MCP

Campaign planning workflow

TL;DR

  • Seven steps, in order: thesis, list, offer, sequence, infrastructure, QA, launch. The order is the point — each step's output is the next one's input.
  • The two most commonly skipped steps are the first and the last, and they are the two that determine whether the campaign was worth running.
  • Write the thesis down. A campaign you cannot state in one sentence cannot be diagnosed later.

The seven steps

1. Thesis. One sentence: we believe [this specific group] has [this problem] right now, and [this offer] is worth a reply. It names the segment, the problem and the reason it is urgent. Everything downstream is a test of this sentence, and if it turns out to be wrong, no amount of copy iteration rescues the campaign — which is exactly what makes writing it down valuable.

2. Success criteria. The number, unit and date from Goals and success criteria, sanity-checked against Outbound math and economics. This is where you discover the list needs to be three times bigger, while that is still cheap to fix.

3. List. Define, source, enrich, verify, in that order — the whole of Audience and lists. The list is the term with the largest leverage in the arithmetic, so it deserves the largest share of the preparation time. Segment it now, because segments are what make the copy specific.

4. Offer and message. What you are proposing and why it beats doing nothing — Offer and messaging. Do this before writing emails: the message strategy is a campaign-level decision that every email inherits.

5. Sequence. Touch count, spacing, and the job of each touch — Sequences and follow-ups — then the individual emails, Writing the email. Write the whole sequence before sending any of it, so the second email is not improvised after the first one underperforms.

6. Infrastructure. Domains, mailboxes, authentication and the warmup runway — Deliverability and compliance essentials and Volume limits and ramp. This step has a lead time of about a month, which is the single most common scheduling surprise in outbound. Start it in parallel with step 3, not after step 5.

7. QA and launch. The checks in Pre-launch QA, then a deliberately small first batch.

Launch small

Send to 50-100 contacts before committing the list. You are not testing copy at that size — the sample is far too small to tell you anything statistically (A/B testing) — you are testing whether the machinery works: variables render, replies arrive where someone will see them, bounces are where you expected, nothing is on fire. Then scale.

After launch

Set the review date when you launch, not when you get nervous. Two checkpoints are usually enough: one early for mechanics (bounces, complaints, obvious rendering failures), one after enough volume for the primary metric to mean something. What you do at the second checkpoint is Diagnosing underperformance.

The step nearly everyone skips is writing down what happened. A campaign whose thesis, list definition and results are recorded turns into evidence for the next one; a campaign that only exists as sent mail teaches nothing — Scaling what works.

References

  1. Marylou Tyler & Jeremey Donovan — Predictable Prospecting (book, McGraw-Hill 2016)