Anatomy of a multi-step campaign
- Trigger: A lead enters the sequence, via manual import, a post-commenter scrape, or a signal trigger.
- Step 1 (Day 0): Send a connection request with a personalized, specific signal reference.
- Wait condition: Up to 5 days for acceptance.
- Branch A (accepted): Continue to Step 2.
- Branch B (ignored after 14 days): Withdraw the invitation and exit the sequence.
- Step 2 (Day +1 after acceptance): Send the first DM: a soft pitch paired with a small piece of genuine value.
- Step 3 (Day +5): Send a value drop, an article or a relevant data point, no direct ask.
- Step 4 (Day +10): A soft close with door-open language, low-pressure, easy to decline without friction.
- Exit conditions: Any reply at any step exits the sequence immediately, handing off to a manual, human-written reply.
Building it in Leadsforlinked
- From the Diamond Pro plan (79 EUR/mo), go to Campaigns → "New campaign"
- Choose a trigger source (post-commenter list, event attendees, or a manual CSV import)
- Add steps via the drag-and-drop builder
- Configure delays, exit conditions, and working hours
- Set a daily volume cap (16/day per account is a reasonable, sustainable default)
- Activate. The campaign runs in your Chrome browser while it's open.
Smart delays and account safety
Leadsforlinked randomizes delays between actions rather than using fixed intervals:
- 2-8 seconds between profile views
- 3-15 minutes between connection sends
- 30-90 seconds between DM sends
- Skips weekends by default
- Pauses automatically if a LinkedIn rate-limit signal appears
The reasoning behind randomization is straightforward: a real person's timing between actions is never perfectly even, so identical, clockwork intervals are one of the more obvious automation signatures a detection system can key on. Randomized delays within a plausible human range, combined with pausing when LinkedIn itself signals a slowdown, keeps activity looking like normal, if efficient, manual use.
Common mistakes
- No exit condition on reply, so the campaign keeps messaging someone who already responded
- Delays set too tight (1-2 seconds looks robotic, not efficient)
- Sending all 5 messages within a single week (too dense, and it reads as pressure rather than genuine interest)
- Identical or near-identical messages across all 5 steps, which signals a template even to a recipient who never noticed the first one
Frequently asked questions
How many steps should a first LinkedIn campaign have?
3 to 5 steps is a reasonable starting range: a connection request, one or two value-oriented DMs, and a soft close. Fewer than that leaves replies uncollected from prospects who needed a second or third touch; many more than five starts feeling like pursuit rather than outreach.
Why do smart delays matter for account safety?
Identical, fixed-interval timing between actions is one of the clearest automation signatures LinkedIn's detection can key on. Randomizing delays within a human-plausible range, and pausing entirely if a rate-limit signal appears, keeps the activity pattern closer to what a person actually does.
What happens if a prospect replies mid-sequence?
The sequence should exit immediately and hand control to a real, manual reply. Continuing to send scheduled follow-up steps to someone who already responded is one of the most common and most avoidable automation mistakes, and it reads as obviously robotic to the recipient.
Should every step in a sequence say something different?
Yes. Identical or near-identical messages across steps signal a template even to a recipient who didn't notice the first one, and they waste the chance each step represents to add a new, specific reason to respond.
Build your first campaign.
100 free leads on signup. Campaigns from Starter, AI Responder on Diamond Pro. Build, activate, and monitor in about 30 minutes.
Start free