LinkedIn Warm-Up Schedule Week by Week

Most accounts that get restricted were not sending thousands of messages a day. They jumped from zero to forty connection requests on day three. That single pattern, a sudden spike from no activity to heavy volume, is what triggers LinkedIn's anomaly detection more reliably than almost anything else.

A proper LinkedIn warm-up schedule is not complicated. It is just slow, and that slowness is the part people skip.

What the Warm-Up Is Actually Doing

LinkedIn's trust model is behavioural. The platform builds a baseline for every account: how many profile views it generates, how often it logs in, what its connection acceptance rate looks like, how frequently it sends messages. Any sharp deviation from that baseline is worth investigating, from the platform's perspective.

When you warm up an account, you are building a baseline that can absorb automation later. A profile that has spent four weeks logging in daily, visiting profiles organically, posting occasionally, and sending a handful of connection requests manually looks completely different to LinkedIn's systems than one that sat idle for six months and then fired off fifty automated requests on a Monday morning.

One thing to be clear about upfront: LinkedIn does not publish rate limits. The numbers in this schedule come from what we have seen work in our own outreach and what consistently causes problems when ignored. They are not official figures from LinkedIn.

Week-by-Week Schedule for a New Account

"New" means the account was created in the last 90 days or has fewer than 100 connections and minimal recent activity.

The "light automation" note for week three: if you are going to introduce a tool, start it with delays set high and daily caps set low. Do not flip it on at full volume. In our own testing, accounts that had the smoothest transitions introduced automation mid-ramp rather than waiting until week six and then switching everything on at once.

Weeks one and two should be entirely manual. Log in from your usual device, on a normal schedule. Fill out your profile completely if you have not already. A sparse profile with no photo, no headline, and no posts sending connection requests at scale is a flag by itself, before the volume even becomes a factor.

Week-by-Week Schedule for a Dormant Account

Dormant means the account exists, has connections, but has had little or no activity in the last three to twelve months. The ramp is shorter because profile age and existing connections provide context that a brand-new account simply does not have.

The most common mistake with dormant accounts is assuming the existing connection count makes them immune to spikes. It does not. The behavioural baseline resets during inactivity. An account that was sending forty requests a day eighteen months ago and then went quiet for a year needs to rebuild that baseline, just more quickly than a brand-new profile does.

The Numbers That Actually Matter During Warm-Up

Connection request volume is the obvious metric. Acceptance rate matters just as much, arguably more.

If you are sending twenty-five requests a day and your acceptance rate drops to a consistently low level, that is a signal to slow down, tighten your targeting, or both. LinkedIn interprets a low acceptance rate as evidence that your requests look spammy or untargeted. Accounts with consistently low acceptance rates are more likely to hit the weekly connection request cap that LinkedIn has been enforcing more aggressively since late 2024.

Keep your message reply rate in mind too. Automated messages sent to first-degree connections that receive no replies and no engagement do register. They are not counted the same way as connection requests, but patterns of zero-engagement sends contribute to the overall behavioural picture the platform builds around your account.

For a detailed look at what LinkedIn's detection systems actually examine, How LinkedIn Detects Automation in 2026 covers the specific signals worth understanding before you start any sequence.

How Ampliflow Enforces This in Practice

We built Ampliflow partly because we were running outbound for Ampliflow itself and kept hitting tools that either had no meaningful rate limiting or had limits you could easily override. Neither is useful when you are trying to protect an account that matters.

The platform caps daily sends at whatever limit you configure. The timing jitter is built into every action: connection requests go out at randomised intervals within your chosen window rather than firing at a fixed cadence. That matters because a perfectly regular send pattern is a detectable signature, and it is one of the easier things for LinkedIn's systems to flag.

Because Ampliflow runs on the Unipile API rather than a browser extension, your laptop does not need to be open for the schedule to run. There is no extension injecting JavaScript into LinkedIn's interface, which is a meaningful architectural difference. Browser extensions vs cloud automation safety explains the specific risk profile of each approach in more detail if you want to go deeper on that.

The real-time account safety scoring watches your acceptance rate, reply rate, and daily volume together. If any combination starts looking anomalous, the system flags it before a restriction happens rather than after. That is the practical difference between a tool built with safety as a constraint and one optimised purely for volume.

The Mistakes We Keep Seeing

People import a Sales Navigator list of five hundred prospects and try to run the full sequence before the account is warm. The import itself is fine. Running a sequence against it on week one is not.

The second pattern: turning off the warm-up early because nothing bad happened in week two. LinkedIn's enforcement is not always immediate. An account can look fine for ten days and then get a soft restriction on day eleven when cumulative volume crosses a threshold. Consistency through the full schedule is the point, not just surviving the first few days.

Third, and this one is easy to overlook: ignoring the content side entirely. An account sending connection requests but posting nothing, commenting on nothing, and viewing zero profiles outside of automated actions looks hollow. Spending five minutes a day on genuine engagement during the warm-up period is worth doing. It is not about gaming an algorithm; it is about looking like a person who actually uses the platform.

For a broader read on what actually triggers restrictions versus what is mostly noise, Is LinkedIn Automation Safe in 2026? An Honest Risk Breakdown takes an honest look at the current risk landscape without inflating the danger.

After the Warm-Up: Sustainable Ceilings

Once you are past week six on a new account, 30-40 connection requests per day is a sustainable ceiling for most profiles. We cap our own sends at 35 per day as a default. The marginal volume above that does not justify the marginal risk, and that is a judgement call we have made for our own outreach, not just a platform recommendation.

Sales Navigator users often assume the paid subscription raises their limits. It does not, at least not in any documented way. What Sales Navigator gives you is better targeting, which improves acceptance rates, which indirectly lets you run at a higher volume without triggering low-acceptance-rate flags. That is a real benefit, just a different mechanism than most people expect.

If you are running multiple accounts or want a more detailed breakdown of per-account versus aggregate volume, Safe LinkedIn Automation Limits in 2026 has the specifics.

A Note on Tools and Pricing

Linked Helper costs $15 a month and Octopus CRM runs about $10. Both are cheaper than Ampliflow's founding price of $19 a month, and if you are comfortable running a browser extension and managing your own rate limits manually, either is a reasonable choice. Cheaper tools being cheaper is just true, and pretending otherwise would be dishonest.

The architectural difference is cloud execution via the Unipile API: no extension fingerprint, no need for an open browser session, and rate limiting that runs at the platform level rather than relying on you to remember to cap it. For founders running outbound on a primary account they cannot afford to lose, that trade-off is worth the price difference. For someone testing outreach on a secondary account and watching every dollar, Linked Helper is not a bad answer.

The founding member price of $19 a month is available for the first 100 accounts before public launch, when Starter moves to $39 a month. That is a saving of about $240 a year compared to the launch price, and roughly $700 a year compared to tools like Dripify ($79/mo) or Expandi ($99/mo) that offer broadly similar feature sets at a significantly higher price point.

The warm-up schedule itself works regardless of which tool you use. The schedule is the discipline. The tool is what enforces it so you do not have to remember to on a Tuesday morning when you are already late for something else.