Goals and success criteria
TL;DR
- Name the one outcome that means the campaign worked, before writing a word. "More leads" is not an outcome; "twelve qualified discovery calls by the end of March" is.
- The goal decides the ask, and the ask decides the copy. A campaign built to book meetings and a campaign built to find the right contact are different emails, not the same email with a different closing line.
- Pick one primary metric and accept that everything else is diagnostic.
Start from the end
The classic framing in the follow-up literature is to start with the end in mind: decide what a successful reply looks like, then work backwards to the sequence that could plausibly produce it. It sounds obvious and it is routinely skipped, which is why so many campaigns end with a disagreement about whether they worked.
Three goal shapes cover most cold outreach, and they are not interchangeable.
A meeting. The most common and the most demanding. You are asking a stranger for time, which is the most expensive thing they have. Everything about the sequence — length, proof, the CTA — is shaped by that price.
A referral or a routing. The goal is to reach the right person, not to sell to the person you emailed. This changes the ask entirely, and it converts far better because you are asking for ten seconds instead of thirty minutes. See Referrals and "not now".
An expression of interest. A reply that says "tell me more" or "send the two-minute version". Cheaper to obtain than a meeting and a legitimate primary goal — data on interest-based versus direct asks lives in Calls to action.
Make the criterion falsifiable
A usable success criterion has three parts: a number, a unit, and a date. "Twelve discovery calls with companies of 50-500 employees, by 31 March, from a list of 1,200." That statement can be wrong, which is what makes it useful — it lets you stop a campaign that is failing instead of extending it indefinitely.
Two failure modes to avoid. The first is the unfalsifiable goal — brand awareness, "getting our name out there" — which no cold email campaign should ever claim, because you cannot measure it and it justifies any result. The second is the borrowed goal: adopting a benchmark you read somewhere as your target without checking whether it was measured the way you measure. Reply rates published by different vendors differ by a factor of seven depending on the denominator, which is Benchmarks' subject.
One primary metric
Choose the metric closest to the money that you can actually observe in useful numbers. For most campaigns that is replies or meetings booked — opens are corrupted (Tracking caveats) and revenue arrives too late and too noisily to steer on.
Everything else — open rate, bounce rate, positive-reply share — is diagnostic: not a goal, but the instrument panel you read when the primary metric disappoints. That distinction is what Diagnosing underperformance depends on.
Before you commit
Sanity-check the goal against the arithmetic. If your list is 400 people and you need twenty meetings, the campaign needs a 5% meeting rate — well above what published data suggests is achievable — and the honest conclusion is that the list is too small, not that the copy needs to be better. Do that calculation first: Outbound math and economics.
References
- Steli Efti — The cold email follow-up plan (blog, updated 2026)