Cold email vs LinkedIn outreach: reach, risk, cost and reply rates

Short answer and comparison table

Tibexa builds cold email systems and uses LinkedIn as a supporting channel, so our default leans toward email. Platform and vendor figures below were checked on 18 October 2026.

Reach and volume limits

Email limits come from the mailbox provider. Google documents a cap of 2,000 messages a day per Workspace user, applied over a rolling 24 hours, with 500 a day on trial accounts, and says the limits can change without notice (Google Workspace). Nobody should send anywhere near that from one mailbox. Woodpecker found that mailboxes averaging 1 to 20 emails a day had a 2.0 percent median reply rate, against 0.5 percent for mailboxes sending 101 to 200 a day (7,676 mailboxes that sent at least 500 emails in 90 days; Woodpecker). That is a correlation, but it matches what we see: volume comes from adding mailboxes, and pushing each mailbox harder costs replies. In our AI feedback SaaS work, 1,000+ emails a day came from a multi-inbox setup (case study).

LinkedIn's limits are less visible. Its help page on invitation restrictions says LinkedIn limits the invitations you can send, for Basic and Premium members alike, and gives no number (LinkedIn Help). The same page sets a 30,000 cap on first-degree connections and says most restrictions lift within a week, but LinkedIn cannot remove one on request, and too many outstanding invitations can mean waiting up to a month. Blog posts quote weekly caps of around 100. We found no such figure on LinkedIn's own help pages, so we do not use one.

Messaging people you have not connected with is a separate route. LinkedIn's help page says free accounts can send up to five personalized invitations of 200 characters (the page gives no time period), Premium accounts can send an unlimited number of personalized notes of 300 characters, and InMail credits depend on the subscription (LinkedIn Help). The Sales Navigator plan page lists 50 InMail messages a month.

Deliverability risk vs account risk

The risk in email is to your sending reputation. Google's sender guidelines require bulk senders (5,000 or more messages a day) to authenticate with SPF, DKIM and DMARC, use TLS, support one-click unsubscribe, and keep the spam rate in Postmaster Tools below 0.30 percent, with 0.10 percent recommended (Google). Microsoft applies similar authentication rules to senders of more than 5,000 emails a day to Outlook.com consumer domains, and from 5 May 2025 sends non-compliant messages to Junk (Microsoft Tech Community). The formal rules start at 5,000 messages a day, but the spam-rate targets are a sensible ceiling at any volume.

List quality is a deliverability issue too. In Woodpecker's data, campaigns bouncing under 2 percent had a 1.6 percent median reply rate, and campaigns bouncing over 20 percent had 1.0 percent. When a domain's reputation drops, you retire it and warm another, which costs warmup time and leaves your people untouched. Our deliverability playbook covers the setup.

The risk in LinkedIn is to a named person's account. Section 8.2 of the User Agreement (effective 3 November 2025) bars using "bots or other unauthorized automated methods to access the Services, add or download contacts, send or redirect messages" (LinkedIn User Agreement). LinkedIn's help page on prohibited software says it does not allow crawlers, bots or browser extensions that "scrape, modify the appearance of, or automate activity on LinkedIn's website," that any member using such tools is in violation, and that accounts may be restricted or shut down (LinkedIn Help). The tools "may become non-operational without notice."

The account at stake holds a person's network and their inbound inquiries. Sending messages by hand inside LinkedIn is ordinary use of the service. Automation tools are what the rules prohibit, and the choice to accept that risk belongs to you and the person whose profile it is. Our advice is to keep LinkedIn manual or close to it.

Cost

Sources: LinkedIn and Smartlead. LinkedIn says its prices are estimates that may exclude tax and can change. Smartlead's FAQ says all plans include unlimited email accounts, and its page also shows a grid at $59 per plan that conflicts with the plan cards, so confirm at checkout.

By our arithmetic, the Pro plan's cap of 90,000 emails for $94 is about a tenth of a cent per email in tool cost. Domains, mailboxes and data are extra and not priced here. The expensive parts of email are the list and the offer. LinkedIn's cost sits in seats and people time, because every seat belongs to one person who can only contact so many prospects.

Reply rates from published benchmarks

These figures are third-party and vendor-reported. Each vendor sells the channel it measures, and the rates use different denominators, so do not subtract one from the other.

Sources: Woodpecker, Smartlead and Expandi. Woodpecker's 1.5 percent headline and its 1.0 percent first-email figure come from different cohorts. Expandi sells a LinkedIn outreach tool that its page promotes for sending "300+ connection requests weekly," its page says it is not an official LinkedIn product, and LinkedIn's help page prohibits tools that automate activity on its site. Its connection-note reply rate fell from 3.5 percent in May 2025 to 2.2 percent in April 2026, and the page does not say why.

A rough illustration per 100 prospects, assuming each accepted connection gets one message and the email sequence is three emails with no overlap in repliers. Email: 1.0 + 0.9 + 0.6 gives about 2.5 replies. LinkedIn: 28.5 accepted connections at a 10.4 percent message reply rate gives about 3.0 replies, plus 3.0 note replies, for about 6.0. On these datasets, LinkedIn gets more than twice the replies per prospect contacted. It also reaches far fewer prospects per week.

A reply is not a meeting, and neither vendor in this table publishes a positive-reply or booking rate. Our planning benchmark is 2 to 3 percent of email sends replying, 15 to 50 percent of replies positive, and 30 to 50 percent of positive replies booking, which assumes a verified list and a tested offer.

Compliance: CAN-SPAM and the LinkedIn User Agreement

The FTC's CAN-SPAM guide lists eight requirements: no false or misleading header information, no deceptive subject lines, identify the message as an ad, tell recipients where you are located, tell them how to opt out, honor opt-outs promptly, remember that subscribers and members can opt out too, and monitor what others do on your behalf (FTC). It states that "the law makes no exception for business-to-business email," that you must honor an opt-out within 10 business days, that your message must include a valid physical postal address, and that each violating email can bring penalties of up to $53,088 (the figure on the page as updated January 2024). The last requirement matters if an agency sends for you, because you cannot contract away your responsibility.

LinkedIn's rules are contractual. Section 8.1 of the User Agreement tells members to comply with all applicable laws, including anti-spam laws, and section 8.2 bans bots and automated methods. Section 3.4 says LinkedIn "reserves the right to restrict, suspend, or terminate" an account that breaches the contract.

CAN-SPAM covers the United States. Recipients elsewhere are covered by other laws that this article does not address, so take legal advice on the countries you target. Put the physical address, a working opt-out and a suppression list into every sequence before the first send.

When to use one channel and when to combine

When we combine them, email goes first because it scales and puts no personal account at risk. After a non-reply, a profile view or a manual connection request follows a few days later, and a short LinkedIn message goes out only if the person accepts. Replies from both channels land in one inbox so nobody answers twice. The sequence design is in the multichannel outbound playbook and the LinkedIn side in the LinkedIn outreach playbook.

Most of the work in email is infrastructure. Our Rapid Lanyard campaign ran on 39 domains with Microsoft and Google inboxes split between them and reached 110+ positive replies (case study). If you want that built and handed over, see our cold email systems service. For the people side of the same decision, read outbound agency vs in-house SDR team.

Questions

Does cold email or LinkedIn get higher reply rates?

In vendor datasets, LinkedIn messages after connecting get higher reply rates (Expandi reports 10.4 percent of messages) than cold emails (Woodpecker reports a 1.5 percent median). The denominators differ, and LinkedIn reaches far fewer prospects per week.

Is LinkedIn automation allowed?

No. LinkedIn's User Agreement bans bots and unauthorized automated methods, and its help page says accounts using such tools may be restricted or shut down. Sending messages by hand within LinkedIn's limits is ordinary use.

Does CAN-SPAM apply to B2B cold email?

Yes. The FTC's guide says the law makes no exception for business-to-business email. You need accurate headers, a physical address, an opt-out, and prompt honoring of opt-outs within 10 business days.

How many cold emails can one mailbox send per day?

Google caps a Workspace user at 2,000 messages a day. Woodpecker's data shows higher reply rates at lower per-mailbox volume, 2.0 percent median at 1 to 20 emails a day against 0.5 percent at 101 to 200.