How to do LinkedIn outreach without getting banned
Watched a friend lose his LinkedIn account last year. Permanent ban, using a "safe" automation tool at "safe" daily limits.
Here's what I learned from it.
1. LinkedIn doesn't care about your daily limits
Everyone asks "how many requests per day is safe", which is the wrong question. LinkedIn's detection is based on acceptance rate and reply rate, not volume. You can send 20 a day that get ignored and get flagged, or send 100 a day that get accepted and replied to and be completely fine.
The numbers that matter are acceptance rate above ~40% and reply rate above ~20%. Below that, LinkedIn's systems start treating you like a spammer, and statistically speaking, you are one.
2. What gets you flagged is bad targeting
My friend blamed the tool and switched three times before realizing the tool was just fine, his list was the problem. When you scrape 2,000 "Heads of Product at SaaS companies" and blast them, 95% of them don't know you, don't care, and ignore you. LinkedIn watches that ignore rate and concludes you're a bot.
3. The fix: stop volume, start targeting high-intent leads
I send 125 requests a week, acceptance rate is at 72% and reply rate at 55%, zero restrictions. The 125 are people who commented on a competitor's post in the last 2 days, posted themselves about the problem I solve, or engaged with specific creators in my niche. Same ICP as everyone else, but with an additional intent filter. I'm catching them when they're already thinking about the problem.
You can identify those people either manually (painful) or you use IbexAI to find them at scale, every day, and on autopilot.
IbexAI scans LinkedIn daily and surfaces people who expressed interest in your product.
___
PS: Follow me on X along my building in public journey.
What are the intent signals you're seeing the most success with? Don't see any in this post, and don't see any on your website.
E.g., an ICP commenting on a competitor's post
What stands out to me is that avoiding a LinkedIn ban is less about finding the “perfect” automation setup and more about making outreach behave like genuine networking.
The biggest mistake I see is optimizing for volume before optimizing for relevance. If the targeting is weak, even a perfectly written message becomes spam. A smaller list of highly relevant prospects, personalized around their role, business, or an actual pain point, will usually create better conversations than hundreds of generic connection requests.
I’d also add that engagement quality should be treated as an important signal. If people are accepting requests but never replying, that’s probably not a volume problem—it’s a targeting or messaging problem.
For me, sustainable LinkedIn outreach comes down to three things: relevance, moderation, and real conversations. Automation can help with repetitive tasks, but it shouldn’t replace the human part of the process.
This is a really interesting idea. I especially like the problem you're trying to solve. If you could improve just one thing before pushing this to more users, what would it be?
The intent-based targeting point is definitely the strongest takeaway here. Reaching out to people who have recently engaged with a relevant problem is much more effective than blasting a broad ICP list. I’d just be careful with the specific acceptance/reply-rate thresholds—LinkedIn doesn’t publicly confirm those as official detection limits. Focusing on relevance, personalization, and genuine engagement seems like the safer long-term approach.
So, Identifying an Intent is the key!
Yes!
The intent-based approach is interesting. Instead of chasing volume, identifying people already discussing a relevant problem seems much more sustainable. GeekyAnts has also been exploring AI-driven approaches to making B2B outreach more targeted.