: “Make the draft shorter and less salesy.”
Before · fictional draft v1

Discover the transformative potential of immersive simulation and unlock the next generation of workforce development with our industry-leading training solutions.

Proposed revision · v2

Give your team a safe place to practice difficult jobs. Let’s talk at AES about what that could look like for your training.

Unchanged: audience, subject and meeting link.

Next: review this revision before saving it to the selected draft. A material edit means another preview.

Discussion: show the changed passages by default, with the full draft available on request?
04 / Preview and human handoff

Ready for review does not mean ready to send.

Example review checklist

Try these controls. They change this page only and reset on reload.

Review incomplete in this example.

Example handoff response

“Your draft and audience are ready for your final check in Mailchimp. Review the rendered email and recipient selection there. An authorized person must send it.”

No Send or Schedule control. The agent cannot launch a campaign, even after someone approves a draft in Slack.

In the real workflow, this handoff would link to the verified draft. No fabricated Mailchimp link is used here.

05 / Plain-English performance reporting

Tell me what happened—and what it means.

You: “How did the AES invitation do?”
Fictional results · first 7 days after human send
240Sent
235Delivered
24Unique clickers
2Unsubscribed

The useful takeaway

People clicked more often than in the comparison campaign. That suggests the invitation earned interest—but clicks do not tell us how many meetings were booked.

Next move: check booking results before deciding whether the invitation drove qualified meetings.

Unique click rate: 10.2% versus 6.0% in the fictional prior campaign. Same seven-day window; different audiences, so this is not proof that the new wording caused the improvement.

Details without the jargon

Delivery failures5
Reported unique opens110
Meeting bookingsNot connected
How these figures are calculated

Delivery rate: 235 ÷ 240 = 97.9%. Click rate: 24 ÷ 235 = 10.2%. Unsubscribe rate: 2 ÷ 235 = 0.9%. Prior campaign: 12 unique clickers ÷ 200 delivered = 6.0%.

Opens can be inflated by privacy protections; security scanners can also inflate clicks. Treat engagement as a signal, not a verified buying decision. All numbers on this page are fictional.

Discussion: short takeaway first, with expandable metrics and caveats underneath?
06 / Post-event follow-up

Continue the conversation without guessing consent.

You: “Prepare a follow-up for people interested in a demo.”

ForgeBot: I’d use the subscribed contacts with a confirmed “Demo requested” tag, rather than everyone who opened the invitation. Please confirm who actually asked for a demo and whether any meetings are already booked.

I can propose a segment and follow-up draft. No tags, contacts or campaigns have been changed.

Example follow-up copy

Subject: Your training simulation conversation

Thanks for your interest in a ForgeFX demo. What task would you most like your team to practice? Share a little context and we can focus the conversation on what matters to you.

Discussion: keep follow-ups as individual requests, or guide the whole pre/post-event sequence together?