Developing
One anecdote, no effect sizes or sample sizes; the author's views are his own.
In 60 seconds
- The example: a quarterly social readout comparing clicks, and two hours of team debate over custom illustrations versus stock photos.
- The team could not explain small performance differences, and a layout treated as a "winner" limited later creative work.
- Each metric should tie to a specific decision; metrics that do not inform decisions should be dropped.
- Audit reporting with three questions: what decision the metric changes, whether the insight justifies the effort and whether the goal has enough time to show progress.
Impact Layer
- What changed
- A practitioner argues for fewer, decision-linked metrics in creative reporting.
- Why it matters
- Over-reading small differences can narrow creative work without producing real gains.
- Who is affected
- Marketing operations, social and creative teams.
- Martech stack area
- Analytics, Content
- Stack impact
- Review your regular reports and remove any metric that no one has used to change a decision in the last quarter.
- Business impact
- Low A contributed opinion piece on MarTech (owned by Semrush) based on one team's experience; the recommendations are general and not tied to measured business outcomes.
- Related technologies
- What to watch
- Your own audit of reporting.
Mike Ukstins, director of strategic accounts and managed services at Cella by Randstad Digital, argues in a MarTech column published October 21, 2026 that chasing small performance gains has a creative cost.
The anecdote
A quarterly social media readout compared clicks and sorted posts into categories. The team spent two hours debating custom illustrations versus stock photos, could not explain small performance differences, and treated a layout as a “winner”, which limited later creative work. No tool or dataset is named, and no effect sizes are reported.
The advice
- Each metric should tie to a specific decision; drop those that do not inform one.
- Data and creative judgement are both needed.
- Audit reporting with three questions: what decision does the metric change, does the insight justify the effort and does the goal have enough time to show progress.
Caveats
The evidence is one team’s experience, not controlled testing. The recommendations are general.
Our read
A report nobody acts on is overhead. Cut it, or tie it to a decision.
Sources
- MarTech: The creative cost of chasing small performance gains (opens in a new tab)
Source for the column's example, advice and caveats.