Day 20: Campaign Retrospective

The concept

Campaigns should make the next campaign smarter.

Too often, they do not. The team launches, reports, moves on, and loses the signal while chasing the next deadline. The result is repeated mistakes: same audience assumptions, same weak offers, same rushed briefs, same channel mix, same vague learning.

A retrospective turns campaign experience into operating knowledge. It captures what happened, why it happened, which decision mattered most, and what should change next time.

Plain English

A retrospective is useful only if it changes the next brief.

AI can help by structuring the review and removing some of the social friction. It can say directly where the campaign missed, where the data contradicts the team's story, and which decision deserves attention.

Focus on decisions, not vibes

Weak retrospective:

Targeting was too broad.

Stronger retrospective:

We chose job-title targeting across three industries instead of narrowing to the segment with the strongest prior conversion signal. That diluted spend before optimisation.

The stronger version identifies a decision. Decisions can be changed.

Ask AI to find:

• the decision that most affected results
• the assumption behind that decision
• what evidence supported it at the time
• what evidence contradicted it later
• what the next campaign should do differently

This makes the retrospective practical rather than therapeutic.

Separate what worked from what won

Even campaigns that miss goals contain useful wins. An email may outperform. One audience may convert well. A message angle may create replies. A landing page may work for one traffic source and fail for another.

AI can help isolate those signals. Carry them forward deliberately.

Similarly, a campaign that hits its goal can still contain weak decisions hidden by one strong channel or favourable timing. Retrospectives are not only for failures.

The critical decision

The critical decision section is the most valuable output. It forces the team to identify the moment that mattered most.

Examples:

• choosing the wrong audience segment
• using an offer with weak intent
• spreading budget across too many channels
• launching without proof assets
• sending traffic to a generic landing page
• failing to split nurture by entry point
• optimising for leads instead of qualified pipeline

Once named, that decision can become a kickoff question for the next campaign.

Create the next brief immediately

A retrospective loses value if it stays separate from planning. The prompt asks for a one-paragraph starting brief for the next campaign so the learning becomes action.

Compare that paragraph to the original brief. If it is not meaningfully different, the retrospective was too vague.

Today's practice

Choose one campaign. Run the prompt. Then capture:

1. The critical decision
2. What we carry forward
3. What we stop
4. The next brief implication

Share those four items with the team. The point is not to assign blame. The point is to make the next campaign more intelligent.

Run a pre-mortem from the retrospective

The best use of a retrospective is improving the next kickoff. Take the critical decision and turn it into pre-mortem questions.

If the previous campaign failed because the offer was too top-of-funnel, ask:

What evidence tells us this offer matches the audience's current intent?

If the previous campaign spread budget too thinly, ask:

Which channel will we deliberately not use this time?

If the previous campaign had weak proof, ask:

What proof must be ready before launch?

AI can generate these questions, but the value comes from asking them before the next campaign is locked.

Store campaign learnings

Create a campaign learning log. For each campaign, save:

• objective
• audience
• offer
• channel mix
• best-performing message
• weakest assumption
• critical decision
• carry-forward practices
• stop-doing list

This log becomes a strategic asset. New team members can understand what has been tried. Agencies can avoid repeating mistakes. Leaders can see how campaign thinking is improving over time.

Keep the tone blame-free but direct

Blame-free does not mean vague. A useful retrospective can be direct without being personal. Focus language on decisions, assumptions, constraints, and evidence.

Instead of:

The team did not plan well enough.

Write:

The campaign launched before the proof asset was ready, which weakened the landing page and sales follow-up.

That sentence is specific, fair, and actionable.

Schedule the retrospective before launch

The best time to schedule a retrospective is before the campaign begins. Put it on the calendar when the launch date is confirmed. This signals that learning is part of the campaign, not an optional postscript.

Also decide what data will be needed: channel performance, conversion rates, spend, pipeline or revenue, audience segment performance, creative performance, nurture engagement, and sales feedback. If you wait until the retrospective to gather data, the meeting becomes reporting instead of learning.

Include the right people

Invite people who can explain the decisions and outcomes: campaign owner, performance marketer, content or creative lead, sales representative, and lifecycle owner if nurture was involved. Keep the group small enough for direct conversation.

AI can prepare the first-pass retrospective, but the team provides context the data cannot show: constraints, trade-offs, stakeholder pressure, timing issues, and execution realities.

Convert lessons into operating rules

The final output should include one or two operating rules for future campaigns. For example: "No paid campaign launches without a proof asset," or "Campaigns under this budget focus on one primary channel." Rules prevent repeated mistakes better than vague learnings.

Share the learning quickly

Retrospectives lose value when they stay with the campaign owner. Share the carry-forward and stop-doing list with anyone who touches the next campaign: content, paid media, lifecycle, sales, product marketing, and leadership.

Keep the share-out short. A one-page summary is usually enough: goal, result, critical decision, carry forward, stop, next brief implication. The shorter it is, the more likely people will read it and use it.

Turn findings into backlog items

Retrospective findings should create concrete work. If the landing page lacked proof, create a task for a proof block. If the offer was too broad, create a task for a sharper segment-specific version. If targeting was weak, create a task to rebuild the audience logic before the next launch.

Use three categories:

1. Fix before next campaign
2. Test in next campaign
3. Document as operating rule

This prevents the retrospective from becoming a conversation that everyone agrees with and nobody acts on.

Revisit the retrospective before the next launch

At the next campaign kickoff, reopen the previous retrospective. Read the critical decision, carry-forward list, and stop-doing list before approving the new brief.

This one habit creates compounding improvement. The team stops treating campaigns as isolated events and starts treating them as a learning sequence.