Sixteen terms this site uses constantly, defined plainly, each one pointing at the page where it is worked through in depth. At the end, how every campaign number quoted here is counted.
Message-market fit is the match between what you say and what a specific market already cares about.
You have it when an interested reply costs 200 to 350 contacts instead of 800 to 1,000. It is found by racing genuinely different angles against each other on the same list, not by editing one message until it sounds better. Most campaigns that fail have working infrastructure and no message-market fit.
An interested reply is a response from a real human writing back with intent, rather than any reply at all.
Opens, clicks and out-of-office replies all miss the bar. So do unsubscribes and referrals to someone else, which is why total reply rate flatters a campaign that books nothing. It is the only reply type that turns into pipeline, and classifying replies by hand is what makes every other number on this site mean something.
Positive reply rate is the share of replies that are interested, measured against replies received rather than contacts reached.
It answers a different question from interested replies per contact. Reply rate per contact tells you whether the list and message are working together; positive reply share tells you whether the people who do answer are the right people. A campaign can run a healthy total reply rate and a poor positive share, which usually means the targeting is close but the offer is not landing.
Angle testing is the practice of racing several genuinely different messages against the same market at the same time.
Ten to fifteen angles, each naming a different problem, run against one list in one window. Losers get cut and the winner takes the volume. The distinction that matters is between an angle and a rewrite: fifteen phrasings of one claim teach you nothing, because they all rest on the same guess about what the market cares about.
Deliverability is whether your mail reaches the inbox rather than the spam folder or a block.
It is decided mostly by sender reputation and recipient engagement, with authentication acting as a gate you must pass but earn nothing for passing. Content sits near the bottom of the list, which is why rewriting copy rarely fixes a placement problem. Deliverability is not the same as delivery: a message can be accepted by the receiving server and still never be seen.
Warmup is the practice of raising a new mailbox's sending volume gradually while generating opens and replies, so it builds a history before it sends anything that matters.
It starts around two sends a day and grows roughly 80% a week, reaching full volume in four to six weeks. The second function is diagnostic: a ramp gives you checkpoints, so a placement problem surfaces at fifteen sends a day where it is cheap to fix rather than at thirty where it is a reputation event.
A sending domain is a separate domain, registered by you and pointed at your real site, used only for outbound campaigns.
It carries its own authentication, inboxes and reputation. The separation exists so reputation damage is disposable: a bad week on a sending domain means you pause and warm another, while the same week on your primary domain puts invoices and password resets in spam. Your primary domain never sends cold email, which is the one rule with no exceptions.
A permerror is an SPF result meaning the record exists but cannot be evaluated, so receivers treat the domain much as they would one with no record at all.
The usual causes are more than ten DNS-querying terms, two records starting v=spf1, or a syntax error. It is silent: nothing bounces and nothing warns you, the record simply stops working. Receivers return it rather than guessing, which is why an SPF record that looks correct can still be doing nothing.
Alignment is the DMARC requirement that the domain authenticated by SPF or DKIM matches the domain shown in the From address.
It is the part SPF and DKIM cannot do alone, because neither looks at the address a human actually reads. Without it, a forger can pass SPF on a domain they control while displaying yours. Relaxed alignment accepts a subdomain and is right for almost everyone; strict demands an exact match and breaks the moment a provider signs with its own subdomain.
Bounce rate is the share of messages a receiving server refused rather than delivered.
It is the metric inbox providers treat as a confession, because a hard bounce means you mailed an address that does not exist. The line is 2%, and a properly verified list runs close to zero. A rising rate made of 5.1.1 codes is a list problem; one made of 5.7.1 codes is a reputation or authentication problem, and re-verifying will not touch it.
A catch-all domain is one that accepts mail to any address, whether or not a mailbox exists behind it.
It protects your bounce rate and defeats verification at the same time, because a typo verifies exactly as readily as the real address. Treat catch-all results as unconfirmed rather than good: the message will not bounce, and it may never reach a person. Segment them separately rather than mixing them into a verified list.
A spam trap is an address that exists only to catch senders who did not verify their list.
Recycled traps are old real addresses reactivated after abandonment; pristine traps were never real and never opted in to anything. Hitting either says you acquired data without checking it, which is exactly what the traps are there to detect. They are the most common reason a sender ends up on a blacklist without any obvious change in behaviour.
A signal is a real, recent event about a company that gives you a specific reason to reach out now rather than at any other time.
Funding rounds, job changes, hiring for a role, a tech install, a product launch. A signal is what separates an email that could only have been written for this company from one that could have gone to fifty thousand. Signals decay: one older than a week or two stops reading as timely and starts reading as research.
An ICP, or ideal customer profile, is the description of the companies most likely to buy, built from closed-won data rather than from what a team believes about its market.
The common failure is treating a stated ICP as the real one. A doc saying B2B SaaS Series B+ often hides a real pattern like post-Series-B vertical SaaS in regulated industries with a VP RevOps in seat. Most teams think they have one ICP and actually have four, each needing its own message.
TAM, or total addressable market, is the count of companies and contacts you could reach, as opposed to the revenue you could theoretically capture.
For outbound it is a practical number rather than a strategic one: it decides how many angles you can test at once and how fast you get an answer. Around 30,000 reachable contacts is where testing becomes fast, because several angles can run side by side. Below that the same process still works, it just takes longer to resolve.
Spintax is a notation for writing several versions of a phrase in one place so a sending platform can pick one at random per message.
Its purpose is narrow: it stops thousands of messages being byte-for-byte identical, which platforms and filters both notice at volume. It does nothing for reply rate. Spin greetings and connective tissue, never the claim or the ask, because spinning the substance means sending different messages under one campaign name and learning nothing from either.
Every campaign figure quoted across this site comes from live client campaigns, shown as unedited dashboard screenshots on the homepage. This block defines what those figures count, so a number lifted out of context still means what it meant here.
| Figure | What it counts |
|---|---|
| Interested reply | A real human writing back with intent. Opens, clicks and out-of-office replies all miss the bar, and so do unsubscribes and referrals to a colleague. |
| 1 per 200–350 contacts | Interested replies against contacts reached, once a message is dialled in. Contacts reached, not messages sent, so a three-touch sequence counts each person once. |
| 1 per 800–1,000+ | The industry baseline for an untested message, used for comparison. Not a GrowthPal figure. |
| 67,934 contacted · 449 tagged interested | One client, one documented period. Contacts reached and replies classified by hand as interested. |
| 83.47% positive | The positive share of replies on the best-performing angle, on 4,303 contacted. It is a share of replies, not of contacts, and it is the top of a 52% to 83.47% spread across angles on the same board. |
| 162,867 sends | Messages sent, not contacts reached, in the best documented month. Ran on roughly 110 inboxes across 40 to 50 domains. |
| 4M+ emails · 40+ companies | Cumulative across all clients since founding, not a single campaign. |
| 25–40% of interested replies become meetings | With same-day follow-up. Meetings booked, not meetings held. |
Excluded everywhere: opens and clicks, which are not tracked at all because the pixel costs deliverability and the number cannot be trusted; out-of-office and auto-responses; and any reply classified as not interested.
Two details are not yet documented in this repo and are deliberately not stated above rather than guessed:
The free GTM plan maps how many companies you could reach, which buyers sit inside them and what a realistic month of outbound looks like on your numbers.