The founder led sales playbook, written while you still sell
What a founder knows but has never written down is what fails at the first sales hire. What to record, what to leave out, and the test that it worked.
Read
Every vendor publishes a number. Almost none of them publish the win rate underneath it, which is the part that decides whether more touches help or hurt.
Ask how many touches it takes to book a B2B meeting and you will be given a number between five and twelve, confidently, by someone with a product to sell. The number is not fabricated exactly. It is an average drawn from that vendor's customer base, which is a population selected by having bought outbound software, and it tells you very little about your own situation.
The useful version of the question is different. Not how many touches, but what each additional touch is doing, and at what point it starts subtracting.
Attribution in outbound is crude. A reply arrives after the fourth message, and the sequence gets the credit, or the fourth message does. Neither is quite right.
What usually happened is that the first message established that you exist and are not obviously irrelevant. The second and third did nothing visible. The fourth arrived at a moment when the problem you named had become live. The fourth message did not persuade anybody. It was present when persuasion became possible.
This matters because it changes what you optimise. If touch four converts, the instinct is to write better fourth messages. The actual lever is relevance in message one, because that is what makes messages two through four legible rather than noise.
Before counting touches, know what the rules require, because the rules are more specific than most sequences account for.
For commercial email reaching US recipients, the FTC's CAN-SPAM guidance is unambiguous on several points people get wrong. Header information must be accurate and identify the sender. The subject line must accurately reflect the content. The message must include a valid physical postal address and a clear explanation of how to opt out. Critically, you must honour an opt-out request within 10 business days, and the opt-out mechanism has to keep working for at least 30 days after you send.
Two details are worth pulling out. The first is that penalties run to $53,088 for each separate email in violation, which is per email rather than per campaign. The second is that hiring someone else to send does not move the liability: the guidance states plainly that you cannot contract away your legal responsibility to comply. If a provider runs your sequences, their compliance failure is your penalty.
None of this tells you how many touches to send. It tells you that the tenth message to someone who asked you to stop after the second is not a persistence strategy. It is an enforcement risk with a number attached.
The other channel has no equivalent regulator, so the platform enforces its own limits, and it does so on a metric most sequence designs ignore.
LinkedIn's help documentation states that restrictions can follow when you send many invitations in a short amount of time, and, separately, when many of your invitations have been ignored, left pending, or marked as spam. That second trigger is the important one. It is not a volume limit. It is a relevance limit, measured by how the recipients respond, and volume only matters because it produces more ignored invitations.
The platform is also explicit that automation is not permitted. Where an excessive number of invitations suggests an automation tool, LinkedIn states it may suspend or restrict the account, and that repeated suspensions may result in permanent restriction. Most standard restrictions are lifted automatically within one week, which is survivable once and expensive as a pattern.
Put the two channels side by side and the same rule appears twice from different directions. Email regulation punishes contacting people who have said no. Platform policy punishes contacting people who do not respond at all. Neither punishes the number of touches as such.
What survives contact with both sets of rules is a sequence built on distinct reasons to make contact rather than repeated attempts at the same one.
The version we use as a starting point is set out in our six touch outbound structure, and the rule it carries matters more than the six: every touch has to say something the previous one did not. A follow-up whose content is "following up on my previous email" is not a touch. It is a reminder that you have nothing further to say, and it is the single most common thing in outbound sequences.
Practical consequences of that rule:
Spacing follows the reason, not the calendar. If touch three exists because a relevant piece of news broke, it goes when the news breaks. If it exists because your tool defaults to three days, it should not go at all.
Channels are not additive. Sending the same message by email and LinkedIn is one message delivered twice, and the second delivery is the one that gets marked as spam. Different channels are useful when they carry different content.
Stopping is a feature. A sequence that ends at six and returns the contact to a nurture list preserves the option of contacting them in nine months when circumstances have changed. A sequence that runs to fourteen removes that option permanently.
If you want your own touch count rather than a borrowed one, it is computable, and the computation is not difficult.
Take your last two hundred sequenced contacts. For each one that produced a meeting, record which touch preceded the reply. You will get a distribution rather than a number, and the shape of it is the answer. In most accounts we have looked at, the distribution has a long tail that thins out sharply, and the honest reading is that touches past the point where the tail thins are being paid for by the ones before them.
Then do the part almost nobody does. Take the meetings and follow them to closed revenue, split by which touch produced them. If late-touch meetings close at a materially lower rate, and they often do, your sequence length is not producing pipeline. It is producing calendar entries, which is a different product.
The measurement discipline for this sits in CRM hygiene, which is unglamorous and load-bearing. Without it the distribution above cannot be built at all.
Touch count is usually a proxy for a decision that has not been made: whether you are running volume outreach or targeted outreach. They have different economics and mixing them produces the worst of both.
Volume outreach accepts a low reply rate and compensates with quantity, which means every additional touch is cheap and the limits above become the binding constraint. Targeted outreach accepts a much smaller list and invests in relevance, which means touches are expensive and the constraint is your own research time. The comparison in cold email versus LinkedIn outreach is really this decision wearing a channel costume, and getting meetings with enterprise clients is almost entirely the targeted case.
We have deliberately not published a recommended number, and the omission is the point. Any number we gave would be an average over our own client base, which is concentrated in MENA B2B services, and it would be quoted back to us in situations it does not describe.
The regulatory position above also covers two jurisdictions and one platform. If you are sending into the EU or the UK, the consent and legitimate interest analysis is different and stricter than CAN-SPAM, and it is worth actual legal advice rather than a blog paragraph. Nothing here is that advice.