Example 02 · Teardown
The same email, marked up.
Demonstration prepared 21 September 2026 · Scipioform
Download the sample teardown PDF
This is Stannary Systems's current cold email, and what we would send back. It is the same document a visitor to our teardown page receives, written out in full. The company and the email are invented; the problems in it are the ones we see most.
The original
Subject: Quick question about your field operations
Hi {{first_name}},
I hope this email finds you well.1 I'm reaching out because I noticed you're the Operations Manager at {{company}}2 and thought you might be interested in what we do3.
Stannary Systems is an all-in-one platform for field service teams4. We help companies like yours with scheduling, mobile forms and reporting dashboards — our customers complete inspections up to 50% faster and save hours of admin every week5.
Would you be open to a quick 30-minute demo this week or next?6 I have Tuesday at 2pm or Thursday at 10am.
Best regards,7
Alex
What is wrong with it
Seven marks. None of them is about grammar — the email is competently written, and that is exactly why it is worth taking apart. The problems are in who it is addressed to, why it is arriving, what it claims and what it asks for.
1
Reason to respond
This is the sentence a reader uses to decide whether the rest is worth reading, and it tells them nothing. It is not offensive; it is evidence that the sender had nothing specific to say and opened with a formality instead. Everything the message needs to earn — that it is about them, now, for a reason — has to happen in this line, and it does not.
2
Audience
Noticing a job title is not research; it is the merge field. It also reveals the targeting: the list is built on a title string, so it contains companies with no statutory inspection obligation at all. Those companies will not reply, and their non-reply will be counted as evidence about the message when it is really evidence about the list. Every rate this campaign produces is uninterpretable until the denominator is fixed.
3
Relevance
The sentence puts the work of finding the relevance onto the reader. A busy operations director in a mobilisation window will not do it. This is the clearest symptom of the underlying problem: the sender does not know why this company, this week.
4
Relevance
Three competitors can write this sentence word for word, so it carries no information. It is also the wrong fight: describing the product as an all-in-one field-service platform invites a feature comparison, and that is the one comparison Stannary loses. The generic tools have longer lists. The thing only Stannary can say — that its evidence export satisfies a named asset owner’s audit — is absent.
5
Unsupported claim
No denominator, no baseline, no source, and "up to", which means the number is the best case and the reader knows it. A director who has evaluated three of these tools has read this number three times this year. It costs credibility and buys no attention. It is also the sentence a procurement reviewer will later ask Stannary to evidence, and it cannot. Either publish the real figure with its denominator and who produced it, or say nothing. Here: nothing, because the figure does not exist.
6
Request size
Thirty minutes is a real cost, and "quick" does not make it smaller — it signals that the sender either does not know that or hopes the reader will not notice. The ask is larger than the trust the message has earned, and it is aimed at the period when this reader has least time. Offering two fixed slots compounds it: it assumes a decision that has not been made yet.
7
Reason to respond
A first name with no company, no role and no way to tell whether this person would be in the conversation or would hand it to someone else. Small, but it is the last thing read before the decision to reply.
The revision
Subject: Your {{framework}} mobilisation — evidence format
Hi {{first_name}},
You picked up {{framework}} with {{asset_owner}} on {{award_date}}. If the usual mobilisation window applies, first inspections are due around {{due_date}}.
The part that catches people out is not the scheduling. It's that {{asset_owner}}'s audit wants the evidence in a specific shape, and most teams find out which shape after the first month's reports come back.
We build Stannary Field, and two of our customers pass {{asset_owner}}'s audit with it today. I've written up the evidence format they require — what has to be captured in the field, and what the monthly submission has to contain.
Want me to send it? It is two pages and it is yours either way. If it shows a gap against how you are set up, I will happily spend fifteen minutes on it with you. If it does not, you have the checklist and you do not need us.
Alex Mercer
Sales, Stannary Systems
The braces are merge fields filled from the award record: the framework name, the asset owner, the award date and the derived due date. Everything in them comes from a published notice, so every sentence can be sourced if the reader asks.
What changed, and what each change is for
What the new ask gives up
A checklist request is a smaller commitment than a demo, so we should expect more people to take it and a smaller proportion of those to become meetings. Some of the extra replies will be people who wanted the document and were never going to buy. That is the cost, and it is real.
We think it is worth paying here for one reason: the checklist qualifies on the way through. Someone who reads it, finds a gap and comes back has told us that they have a gap, which is more than a politely accepted demo tells us. Someone who reads it and does not come back has cost us a document and has not cost Stannary’s salesperson half an hour.
It is worth being clear that this reasoning is why we would test it, not evidence that it works. If test 3 shows the checklist ask produces more replies and no more qualified conversations, the trade failed and we go back to asking for the fifteen minutes directly.
What the revision is testing
A message that names the specific framework award, the asset owner and the mobilisation date will produce more qualified conversations than the same message without those three facts.
Measured by. Qualified conversations, against the written definition, per 100 delivered. Not replies: the revision is expected to change the mix of replies as well as the number, so reply count alone cannot answer this.
It is wrong if. If the trigger variant does not out-perform the no-trigger variant on qualified conversations, the expensive part of this programme — finding, dating and verifying awards — is not earning its cost, and should be stopped rather than scaled.
We are not claiming the revision improves response rates. No test has been run on it. The reporting example shows what the answer would look like and what decision it would force. The reporting example is what came back after six weeks, and the threshold was missed.
Want yours marked up like this?
Send us your current outreach and we will do the same thing to it, in writing, within 5 business days. It stays private unless you tell us otherwise.