What works

    Good Outreach

    The archive only shows failures because failures get submitted. This section corrects the imbalance - annotating high-scoring messages and the specific signals that earned them.

    Important caveat: These are real messages that scored B or above. The bar is not "no flaws" - it's "meaningfully better than the baseline." The sender is still redacted. What survives is the structure.
    The four things every good message has
    01
    A real signal
    Something specific, verifiable, and recent. A post, a hire, a launch. Proves someone looked.
    02
    A named ICP
    Industry, stage, role. Specific enough to exclude most companies. Addressed to this person.
    03
    A customer outcome
    One real customer. One real number. Not a feeling - a claim the reader can evaluate.
    04
    A proportionate ask
    Earned by what came before it. Sized to the value established, not to the seller's goal.
    // Case #0035 · Highest score on file

    Decent Effort

    LinkedIn InMail · Fintech · 2026-02-05
    Case #0035LinkedIn InMailFintech
    Decent Effort

    A message that opens with a verifiable signal, names a specific ICP, leads with a real customer outcome, and asks for something proportionate to what it established. One minor offense - a premature calendar ask - prevents an A grade.

    REAL-SIGNALNAMED-ICPCUSTOMER-OUTCOMEPREMATURE-ASK (minor)
    B6.9/10
    High score
    Annotated message
    The message
    Hi Sarah,Noticed you mentioned the SEC ruling on private credit in your last post - that's exactly the compliance gap we built our reconciliation layer around.
    We work with Series B credit funds that are navigating the new reporting requirements. Kayne Anderson cut their monthly close from 11 days to 3 after implementing ours.
    Worth a look? Happy to send the one-pager first if that's easier.
    What works
    Real signal: References a specific post on a specific topic. Verifiable. Proves the sender looked at her actual content, not just her title.
    Named ICP: "Series B credit funds navigating new reporting requirements" - a description that excludes 99% of companies.
    Customer outcome: Kayne Anderson, 11 days to 3. Real company, real metric, real timeframe. The reader can evaluate this claim.
    Minor issue: "Worth a look?" is a soft ask but still unprompted. The one-pager offer mitigates it.
    // Case #0037

    Almost There

    LinkedIn InMail · MarTech · 2026-02-15
    Case #0037LinkedIn InMailMarTech
    Almost There

    Opens with strong specific research, names a precise ICP. Falls short on the value prop - the outcome claim exists but lacks a number.

    REAL-SIGNALNAMED-ICPWEAK-OUTCOMEPREMATURE-ASK (minor)
    C5.2/10
    Above avg
    Annotated message
    The message
    Hi Marcus,
    Saw your post on attribution modeling for B2B SaaS - specifically the part about first-touch vs multi-touch being a false binary. We've been building exactly around that thesis.
    We work with growth teams at Series A-B SaaS companies trying to get accurate pipeline attribution without a data team. Our customers have seen meaningful improvements in forecast accuracy.
    Open to a 20-min call this week if the timing is right?
    What works / what doesn't
    Strong open: References a specific argument from a specific post, not just the topic. Shows comprehension.
    Solid ICP: "Series A-B SaaS, no data team" is specific and exclusionary.
    Weak outcome: "Meaningful improvements" - how meaningful? A number would have earned an A.
    Ask slightly early: The value prop didn't fully land. "If the timing is right" softens it but doesn't fix it.
    The difference one number makes
    Without a numberC · 5.2
    "Our customers have seen meaningful improvements in forecast accuracy."
    C · 5.2
    Vague. "Meaningful" is unmeasurable. Reader cannot evaluate the claim.
    With a numberB+ · 7.8
    "Lattice went from 40% forecast accuracy to 74% in one quarter."
    B+ · 7.8
    Specific. Falsifiable. Reader can assess whether this outcome matters to them.

    What makes a good cold email

    Four things. A specific, verifiable reference to the recipient. An ICP narrow enough that most companies don't qualify. A real customer outcome with a real number. And an ask that matches the value shown, not the sale being chased.

    The gap between a D and a B is almost never the product. It's the structure. Same product, same recipient, but with actual research and a right-sized ask, the score jumps 4-5 points.

    The best cold email we've seen scored a B (6.9/10). Nobody has hit an A. Most common sticking point: even strong messages include a calendar request that's bigger than what the message earned.

    Messages that name a specific customer with a specific number score 2.3 points higher on average than messages with vague social proof. One real number beats three logo drops.

    Frequently asked questions

    What does good cold outreach look like?

    Four things. It mentions something specific and verifiable about the recipient. It names an ICP narrow enough to exclude most companies. It includes a real customer outcome with a real number. And the ask is sized to match the value shown, not the sale the sender wants.

    What score does good cold email get on YourOutreachSUCKS?

    The best we've seen is a B (6.9/10). Nobody has scored an A yet. The most common reason: even strong messages include a premature calendar request that doesn't match the value delivered in the body.

    What is the difference between a D and a B cold email?

    Almost never the product. It's the structure. Same product, same recipient, but with real research and a right-sized ask, the score jumps 4-5 points. One named customer with a number beats three logo drops every time.

    // Principles from high-scoring cases

    What separates B from F

    01
    Research that can be verified
    Not "I've been following your work" - a specific post, a specific argument, a specific moment.
    02
    An ICP that excludes
    "Series B credit funds" is good because most companies aren't that. "Companies like yours" is bad because every company is that.
    03
    A number, not a feeling
    "11 days to 3" is verifiable. "Meaningful improvements" is not. The number earns the ask.
    04
    An ask sized to the evidence
    A one-pager costs nothing. A 30-minute call costs something. Only ask for what you've earned.
    05
    No manufactured pressure
    Good outreach doesn't need countdown timers or competitor threats. If the offer is relevant, urgency is built in.