In a cold outbound campaign, contact 2 or 3 people at each target company in the first wave. At a small founder-led company, 1 or 2 is enough. At a large account where one contract could carry your quarter, go to 3 or 4.
Two or three. Not ten.
Do not send to all of them on the same day. Start with the person most likely to own the problem, bring in the second a few days later, and keep the third in reserve until the first sequence has run its course.
That is the starting number, before anyone has answered. It is much smaller than the 10 to 17 people you see quoted in deal research, and the difference matters. Those bigger numbers count the people involved in a deal by the time it is won, and a deal only exists once somebody at the company has agreed to talk.
What the deal data really measures
Contacting more than one person at the same company is what sales teams call multithreading. The case for it is strong.
Gong studied 1.8 million deals for new business that closed during 2024. In that set, 77% of deals were multithreaded, and the deals that were won had twice as many buyer contacts as the ones that were lost.
In deals over $50K, multithreading boosted win rates by an average of 130%. Large strategic wins averaged 17 contacts. Gong also writes that a won deal in the $50K to $250K range typically involves "at least 10 stakeholders".
Keep in mind who produced these numbers. Gong sells revenue software, and both pages come from its own research, so they count as one source.
The bigger point: every one of those figures describes a deal, which means a sales opportunity already exists. Your cold list has no deal yet. Nobody has replied, nobody has agreed to a call, and nobody inside the company is vouching for you.
So the research makes a strong case against one contact per account. It does not tell you to email ten strangers at the same company on Monday morning.
How many to contact, by account size
The first wave exists to find one person who will talk to you. Everyone after that is easier to reach through them.
The numbers below are planning guidance for a first wave, worked out from that logic. They are not results logged from one campaign.
| Account | First wave, before any reply | Held in reserve | When the reserve goes in |
|---|---|---|---|
| Small founder-led company, under about 50 people | 1 or 2 | 1 | Only if the first sequence ends in silence |
| Mid-size company, about 50 to 500 people | 2 or 3 | 1 or 2 | When a sequence finishes with no reply, or someone answers "not my area" |
| Large account, or a deal worth a quarter of your revenue | 3 or 4, across two teams or two locations | 2 or more | Preferably through the person who replied, not as more cold sends |
At a 20-person company, three emails from the same stranger in one week can land on three desks in the same room. At a 5,000-person company, three people in different offices are far less likely to compare inboxes.
Space the first wave out
Cold email sequences usually run four steps. In our campaigns the meeting tends to come from the third follow-up, and only if the messages are spaced out with real gaps.
Adding a second or third person does not change that spacing. I lay out the full logic in the follow-up sequence my team builds.
Here is one way to spread it across a single account, as an example calendar:
- Day 1. Person 1 gets step 1 of the sequence.
- Around day 4 or 5. Person 1 gets step 2. Person 2 gets step 1, with a different angle.
- The following week. Both sequences keep running on their own spacing. Nobody at the account gets two emails from you on the same day.
- After person 1's fourth step. If the account is still quiet, person 3 comes out of reserve.
- The moment anyone replies. Every other sequence at that company stops. From here, new names come from the conversation.
There is a sending reason for this as well as a human one. Google's sender guidelines ask senders to raise volume gradually and avoid sudden spikes, keeping the spam rate below 0.1% and never reaching 0.3% (checked September 2026).
Every extra person you add is one more inbox that can press "report spam". Tripling the names per company overnight also triples the sends from the same mailboxes.
Adding colleagues also means more addresses you had to find or guess. Every address gets verified before it goes into a campaign, and the bounce rate has to stay below 0.5%. A guessed colleague address counts as unverified until it has been checked.
Who goes second and third
Pick the next names by the part of the problem each person owns. One person uses what you sell every day. One signs off the budget. One has to live with the rollout once it lands.
Each of them gets their own email, about their own part of the problem. Same offer, new angle.
Here is an illustration, not a real campaign. Say you sell software that takes the manual work out of month-end invoicing at consulting firms. The first line to each person could read:
- To the person who does the invoicing: "How much of month-end still happens by hand in a spreadsheet?"
- To the finance director who owns the budget: "Invoices that go out sooner get paid sooner, and that shows up in your cash every month."
- To the operations lead who would run the switch: "Changing invoicing tools usually stalls at the data import, so here is how that part works."
And watch the replies. Plenty of them are neither a yes nor a no. Some point you to a colleague, which can become your next name. Answer while the thread is warm.
Somebody has to own that inbox, which I explain in my post on the first three months of a campaign.
Keep your own team off the same prospect
This part sits on your side of the table.
The moment you contact three people per company, the chance that two of your own people land on the same account goes up. One rep on LinkedIn, another by cold email, an agency on top, the founder sending "a quick note" from his own inbox. The prospect sees four people from one company who clearly have not talked to each other.
I wrote the rule into my post on sending setup:
"Split the territories too, so two of your people are never sitting in the same prospect's inbox in the same week. That comes out of the list, not out of the tool."
Run this check before the first wave goes out:
- Assign whole companies. Once an account belongs to a rep, every person at that company goes through that rep.
- One owner per channel mix. LinkedIn and email to the same person should come from the same sender, planned as one sequence.
- Check the reps against each other. Ask your two most active reps to list the last companies they contacted. Any company on both lists is a territory problem.
- Include outside senders. If an agency, a partner or the founder is also emailing, their accounts go into the same split.
- Stop everything at the account on the first reply. Otherwise a colleague of the person who just said yes gets your step 3 that afternoon.
Dividing territories company by company, so nobody reaches the same person twice, is one of the list checks I describe in what goes wrong in a B2B lead list.
What depth costs you in breadth
Contacts per company is a trade-off, because in any given month your sending capacity is fixed.
For context, Smirnov Consulting Group is a Prague-based B2B outbound lead generation agency that runs cold email and LinkedIn campaigns for founder-led B2B companies and books qualified sales calls. Here is one month of that work, for a German software development client in May 2023.
On the cold email side alone, the campaign reached 5,000 people. It took 15,000 emails from 20 addresses, which works out at about 25 a day for each mailbox. The LinkedIn side of that campaign ran separately.
I walk through that volume math, and how one team split its profiles and mailboxes between reps, in my post on domains, inboxes and sending volume.
Now run a planning calculation on that capacity. At 1 person per company, 5,000 people means 5,000 companies. At 3 per company, the same 5,000 people cover only about 1,670 companies.
You have bought depth with breadth. In a small market that is the right trade: several decision-makers at each company, longer sequences and more patience, as I explain in the playbook for a market that is running short.
In a large market with a long list, the same capacity at 1 or 2 per company opens far more new doors.
LinkedIn tightens the same equation. We plan around roughly 2,800 people a month per LinkedIn profile, which is why a real team adds profiles and mailboxes instead of pushing one harder.
For one financial advisory client in Australia in 2023, the build was 5 profiles and 20 email accounts. Each rep got a LinkedIn profile, one mailbox reserved for follow-ups, plus three cold mailboxes.
Mistakes that burn an account
Everyone on day one. Three near-identical emails land in one building on the same morning. To the people reading them, it looks like a campaign aimed at the whole company. Because it is one.
One message for every title. Your budget owner and your daily user care about different things. One generic email written for both of them lands with neither.
The same text, word for word, to every contact. No variants. No spintax. That sits on my list of what breaks a sending setup, and colleagues notice faster than you think.
Adding names instead of fixing the list. Nobody replied at 3 per company? Do not reach for a fourth. Sort the list by title. Count the people who could approve a deal on their own. If that is a minority, a fourth name from the same list is one more wrong person.
FAQ
How many people should I contact at one company when cold emailing?
Two or three in the first wave for most B2B accounts, 1 or 2 at small companies and 3 or 4 at large ones. Keep one or two more in reserve for when the first sequence ends without a reply.
Does contacting multiple people at the same company hurt deliverability?
It can if you add them all at once. More people means more sends and more chances of a spam report, and Google asks senders to avoid sudden volume spikes. Stagger the sends and verify every address before it goes in.
What if I don't know who else to contact at the account?
Start with the one person you are sure about. If they reply, ask who else should be part of the conversation. A name handed to you from inside the company arrives with context that a cold name lacks.
Should I use the same message for everyone at the account?
No. Keep the offer the same and change the angle to fit each person's part of the problem. Three copies of one email in one building read as a mail merge.
