Six ways outbound fails, and how the numbers tell you which one you have

A campaign that produces nothing produces no information either, which is what makes it expensive twice. You spent the money and you still cannot say what to change. Founders respond by rewriting subject lines, because that is the visible layer.

Outbound fails in six places. Each one has a distinct fingerprint in the numbers, and the numbers tell you which one you have before you touch the copy.

Read the funnel backwards

Four rates, measured per segment rather than across the whole campaign.

Find the first rate that falls below range. That is where the campaign broke. Everything downstream of it is a symptom.

Failure one: the mail never arrived

Delivery under 97 percent, or delivery fine but replies near zero across every segment. The message is landing in Junk, or the addresses were never real.

You cannot see this from inside your own sending tool, which reports delivery as accepted by the receiving server. Accepted and inboxed are different outcomes. Seed a handful of real mailboxes at Google, Microsoft and one European provider, and look at which folder your own campaign lands in.

The tell that separates this from a copy problem: reply rate is uniformly terrible across segments that should behave differently. Real copy problems produce variation.

Failure two: the list is not the market

Delivery good, reply rate 1 to 2 percent, and the replies that do come back say some version of "wrong person" or "we do not do that."

The list was built from a dropdown. Industry equals manufacturing, size equals 50 to 200, country equals Germany. That describes 40,000 companies and characterises none of them. Nothing in that filter set says the company has the problem you solve.

The fix is upstream of the copy. Pick an observable event that only happens to companies with your problem, then build the list from companies where that event occurred in the last 90 days. Fewer names, and the message can reference something true about them.

If your first line would be equally true sent to any company on the list, the list is doing no work. You are paying for volume to compensate for the absence of a thesis.

Failure three: the offer does not survive a stranger

Reply rate healthy at 5 to 7 percent, positive replies under 0.5 percent. People are reading and answering. They are answering no.

This is the most useful failure, because the market is telling you something specific. Read the negative replies as data rather than rejection. Three patterns recur:

They already have a supplier and see no reason to switch. Your message competes with an incumbent and never mentions why anyone would move.

They agree the problem exists and say it is not a priority. You have no answer to why this quarter, which means your targeting signal was not really a trigger.

They do not recognise the problem in your words. You described your solution's category instead of their Tuesday.

Failure four: interest that does not convert to a calendar

Positive replies at 2 percent, meetings at 0.2 percent. People say yes and then vanish.

Look at what you ask for in the reply. "Happy to send over a proposal" and "Would 30 minutes on Thursday work" produce different outcomes from the same interested person. Ask for the smallest next step that still moves the deal, and put a concrete time in the sentence rather than a booking link and a hope.

Response speed matters more here than anywhere else in the funnel. A reply answered within an hour books at roughly twice the rate of the same reply answered the next morning. If nobody owns the inbox between nine and six, this failure is structural rather than a copy problem.

Failure five: meetings with people who cannot buy

Meetings happening, opportunities not. Under 20 percent of meetings turn into anything.

You targeted the person who feels the pain instead of the person who signs. Both matter and they are rarely the same human. An operations manager will take the meeting, agree enthusiastically, and then discover there is no budget line for it.

The fix is not to stop talking to the operations manager. It is to write to them with a different ask, one that ends in an introduction rather than a demo, and to run a parallel sequence to the budget holder with a different message.

Failure six: the campaign stops before it starts working

This one has no fingerprint in the numbers, because there are not enough numbers yet.

A European B2B campaign at 800 sends a month produces roughly 40 replies, 12 positive, and 6 meetings. Half of those slip. Judging a campaign on three weeks and 200 sends is judging noise. The sample is too small to distinguish a bad message from an ordinary run of bad luck.

Set the review point before you start: 1,500 sends per segment or eight weeks, whichever comes first. Then hold to it. Changing the message every ten days guarantees you never accumulate enough of any one version to learn from it.

The diagnostic in one table

Copy is the last thing to change, not the first. In a campaign where the list, the offer and the infrastructure all hold, rewriting the subject line moves reply rate by a fraction of a point. In a campaign where any of the three is broken, no subject line saves it.