Omnichannel Outreach for B2B: Order, Handoffs, Stops

Omnichannel outreach is one conversation with one prospect, carried across LinkedIn, email and sometimes a phone call. Each channel knows what the others already did. A reply on LinkedIn stops the email follow-ups. An accepted invite changes what the next email says. An unsubscribe closes the thread on every channel at once, while a bounce only ends email to that address.

For a B2B team, running it well comes down to four decisions:

1. Which channel makes the first touch.
2. How many touches the sequence gets, and how far apart they sit.
3. What happens the moment a prospect reacts, on any channel.
4. Whether a warm call belongs in the mix, and under which rules.

My answer to the first decision is fixed. LinkedIn opens, and email follows as a warm follow-up, so the prospect has already seen my name by the time it arrives. The tables below cover the other three decisions. Three client programs of ours, for a US marketing agency, a Canadian financial advisory firm and a UK social media agency, show the volumes and how long the ramp takes.

How is omnichannel different from multichannel outreach?

Omnichannel means the prospect deals with one sender who has one memory. You may reach them on three channels, but every message is written knowing what happened on the other two.

Multichannel is the weaker version. The channels run side by side. The LinkedIn tool sends its steps, the email tool sends its steps, and neither knows the prospect answered yesterday. That is how a founder who replied "send me details" on LinkedIn gets a fourth cold email asking whether they saw the first three.

Why add a second channel at all? SmartReach, a company that sells outreach software, looked at more than 1,800 LinkedIn campaigns run by over 160 businesses between January and June 2026.

LinkedIn inside a multichannel sequence got a 1.6x reply lift against LinkedIn alone. Per 1,000 prospects, the report counted about 4 meetings from LinkedIn alone, about 14 with email added and about 21 with email and calls (SmartReach).

Never take a vendor's own numbers at face value. Read them as a direction of travel, because teams that run several channels may also differ in other ways. Then test it small yourself before you trust it, and grade any multichannel outreach statistics you quote before you rely on them.

Should you start with LinkedIn or email?

Start with LinkedIn when the prospect has never heard of you, which in cold outbound is nearly always. A connection request costs them nothing to look at, and the profile is what they can check before they accept. Buyers audit you before they buy, and the LinkedIn touch gives them a face and a profile to check before the warm email lands.

Not every guide agrees, and the disagreement is fair. The lemlist sequence template starts with email because email has the "highest volume capacity", up to 50 sends a day per inbox in its guidance. If the first emails get no opens within 3 days, it moves the prospect to a LinkedIn invite. If the invite is not accepted in 7 days, it goes back to email (lemlist).

That logic suits a team chasing reach.

It has one weak joint: it branches on opens.

In its sender guidance, Google states plainly that it doesn't track open rates and adds that a low open rate is not necessarily an accurate sign of a spam or deliverability problem (Google Workspace Admin Help). A rule that waits for an open can push someone who read every word onto another channel as if they had ignored you.

In Czechia, Germany and Austria my team relies on LinkedIn and sends no cold email unless a compliant basis exists.

The UK rows come from the regulator. According to the ICO, PECR (the UK's Privacy and Electronic Communications Regulations) lets you email corporate subscribers without consent, as long as you state clearly who is writing and include a working address where the recipient can opt out.

Sole traders and some partnerships are treated as individual subscribers, which means you need their specific consent or must meet the soft opt-in conditions before emailing (ICO).

The UK GDPR still applies to the personal data, and any business contact can object to direct marketing use of it, with no exceptions. This is not legal advice.

Whichever channel opens, the goal is a conversation. Conversations come first. Calls come out of conversations. Clients come out of calls. Set the monthly target at 50 or more conversations started rather than a count of booked calls, because a team chasing calls pitches in message one. If you already see engagement before the first message, you can use it to warm the contact up on LinkedIn without pitching.

How many touches should an omnichannel sequence have?

Five to six touches across both channels is the minimum I plan with. Two messages and then silence is not a strategy. It's a lottery ticket.

In our campaigns, the third follow-up, with real time left between the messages, is usually the one that works best. Most teams polish the first message and never get around to sending the third.

SmartReach's report makes a related point about patience: in its data, the first follow-up added almost nothing, while the second added about 4 percentage points.

For spacing, the lemlist guidance gives 3 or more days between early messages and 7 to 14 days for later ones, to keep spam risk down.

If the invite is never accepted, the LinkedIn steps drop out and the email steps carry on alone.

Every step asks for the next small step and never for the service itself. Pitching in the opening message is the fastest way I know to be ignored.

Here is a made-up example for a supplier writing to a contracts manager at a construction company. Touch 2, on LinkedIn:

Thanks for connecting, Sarah. Quick question: when your team prices a new residential package, who signs off on edge protection, you or the site manager?

Touch 3, the warm email:

Subject: Edge protection on your Leeds sites

Hi Sarah, we connected on LinkedIn last week. I saw your team has two residential sites running in Leeds this year. We supply edge protection to contractors on blocks of a similar height across Yorkshire, and we can quote from a drawing. Would a 15-minute call next Tuesday help, before your next package goes out?

Increasing the volume is the last step, not the first. A sequence that cannot win replies at a small scale will not win them at a large one. Doubling the sends on a sequence that has no stop rules only doubles the mistakes.

What should happen when a lead replies on LinkedIn or email?

Write the stop rules before launch, and apply each one to the person, on every channel.

In lemlist, the automation settings let you choose Continue, Pause lead or End lead when a person replies or clicks a link. Pause holds every later follow-up until someone restarts it by hand, and End removes the person from that campaign flow (lemlist).

Its status guide counts a reply on either channel as Replied and separates Failed (the tool could not attempt the send) from Bounced (the email went out and the receiving server rejected it). A failed send is a setup problem; a bounce is a list problem.

One detail to test before launch: the same guide files an accepted LinkedIn invitation under Clicked in multichannel campaigns (lemlist). If your click rule ends leads, check whether an accept triggers it.

An accept is permission to send the next step, nothing more. In the SmartReach data, 4 in 5 people who accepted a connection request went on to write nothing at all.

Tools cannot read a two-word answer and decide whether it means curious, annoyed or out of office. On every build I want a human on replies, an inbox manager, so a "maybe" never sits unread. Every conversation goes into a CRM as well, and I do not care which CRM, only that every exchange is logged in it.

This handoff is the strongest argument for one team owning both LinkedIn and email, and for how fast a yes should turn into a booked call.

When can you add a warm call to cold email and LinkedIn?

Nearly all of my team's outreach is cold email and LinkedIn; cold calls are not part of it.

If a client wants a phone step, we call only people on a warm list, meaning they replied, engaged or opted in first, and we treat it as an optional add-on. The SmartReach meeting figure that includes calls is the vendor's own data, and my team has no phone results of its own to set against it.

A warm call earns its place when three things are true:

1. The prospect has already replied, engaged or opted in on email or LinkedIn.
2. You obtained the number properly and screened it against the registers below.
3. The call has one job, such as confirming a meeting or answering a question the prospect raised.

The UK regulator leaves little room here. Without the person's specific consent, a number listed on the TPS or the CTPS (the Telephone Preference Service and its corporate version) is off limits for marketing calls.

Screen every list against both registers plus your own "do not call" list, and never call someone who has said they do not want your calls.

Automated marketing calls need consent that specifically covers automated calls. On a live call, let your number display, say who is calling, and hand over a contact address or a freephone number when the person asks for one (ICO).

In the US, the Telemarketing Sales Rule covers a campaign to sell by telephone that involves more than one interstate call. It bars calls to numbers on the National Do Not Call Registry unless the seller has the person's express written agreement or an established business relationship. Sellers must also keep records of their campaigns (eCFR, 16 CFR Part 310).

Section 310.6(b)(7) exempts most calls between a telemarketer and a business, but the exemption has carve-outs (retail sales of nondurable office or cleaning supplies among them), so read it against your own case before relying on it. This is not legal advice.

What jobs does the tool stack have to cover?

Buy tools for jobs, and give every job an owner. A tool nobody owns is a tool nobody checks. Take the suppression list. If nobody owns it, the stack keeps emailing people who already said stop.

Before anything sends, my team verifies every address and keeps bounces under 0.5%. SPF, DKIM and DMARC go on every sending domain before volume, whatever the daily count.

Gmail sets the floor for the email side. Since February 1, 2024, anyone sending to Gmail accounts needs SPF or DKIM, a TLS connection and a spam rate in Postmaster Tools below 0.3%.

Cross 5,000 messages a day and the bar rises: all three records (SPF, DKIM and DMARC), a From domain that aligns with the SPF or DKIM domain, and one-click unsubscribe on marketing messages (Google Workspace Admin Help).

What do real campaign numbers look like?

The client was a US marketing agency, and for 60 days we ran LinkedIn outbound with 3 sequence scenarios alongside email. The volume was 1,735 LinkedIn InMails and 3,143 emails. That produced 962 replies, and from those came 277 sales calls, then 34 contracts and $255,000+ in sales.

Three things to know before you compare your numbers with these:

1. Every one of the 277 was a meeting booked through LinkedIn or email messages.
2. Booking a meeting and closing a deal are separate jobs. The case counts both, 277 meetings and 34 contracts, and you should judge your own program on both numbers.
3. Your numbers will not match these. The result depends on an offer the market actually wants and on you answering the leads quickly.

With 962 replies in 60 days, one record and one owner per prospect is the only way nobody gets a fourth email after saying yes.

Scenarios are how a sequence stays coordinated. When I build them, prospects split by what they did: someone who accepted and went quiet does not get the same message as someone who answered "not this quarter", though both are offered the same thing.

A Canadian financial advisory firm shows the ramp over a longer period. Its program mixed LinkedIn organic, outbound and warm emails, with the advisors' profiles optimized. Over 12 months it produced 421 sales calls, all of them meetings booked through LinkedIn and email. The first quarter was a ramp: 19 calls in month 1, then 26, then 32.

A UK social media marketing agency followed the same curve. Its program combined LinkedIn outbound and warm emails with cold email across 3 niches, with no paid ads. Month 1: 44 calls. Month 2: 57. Month 3: 68. Over 12 months it reached 714 sales calls, every one booked through messages. Judge any omnichannel program over months, not its first weeks.

What should you check before launching omnichannel outreach?

Run this list before the first touch goes out:

1. Each LinkedIn account and each rep has its own slice of the market, and no prospect appears in two slices.
2. Every sending profile reads like a page for buyers (who you help, how, and proof) rather than a CV.
3. The list is verified and deduplicated across channels, with one record per person.
4. One person owns replies and one person owns the CRM, by name.
5. The stop rules from the table above are set in the tool and tested on a dummy lead on both channels.
6. Every tool reads from the same suppression list.
7. Every sending domain passes authentication and meets the Gmail requirements.
8. A written re-entry rule exists: finished prospects stay out unless you deliberately reactivate them with a new offer or a new reason, a fresh sequence and a CRM note saying why.

If you test only one thing before launch, test item 5. Reply to your own campaign from a test LinkedIn account and watch whether the next email still goes out. If it does, fix that before you add a single prospect. Order first, stop rules second, volume last.

Not sure which channel should open for your market? Book a call and tell me who you sell to, and I will help you set the order and the stop rules.