Campaign benchmark verdict

Purpose

A reply rate of five percent can be excellent for cold outreach to senior executives and poor for re-engaging existing users. A number alone supports no decision. This skill supplies the comparison that gives it meaning, and turns the result into a verdict the team can act on: celebrate, optimize, or stop.

When to use

• Before setting a target for a launch.
• When a campaign's numbers need a verdict for someone who lacks context.
• When comparing segments, clients, or time periods.
• When a number has dropped and the team needs to know whether it matters.

Route elsewhere when:

• The question is how to measure. Use paid-measurement-scorecard.
• The question is which optimization to make. Use paid-optimization-signals.

Inputs

• The number to judge, with its metric, denominator, and period.
• The team's own history for a comparable campaign: same channel, audience, offer, and season.
• Any target set before the results were known.
• External averages, used only as a rough guide and labeled as such.

Procedure

Step 1: Define the number

Define the number exactly: the metric, the denominator, and the period.

Step 2: Compare against the team's own history first

Find the comparable history. Match the channel, audience type, offer, and season. If no comparable history exists, say so.

The team's own history for the same kind of campaign is the most useful layer, because it reflects the team's own audience and product. State the figure, the historical range, and where this result sits.

Step 3: Compare against the target

A target is a useful comparison only if it was set before the results were known. Compare against it if one was set before launch.

Step 4: Compare against external averages

External averages are published figures aggregated across different sources, audiences, and dates. Use them last, only if the team has no history or target, and only as a rough guide, because they often describe a different kind of campaign. Note that the figure may not be like for like.

Step 5: Give the verdict

Name the layer used in every verdict.

• Celebrate if the result is clearly above the comparable history and target.
• Optimize if it is within range but below target, and the gap is worth closing.
• Stop or rethink if it is clearly below history and target for a sustained period.

Step 6: State the action

State the action the verdict implies, and the measure that will show whether it worked.

Output

• The metric, the comparison layer used, and the figure against it.
• The verdict: celebrate, optimize, or stop or rethink.
• The action, with its check date.

Quality bar

Every verdict names its comparison layer. Team history is used before external averages. A target counts only if it was set before launch. Like is compared with like. Watch for a verdict based on one external figure, a target set after the results were known, and a comparison across different campaign types.