By Most of the outreach advice still circulating on LinkedIn was written for a platform that no longer exists. I run Linkdify, which sends connection requests and messages for solo founders and small B2B teams, so I get to watch what happens after the advice gets applied. The same six tactics keep showing up in campaigns, and they keep underperforming. Here is what stopped working, and what I see working instead.
1. The personalized connection note
This one hurts because every guide still recommends it. We looked at 8,400 connection requests sent through Linkdify in August and September. Requests with a note were accepted 8.6% of the time. Requests without a note were accepted 48.6% of the time. The blended average was 25.2%, about one in four, but the gap between the two groups is the whole story. The numbers are in our outbound benchmarks report.
My read: a note turns a neutral request into a pitch. People accept neutral requests from strangers all day. They decline pitches. Save your words for after the accept.
2. "I noticed we're both in SaaS"
Manufactured familiarity. You are not the first person to notice they work in SaaS, and they know a tool pulled that line off their profile. In 2023 these openers got replies because they were rare. Now the recipient has forty of them in the inbox and recognizes the shape before finishing the first sentence.
What works now is a message that would only make sense sent to this person this week. A post they wrote. A hire they announced. Even a plain question about something on their profile beats fake kinship. If you cannot find anything specific, send something short and plain and skip the familiarity entirely. Honest and generic beats personalized and false.
Want LinkedIn outreach that runs itself?
Linkdify finds your leads, writes personalized messages, and sends them at safe, human-paced limits.
3. The cold email cadence, pasted into a chat window
Cold email taught a generation of sellers to send "just bumping this" and "should I close your file?" LinkedIn messaging is a chat thread. Five unanswered messages stacked in one thread look exactly like what they are, and the recipient scrolls past the pile feeling pressured, which is the opposite of what a connection-based channel is for.
Our timing data points the same way. Across 2,511 messages, the median reply came in about 14 hours, and most replies landed within the first day. If a thread is quiet after a few days, one different follow-up is reasonable. A fourth one is noise. I would rather a founder spend that effort on the next twenty people than on wearing down the last one.
4. Sending 100 requests a day
Three years ago you could run a fresh account hard and mostly get away with it. LinkedIn has since tightened invitation limits and gotten much better at spotting pacing that no human produces. An account that goes from zero to a hundred requests a day gets flagged, and a flag on the account you sell from is a different kind of expensive than a bad campaign.
This shaped how I built Linkdify. We cap daily sends at a level LinkedIn tolerates, spread actions across the working day, and slow cold accounts down further. Founders sometimes ask me to raise the limit. I usually point out that at a one-in-four acceptance rate, 20 requests a day is roughly a hundred new connections a month. That is already more conversations than one person can run well.
5. Automated likes and the "congrats on the new role" message
Engagement automation promised warmth before the pitch. Like a few posts, endorse a skill, then connect. From the other side, the sequence is obvious. A stranger who liked your last three posts inside the same minute and then sent a request is not a fan. The congratulations message that lands the day a job change hits LinkedIn has become so common that it now reads as a sales trigger, because it is one.
Job changes are still a strong signal. New role, new budget, new problems. I just think you should use the signal as a reason to send something useful, not as a pretext for a congratulations nobody asked for.
6. Reporting on acceptance rate
Dashboards love acceptance rate. The number is big and it moves. It is also the wrong number. An accepted connection that never replies is worth nothing, and an acceptance rate inflated by no-note requests tells you nothing about whether your pitch lands. The number that pays rent is replies. On our platform the reply rate across those 2,511 messages was 14.6%. That is the metric I watch, and it is the one the Linkdify reports are built around.
The second-order effect is on behavior. When you optimize for accepts, you write blander requests. When you optimize for replies, you write a better first message and you answer fast, because the reply window mostly closes within a day.
Want LinkedIn outreach that runs itself?
Linkdify finds your leads, writes personalized messages, and sends them at safe, human-paced limits.
What all six have in common
Every tactic on this list worked once because it was rare. Then it got automated at scale by people who did not care who was on the other end, and it stopped. That is the pattern, and the next clever trick will follow it. So the question I ask about any new outreach tactic is simple: would it still work if ten thousand people did it tomorrow? If the answer is no, treat it as short-term arbitrage and plan to drop it. If the answer is yes, it is usually because the tactic requires actually paying attention to the person. Nothing on LinkedIn has stopped rewarding that.