CampaignStack
Account Safety

Your LinkedIn accounts are safer with CampaignStack than doing it by hand.

CampaignStack caps every LinkedIn action at 80% of the platform's practical limit. Connection requests follow a dynamic daily budget of roughly 3% of your connection count (15 to 50 per day, scaled by account tier) under a rolling weekly cap of 100 to 200. New accounts warm up from 10% of budget, growing 4 points per active day, and accounts under 300 connections send nothing at all. Every account gets a real-time dashboard showing what's running, what's throttled, and why.

Start free trial
01 · Daily Budget System

Every action type has its own independent budget.

Connections, messages, likes, and enrichments are all tracked independently. Budgets adjust per account tier (Free, Premium, Sales Nav, Recruiter) and are recalculated daily at midnight.

Today's budgetWeekend: sends reduced 70%, research 30%
Activity timeline
Mar 22Mar 29Apr 5Apr 12Apr 19125102050100Weekend
  • Connections
  • Messages
  • Profile views
  • Post likes
  • Comments
  • Enrichments
80% of LinkedIn's cap, never more
Every action caps at 80% of LinkedIn's maximum. The remaining 20% is headroom for Signal Agent replies, manual actions, and safety-check retries.
Weekly rolling cap
Connection requests have an additional rolling 7-day cap: 100 (Free) to 200 (Recruiter). Today's daily limit is clamped to the remaining weekly allowance.
Under-300 gate
Accounts with fewer than 300 connections get all write budgets set to zero and reads halved. New accounts are protected until they build a natural connection base.
02 · Usage-Based Warm-Up

New actions start slow and ramp up with usage.

Each action type starts at 10% of its base limit. Active days add +4 percentage points. One idle day gets a grace period. After that, limits decay at 18% per idle day. Three weeks of consistent usage reaches 100%.

Warm-up progression
New action type: connection requests
Mar 1Mar 8Mar 15Mar 22Mar 290%25%50%75%100%100% capacity
10% cold start
+4% per active day
x0.82 decay per idle day
10% cold start
A brand new account sending its first connections gets 1 to 2 per day, not 15. The system treats every action type independently, so warming up messages doesn't warm up likes.
Automatic decay on inactivity
Stop using an action type for 7+ days and the limit drops significantly. 14 idle days hit the floor. This mirrors how a real human would re-engage gradually after a break.
Manual advance controls
Need to move faster? Each account's detail page has controls to advance warm-up by +15% or skip to full capacity. You stay in control.
03 · Human-Like Timing

Activity that looks like a real person at work.

Actions are distributed across business hours in the account's local timezone, following a weighted curve that matches real human work patterns. Weekends are quieter, mornings busier.

30 / 40 / 30 time-of-day distribution
30%
40%
30%
Morning
8:00-11:20
Midday
11:20-14:40
Afternoon
14:40-18:00

Daily start varies by up to 30 minutes so the account never clocks in at the exact same time. Each action is staggered 45 to 120 seconds apart with random jitter.

Timezone-aware business hours
Weekdays: 8 AM to 6 PM local. Weekends: 9 AM to 2 PM local. Actions outside this window are deferred to the next morning, never dropped.
Micro-breaks every 20 to 30 actions
After a random 20 to 30 actions, the system inserts a 5 to 15 minute pause. The threshold is re-randomized daily so there's no detectable pattern.
Weekend reduction
Sends drop 70% on weekends, research drops 30%. Just enough activity to keep the account warm without looking automated on a Saturday.
45 to 120s between actions
Consecutive actions for the same account are staggered with a 45-second floor plus up to 75 seconds of random jitter. No burst patterns.
+/-30 min daily start variance
The business-hours anchor shifts randomly each day, so the account never starts its activity at exactly the same time.
Wait node jitter
Workflow wait steps add 5 to 10% random jitter to their configured duration, capped at 30 minutes. Two leads never move through a workflow in lockstep.
04 · Dispatch Gate & Rate Limits

Four checks before every single action.

No action reaches LinkedIn unless it passes every gate. If any check fails, the action waits. It doesn't fail, it doesn't retry blindly. It waits until the gate opens.

4-check dispatch gate

Every LinkedIn action must pass all four checks before dispatch. If any check fails, the action stays in WAITING. Never failed, never lost.

✓
Session valid
Cookie verified daily against LinkedIn API
✓
Business hours
8 AM to 6 PM weekdays, 9 AM to 2 PM weekends (local time)
✓
Daily budget
used < effectiveLimit for this action type
✓
Weekly cap
Rolling 7-day sum for connection requests
-> Action dispatched
Instant rate-limit response
When LinkedIn returns a 429 or throttle signal, the budget for that action type is capped to current usage immediately. No more actions of that type go out today.
Automatic recovery
Temporary throttles restore after 1 hour. Daily limits reset at midnight PST + 1 hour buffer. Actions waiting in the queue auto-resume when capacity returns.
Session health checks
Every weekday morning, each account's LinkedIn cookie is validated against LinkedIn's API. Expired sessions are instantly suspended with all budgets zeroed. No actions go out on a dead cookie.
05 · Infrastructure Safety

Safety is more than limits. It's how the system runs.

From dedicated IPs to orphaned job cleanup, the infrastructure is designed to make every action look like it came from a human sitting at their desk.

Dedicated residential proxy per account
Every LinkedIn account gets its own residential IP bound to a consistent geography. No datacenter IP pools, no rotation, no fingerprint leaks across accounts.
Consistent browser fingerprint
Each account uses the same Chrome instance with a stable user agent, viewport, and cookie jar. LinkedIn sees one device, not a rotating fleet.
Encrypted cookie storage
LinkedIn session cookies are encrypted at rest. Decryption happens only at execution time, in memory, never logged.
Orphaned job cleanup
Jobs stuck in processing for more than 15 minutes are automatically swept to failed and the browser session is released. No ghost sessions, no resource leaks.
Budget-aware automation scheduling
Automations are scheduled within budget capacity. When an action type is at 80%+ usage, non-essential automations are deferred to the next day.
Tier-aware limits
The system auto-detects Premium and Sales Navigator subscriptions during daily health checks. Limits scale up with your LinkedIn plan with no manual configuration.

Ready to automate without the risk?

Start a 14-day free trial. Full safety system included on every plan.

Start free trial