Our dashboard once showed a 733% open rate.
No, the copy was not that good.
An open rate above 100% means the denominator is broken, not the message. When leads get moved between sequences by hand, the reporting stops describing reality.
Here's the mechanic most teams miss. Any reply exits a sequence. That includes "thanks", "hi" and "sure". So a friendly one-word answer quietly cancels every follow-up planned behind it.
How we wire it now:
- Connection request campaign. Acceptance is the trigger. Nothing else happens here.
- DM campaign. Starts automatically when the request is accepted.
- Extension campaign. A polite but empty reply moves the lead here, so the conversation keeps going.
- Naming. Same name as the parent campaign plus "_Extension", so the numbers stay attributable.
The day we fixed the wiring, the vanity numbers got smaller and the real ones got clearer.
If a metric looks too good, check the plumbing before you celebrate.
What did the dashboard show?
The real screen, redacted.
| Stat | Value |
|---|---|
| Leads in campaign | 758 |
| Leads launched | 100% (758) |
| Leads reached | 54.2% (411) |
| Messages sent | 33 |
| Messages not sent | 0 |
| Delivered | 100% |
| Open rate | 733.3% (242) |
| Clicked | 0% |
| Replied | 242.4% (80) |
| Invitation accepted | 58.9% (242 of 411) |
| Interested | 0% |
Only one number on that screen was real and worth reading: 58.9% invitation accepted. Everything below it inherited the broken denominator.
How can an open rate be over 100%?
- 242 "opens" against 33 sends.
- The tool counted every profile visit after acceptance as an open.
- Leads had been moved between sequences by hand, so the denominator no longer described anything.
Why do one-word replies kill your follow-ups?
Any reply exits a sequence. That includes "thanks", "hi" and "sure". A friendly one-word answer quietly cancels every follow-up planned behind it. Most of the 80 replies were one word, and most were dead ends.
How should LinkedIn outbound campaigns be wired?
Three campaigns, one name.
| Campaign | Trigger | What it does | Name |
|---|---|---|---|
| Connection request | — | Sends the request. Acceptance is the only trigger. | Rep_Geo_Aud_CR_Mon |
| DM | Request accepted | Starts on its own | Rep_Geo_Aud_DM_Mon |
| Extension | Polite but empty reply | Keeps the conversation going | Rep_Geo_Aud_DM_Extension_Mon |
Before: one sequence, moved by hand. The rep moved leads, and the denominator broke.
After: three wired campaigns feeding one dashboard, with real rates and the same denominator. Nothing depends on someone remembering.
What happened when we fixed the wiring?
- The vanity numbers got smaller and the real ones got clearer.
- Open rate fell to something believable.
- Reply rate split into "any reply" and "interested".
- Meetings stayed the same. That was the point: nothing real had changed.
Why does the campaign name matter here?
The extension carries the parent's name plus "_Extension". Without that, the extra campaign lands in the error bucket and the dashboard shows one profile doing half the work it did.
If a metric looks too good, check the plumbing before you celebrate.