Why Your KPIs Are Lying to You

⚠️ When KPIs Become the Goal, They Stop Working Gradually & Slowly

Marketing & Media has the same problem, except that we call it “optimization.” 😅

• We cut the budget, & ask for more
• We cut the headcount & also ask for even more
• We cut the quality so that we meet the deadline!!!
• We pray & pay for optimization!!!

We workship “OPTIMIZATION”, we workship FAILURE as well (Tommy Nguyen)

🔄 How It Corrupts Marketing Metrics

Every KPI is a proxy: an approximate representation.

When that proxy becomes a target, 3 things happen in sequence:

1️⃣ Proxy Problem: The gap between “what we measure” & “what we actually want” gets exploited.

2️⃣ Optimization Pressure: People find the shortest path to hit the number: optimizing the metric, not the outcome.

3️⃣ Feedback Loop:

• Good numbers trigger positive feedback.
• Gaming behavior gets reinforced.
• Data loses meaning
• Strategic decisions go wrong.

🎯 The 5 Most Vulnerable Marketing KPIs

👉 CTR: Clickbait headlines boost CTR while bounce rate skyrockets. The click happened so the value did not.

👉 Reach: Facebook counts a 1-second scroll as a reached user. The number looks great, however awareness does not move

👉 ROAS: Retarget existing customers to inflate ROAS. High return, zero new customer acquisition.

👉 Engagement Rate: Like and comment farms deliver beautiful engagement metrics. Brand value stays at zero.

👉 Brand Awareness %: Top-of-mind surveys are easy to influence. The metric is accurate. The insight is not.

🇻🇳 Case Study: The View Count Problem

Many brands celebrate 5 million monthly reach while brand trackers show zero or little awareness lift.

KOL view farming becomes standard when view count becomes the contract KPI.

The metric was hit. The objective was missed. 😔

🛡️ How to Fight Goodhart’s Law

– For Media Planners: When briefed to “increase CTR,” always ask the reverse question: will CTR growth come with quality traffic growth?

Set up A/B tests comparing CTR-optimized vs. quality-optimized campaigns.

– For Brand Managers: Ask: “Does this ROAS include returning customers or only new ones?

Is brand search excluded?”

– For Marketing Directors: Build a Balanced Scorecard with brand health, short-term sales, and efficiency metrics in parallel.

Triangulate from multiple measurement sources.

💡 Do not ask whether the campaign hit its KPI. Ask whether that KPI is still measuring what you actually want to achieve.

---

❓ Frequently Asked Questions

What is in marketing?

When a KPI like CTR, ROAS, or engagement rate becomes the optimization target, teams find shortcuts to hit the number without achieving the actual business objective.

The metric improves on paper while real performance stagnates or declines.

Which marketing KPIs are most vulnerable to gaming?

The five most commonly gamed marketing KPIs are:

• CTR (clickbait inflates it),
• Reach (platform definitions are loose),
• ROAS (retargeting existing customers inflates it),
• Engagement Rate (bot farms)
• Brand Awareness surveys (easy to influence with priming).

Incrementality testing is the strongest antidote.

How can marketers avoid?

Use a Balanced Scorecard that combines brand health, sales, & efficiency metrics instead of optimizing a single number.

Triangulate from multiple measurement sources, run incrementality tests, and always ask whether the KPI still reflects the actual business goal.

---

TOMMY Nguyen (H.Thinh)