The Follow-Up Needs Its Own Angle

Why the second email gets the same no

Writing again after an unanswered email asks the same person for the same thing. They decided the first time, and repeating the request does not change that decision. This is the most common and the most expensive mistake in outbound, and it hides behind a polite surface.

In practice it reads like this. The second message opens with a phrase such as “just circling back” and then restates the same offer as the first one. Formally it is a new email. In substance it is a second attempt with unchanged content, and that is exactly how it lands.

Our position: silence is an answer, not a gap. Treating it as a gap sends teams off to test subject lines and send times against a problem that sits one level above both. The real question is not how often you write, but whether the second message has a reason of its own.

The result of that shift is undramatic and effective. Every message in the sequence has to justify why it exists. One that cannot is cut, not rescheduled.

The angle map, decided before anything is written

Most sequences are invented while being written. Someone opens an empty document, drafts the first email, and the follow-ups are assembled afterwards from whatever is left over. That is why they read like leftovers. We reverse the order and settle the substance before any prose exists.

The instrument is a single table. For each touch it records which sub-problem it addresses, with which argument or proof, which personalised detail it uses, and how much commitment its ask carries. Copy is written only once that table stands and has been signed off.

Our position: this map is a gate, not a planning document. It is signed off before a line of copy exists, because once something is written nobody wants to part with a good paragraph. The order of the problems and the choice of proof are substantive decisions, not questions of wording.

The effect shows up immediately. If the sub-problem column says the same thing twice, you can see it in the table. In finished copy you cannot, because both passages are phrased differently while saying the same thing.

The commitment ladder ends with a door

Many sequences get more insistent towards the end. The last message raises the pressure because it is the last one. That costs more than it earns: someone who feels pushed does not reply with a no, they mark the message as unwanted. That hits the deliverability of every campaign that follows.

We build the ladder the other way round. The first message asks for very little — a reaction, an opinion, one line back. The middle ones get more specific and propose a time. The last one closes the door cleanly and explicitly offers to stop sending.

Our position: a clean opt-out is not weakness, it is the price of still reaching inboxes a year from now. A no is a usable outcome. A spam complaint is damage that the next list inherits.

Every personalised detail needs a named field

The most expensive visible error in a sequence is an empty bracket in the inbox of someone you wanted to win. It always happens the same way. A personal touch reads well while drafting, and nobody checks whether the underlying detail actually exists for every recipient.

So each personalised detail is traced back to a named source before it is allowed into the copy. Either it sits as a field on the enriched list, or it is entered per recipient by hand. If neither is true, the detail is cut. We build that list in Clay, where the company and people data come together.

Our position: personalisation without a data source is not personalisation, it is a risk that feels good. The test is simple and happens before sending: for how many rows of the list is this field filled? Anything short of all of them needs a fallback or goes.

That groundwork sits in the list, not in the copy. How we check a list before sending is covered in List Hygiene Before the Outbound Send, and where the fields come from in Lead Enrichment Before the First Email.

A claim that holds for every recipient

A sequence rarely goes to exactly one audience. As soon as two segments share a sequence, every fixed claim inside it has to hold for both. Whatever is true of only one is read by the other half as a statement about themselves — and it is wrong.

In practice that means checking every sentence that asserts a fact against the recipient list. A statement about company size, the system in use, or the role in the buying process either becomes a variable, or it is worded broadly enough to hold for everyone who receives it.

Our position: a claim that is true for eighty per cent of recipients is a false statement to the other twenty. In a sequence going to a thousand people that is not a rounding error, it is two hundred conversations starting with a misunderstanding about the person you are writing to.

The proof belongs to the client, not to the agency

Proof only works while it can be checked. The temptation while drafting runs the other way. A round number reads better than an awkward one, and a sentence about impact is written faster than anyone can evidence it. That is exactly where claims come from that fall apart on the first call.

So we keep two things apart. How a product is different is the client's claim, taken in their own words. What we infer from it is marked as an assumption and put in front of them, rather than landing in the copy as fact. And a number without a named source stays a visible placeholder until it has one.

Our position: the one piece of proof a competitor cannot copy is your own project, with a name and a link on it. That is why our sequences carry a named customer example rather than an industry statistic that sits on every second website.

LinkedIn belongs in the same map

The recipient does not experience two channels. They see one company writing to them. When the email sequence and the network messages are planned separately, the same person gets two unconnected approaches — and both read as more generic than either would on its own.

So the LinkedIn touches sit in the same table as the emails, held to the same standard of having their own angle. The message there is shorter and more personal, but it does not repeat the email from two days ago. That this motion carries deals of its own is shown by the case study linked above: 400k Deal closed via LinkedIn GTM in 4 Months.

Our position: the channel is a property of the touch, not a campaign of its own. Planning them separately doubles the contact frequency and halves the relevance.

Before the sequence enters the sending tool

Between a finished map and the send sits a step many teams skip: settling how the client's house actually speaks. How formal is the address, what salutation is used, and how does the company name read in the middle of a sentence?

We record those answers once per client and read them before each campaign instead of guessing again every time. Only then does the signed-off map move into the sending tool — for us usually Lemlist, where email and network touches run as one sequence.

Our position: an inconsistent salutation is the clearest sign of a mass send a message can carry, clearer than anything in the wording. It is also the cheapest to fix, because the decision is made once and only read thereafter.

To place the two tool names above: Clay lists us in the Clay Solutions Partner Directory as Artisan, the entry tier of that programme.

Lemlist we use as a sending tool in our own stack.