Cold Email vs LinkedIn Outreach: Which One Actually Works for You

The argument is usually framed as a preference. It is not. The two channels have different physics — different reach, different ceilings, different failure modes — and which one wins is decided almost entirely by who you sell to.

Here is the comparison without either side's marketing.

The short version#

	Cold email	LinkedIn
Who you can reach	Any business with a published address	Only people with an active profile
Coverage of local trades	High — most have a website	Low — plumbers and salons are barely on it
Coverage of corporate roles	Medium — named addresses are harder	High — this is what it is for
Realistic daily volume	30–50 per inbox, more with more inboxes	~20 connection requests a day, hard capped
Marginal cost per contact	Near zero	Near zero, but time-capped
What breaks when you push it	Domain reputation, recoverable slowly	Account restriction, sometimes permanent
Who owns the channel	You do — it is your domain and your list	LinkedIn does
Attachments, links, detail	Fine	Suppressed and awkward
Paper trail	Searchable, forwardable, exportable	Trapped in the platform
Legal framework	Real and specific per country	Platform terms, plus email law if you export

Two rows do most of the deciding: coverage of your market, and who owns the channel.

Where each channel actually wins#

Cold email wins when the business is the buyer#

If you sell to businesses rather than to roles — clinics, garages, restaurants, letting agents, shops, trades — email is not the better channel, it is the only serious one. A plumbing company with eleven vans has a website, a published address and a person who reads it. The owner may have a LinkedIn profile they last opened in 2019.

It also wins whenever the pitch needs detail: a proposal, a number, a link to something you built, an attachment. LinkedIn messaging is built for conversation and actively poor at documents.

And it wins on ownership. Your list is a file you keep. If LinkedIn changes its limits next quarter — it has, repeatedly — nothing about your email pipeline changes.

LinkedIn wins when the role is the buyer#

If what you need is "the Head of Partnerships at a 200-person software company", email struggles at the first step. That person's address is not published anywhere, the company's is info@, and info@ will not forward your message to them.

LinkedIn solves exactly that: it is the only mainstream source where a job title at a specific company, right now, is a searchable thing. For enterprise and mid-market roles that is decisive.

It also wins on context. The recipient can see who you are, who you both know, and what you post, before deciding whether to answer. On email you are a string of text from an unknown domain.

Where both lose#

Neither channel reaches a business that is genuinely offline — no site, no profile, a phone number on a van. That market exists and it is large. The answer there is the phone, and pretending otherwise wastes weeks.

The volume ceilings, honestly#

This is where the two diverge most and where most plans go wrong.

Cold email is limited by reputation, not by rules. A new domain should stay under about 50 sends a day per inbox for the first few weeks. That ceiling then rises. Adding inboxes and domains raises it further — the practical constraint becomes how many verified, well-matched contacts you can generate, not how many you may send.

LinkedIn is limited by hard caps that no amount of good behaviour lifts. Roughly 100 connection requests a week, throttled further if your acceptance rate is poor. Free accounts also hit a monthly search limit. InMail is paid and metered.

The consequence: LinkedIn's ceiling is somewhere around 400 new conversations a month with real effort, and it does not scale past that on one account. Email's ceiling is set by your list and your infrastructure. If your model needs a thousand conversations a quarter, that decision is already made for you.

Reply rates, and why the comparison is unfair#

LinkedIn connection acceptance rates of 30% and message reply rates of 10–20% are commonly reported. Cold email reply rates of 3–8% are considered good. That looks like a rout — until you notice the denominators are different.

LinkedIn's numbers are computed on a small, hand-picked set of people you searched for individually. Email's are computed on a list of several hundred built in an afternoon. Compare like for like — the same fifty carefully chosen prospects on each channel — and the gap narrows sharply.

The other asymmetry: a LinkedIn reply is often a conversation that goes nowhere, because the medium invites low-commitment politeness. An email reply from a business owner is more likely to be a decision. Measure booked calls, not replies, and the two channels look much more alike than the headline rates suggest.

The risk profile is not symmetrical#

This is the part LinkedIn automation vendors do not put on the pricing page.

Email failure is recoverable. Push too hard, and your domain's reputation degrades. You slow down, clean the list, warm up again, and in a few weeks you are back. If it is truly unrecoverable you move to another sending domain — which is exactly why cold outreach should never go out from the domain your invoices come from. That reasoning, and the DNS records that go with it, are in the deliverability guide.

LinkedIn failure is not. Automation tools that run against your logged-in account violate the user agreement. Enforcement is account restriction, and a restricted account takes your network, your history and your social proof with it. There is no second domain to move to — you would be starting a new profile from zero.

So the sensible posture is asymmetric too: automate email, do LinkedIn by hand.

The sequence that uses both#

For a list where both channels are viable, the ordering that works:

Day 0 — view the profile. No message. Profile views generate a notification, and a small share of people look back at you. Free, invisible, zero risk.

Day 1 — email. The full pitch, four to six lines, one question. Segment-level personalisation is enough at this stage; the shapes are in cold email templates that get replies.

Day 4 — connection request, no note, or a very short one. Requests with no note are accepted more often than requests with a pitch attached. Now your name has appeared twice in a way that does not read as pursuit.

Day 6 — email follow-up in the same thread. Adds something new; never "just bumping this".

Day 10 — LinkedIn message, only if the connection was accepted. Two lines, referencing nothing about the emails. Acceptance is a signal; treat it as a fresh, warmer start.

Day 14 — close the loop by email. The graceful exit, which is where a surprising share of replies actually arrive.

Five touches across two channels in two weeks. That is enough to be memorable and short of being a nuisance — and each touch is cheap because the list, the segment and the message were built once.

Which to start with, if you have to pick one#

Answer three questions:

1. Does your buyer's business have a website with a published address? If yes, email. Local and mid-market B2B is an email market.
2. Do you need a specific role inside a larger organisation? If yes, LinkedIn — for finding the person, at least. The address usually still has to come from somewhere else, and prospecting on LinkedIn without Sales Navigator covers that gap.
3. How many conversations do you need per month? Under fifty, LinkedIn can carry it. Above a couple of hundred, only email can.

For most people reading this — agencies, small software companies, local service businesses — the answer is email first, LinkedIn as the research layer and the second touch. The exception is a genuinely enterprise sale, where the named individual is the entire problem and volume is irrelevant.

What to measure, either way#

The same four numbers on both channels, so they are comparable:

• Delivered / accepted rate. Email: sends minus bounces. LinkedIn: connection acceptance. Both tell you whether your targeting is plausible before any copy matters.
• Reply rate, positive and negative separately. "Not for us" is information; silence is not.
• Conversations booked. The first number that means anything commercially.
• Cost per conversation, in time. LinkedIn's cost is nearly all your hours; email's is nearly all list quality. That difference is usually what decides which channel your business can actually sustain.

Track those for two weeks per channel on the same fifty prospects, and you will have a better answer for your market than any general comparison — including this one.

FAQ#

Is cold email or LinkedIn better for B2B?#

For selling to businesses with a public presence — local services, clinics, trades, retail, small agencies — email, because coverage is far higher and the volume ceiling is much higher. For reaching a specific role inside a larger company, LinkedIn, because job titles are searchable there and nowhere else.

How many LinkedIn connection requests can I send per day?#

Roughly 100 a week, which works out at around 20 a day, and less if your acceptance rate is low. The limits are not published precisely and are adjusted by LinkedIn without notice. Exceeding them results in restrictions rather than a warning.

Can I automate LinkedIn outreach safely?#

Automation against your logged-in account breaks the user agreement, and enforcement is account restriction, which is often permanent. Given the asymmetry — a burned email domain is replaceable, a burned profile is not — the sensible split is to automate email and do LinkedIn manually.

Do I need both channels?#

No, but sequencing them raises response rates without much extra work, because the list and the message are built once. Start with the one that matches your market, add the second only when the first is running consistently.

What reply rate should I expect from each?#

For a targeted, verified email list, 3–8% replies is a realistic band. LinkedIn typically shows higher percentages on much smaller, hand-picked sets. Compare booked conversations rather than reply rates, because the denominators are not equivalent.