Multi-channel outbound sequence: a 3-week plan across email, LinkedIn and phone

Give each channel a job

Email carries the volume and the offer. LinkedIn adds a second place to see your name and a profile that backs up the email. Phone gives the fastest reaction but costs the most time per contact, so it goes to a small slice of accounts. Everything is sequenced against one list and one suppression list. The rules for LinkedIn steps, including what LinkedIn prohibits, are in the LinkedIn outreach playbook.

The case for spreading touches across channels starts with email itself. Later emails return less. All figures here were checked on 18 October 2026.

Smartlead's figures come from more than 850 million emails sent through its platform, covering January to June 2026, and it advises planning for three or four emails (Smartlead benchmarks). Woodpecker's come from campaigns of four to six emails sent from mailboxes with at least 100 emails sent, in the last 90 days of 2025, and its page says every follow-up is less effective than the one before it (Woodpecker). That is a reason to spend touches five onward on a different channel, though it does not prove the other channel will do better.

Be skeptical of claims that multichannel sequences produce two to four times the replies. One vendor guide states the multiple without naming a dataset or method (Smartlead guide). Treat it as unproven and test it on your own list.

A three-week sequence, day by day

Days below are business days, so day 15 is the end of week three. The table is an illustration of our default shape; we have no benchmark showing it is optimal. Tier A means an account with a high fit score and a fresh signal, as defined in the signal-based outbound playbook.

A tier A contact gets ten touches: four emails, a request, two messages and three calls. Tier B gets seven. Tier C gets the four emails alone. Tier B is an account that clears your fit cutoff but lacks either a top fit score or a fresh signal, and tier C is the remaining accounts worth emailing at all.

The order has reasons. The request follows email 1 by a day, so the name is already familiar when the notification arrives. Calls start in week two, after two emails, so the opener can refer to a note the person has had time to see. Email gaps run three, four and five business days, which gives each message room before the next. No contact gets more than one touch a day.

If someone accepts late, shift the LinkedIn messages with the acceptance, but send none after day 15. Resist lengthening it. Later touches return less, and in our experience a better list or offer beats a fifth touch. For planning, our benchmarks are 2 to 3 percent of sends replying, 15 to 50 percent of replies positive and 30 to 50 percent of positive replies booking.

Rules for stopping a sequence

A sequence needs exit rules written down before launch, and they need to work across channels. If a reply on LinkedIn does not stop the email, you will email someone who has already answered you. Rules that rest on a law or regulator are marked as such. The rest are our own.

These rules are easiest to enforce when every channel's replies land in one place and one status field. That routing is the job of our automation service, and the next step after a reply is covered in handle replies and book the call.

The legal rows come from the FTC's CAN-SPAM compliance guide, the ICO's guidance on live marketing calls and LinkedIn's invitation restrictions page, all checked on 18 October 2026. The ICO page notes its guidance is under review. The FTC's Do Not Call Q&A says most calls to a business made to solicit sales are exempt from the Do Not Call provisions (FTC), but state laws, mobile numbers and other countries have their own rules. This is not legal advice, so check the rules for each market you call into.

When a phone step is worth it

Phone is worth it for a small set of accounts where each meeting is worth a lot and you hold a number that reaches the right person. It is rarely worth it for a long list of low-value accounts, where the same hour buys many more emails.

The published numbers are thin and mostly vendor data, so read them as ranges. Cognism's 2025 cold calling report uses data from WHAM (We Have A Meeting): 41,936 connected calls and 27,513 conversations worldwide. It reports a 2024 rate of 4.82 percent of conversations ending in a booked meeting, says the 2025 rate is almost half of that, and puts the industry standard at around 2 percent (Cognism). This is vendor-reported data, and the page does not give total dials in text. Using its range, 100 conversations yield roughly 2 to 5 meetings (our arithmetic).

The same report counts how many attempts it takes to reach someone. Its figures say three calls captured 93 percent of conversations and five calls 98.6 percent, with little return beyond that. That is why the example caps phone at three calls.

Gong's published cold call statistics add two findings about the opening. Calls that opened with "Did I catch you at a bad time?" had a 0.9 percent success rate and cut the chance of booking by 40 percent, and stating the reason for the call raised the success rate 2.1 times. The page does not give the number of calls analyzed, so treat these as a hint (Gong).

Two engagements from our work show the scale email can reach. One for an AI feedback SaaS reached 1,000+ emails a day over multi-inbox infrastructure and booked 40+ meetings a month with AI qualification (case study), and one for an enterprise AI SaaS built $3.7M in pipeline with multi-stakeholder outreach (case study). The infrastructure behind that is on our cold email and outbound systems page.

Measure the sequence against an email-only group

The honest test is a holdout. Within one tier, send part of the list the full multi-channel sequence and part the email steps only, drawn at random from the same list, then compare booked meetings per 100 contacts. Compare meetings and ignore opens.

Plan for small numbers. At the benchmarks above, 500 email-only contacts give roughly 10 to 15 replies, 2 to 8 positive replies and 0 to 4 meetings (our arithmetic). At 5 percent significance and 80 percent power, the standard two-proportion approximation says telling a 0.5 percent meeting rate from a 1.5 percent one takes about 1,550 contacts per group, and telling a 2 percent reply rate from a 3 percent one takes about 3,800. Most tier A lists are far smaller than that, so you will only see very large effects. Decide in advance what lift would justify the extra work, and run the test over several batches. The outbound metrics playbook covers the rest of the scorecard.

Questions

How many touches should a three-week outbound sequence have?

Our example uses four emails for every contact, plus a LinkedIn request and two messages for tiers A and B, and up to three calls for tier A. That is four, seven or ten touches over 15 business days depending on tier.

Does a multichannel sequence beat email alone?

The multiples quoted in vendor guides come without a named dataset or method, so treat them as unproven. Run an email-only holdout on your own list and compare booked meetings.

How many times should I call a prospect?

Up to three. Cognism's 2025 report says three calls captured 93 percent of conversations and five captured 98.6 percent, so extra attempts return little.

Should one reply stop the whole sequence?

Yes. A reply on any channel stops every channel and goes to a person. An email opt-out should suppress the contact everywhere, even though the US legal limit for honoring it is 10 business days.