Stop counting empty calories. Start feeding your funnel.
This 5-part series takes you from vanity metrics to revenue reality. We’ll cut through the dashboard BS to find the numbers that actually grow your business.
The Empty-Calorie Campaign
“We got 1M impressions!”
Cool story. But how many humans actually saw your ad, remembered it, or did anything about it? Stuffing dashboards with vanity metrics feels good but does nothing for growth. Let’s swap the junk for nutrients.
1. Impressions vs Viewable Impressions
- Impressions = ad code loaded (could sit below the fold or flash for 0.1 seconds)
- Viewable impressions = at least 50% of pixels on-screen for ≥1 second (video: 2 seconds)
Why it matters: Bots, background tabs, and lightning-fast scrolls inflate raw impressions. If viewability is below 70%, your “reach” has a crater in it. (Industry fact: 56% of ads never get seen by humans.)
Attribution note: iOS 14+ privacy changes make viewability even more critical—it’s one of the few metrics you can still trust.
Action: Always show both numbers. Fix viewability before celebrating reach.
2. CPM vs eCPM: The False Discount
- CPM (cost per thousand impressions) = the sticker price
- eCPM (effective CPM) = what you actually pay after filtering out non-viewable, brand-unsafe, or fraudulent impressions
The math: eCPM = (Total Revenue / Total Impressions) × 1000
Reality check: A $2 CPM with 40% viewability = $5 eCPM. Cheap inventory, expensive results.
Action: Track the CPM-to-eCPM spread. Widening gap = sketchy inventory.
3. Frequency Caps: Dodging Ad Fatigue
- Sweet spot: 3–6 exposures per user per week
- Red flags: Falling CTR, rising CPC, or angry comments = oversaturation
- Fix: Cap frequency at campaign level. Refresh creative every 7–10 days
4. The Bounce Rate Myth
High bounce rate ≠ automatic failure. Single-page lead forms or click-to-call landers should bounce quickly.
Judge intent, not averages. Pair bounce rate with:
- Scroll depth
- Time to first action
- Micro-conversions
See if users bailed too soon or completed their task.
5. When CTR & ROAS Lie to Your Face
Click-bait can spike CTR. Targeting only past buyers inflates ROAS. Neither grows your business.
- CTR lies when sensational headlines or fat-finger taps pump clicks—but post-click engagement flatlines
- ROAS lies when attribution windows are too short or branded-search lift sneaks into the math
Truth serum: Check dwell time, scroll depth, and run holdout tests to prove incremental lift.
6. Engagement Rate: Quality Over Quantity
- Calculate engagement against people reached, not impressions served
- Weight high-value actions (shares, saves, form fills) over passive likes
- Shallow reactions don’t move revenue
The Reality Check Checklist
- Report viewability and eCPM alongside raw reach & CPM
- Cap frequency near 5 views/user/week to dodge fatigue
- Layer intent metrics (scroll depth, dwell time) on top of clicks
- Run incrementality tests quarterly to validate ROAS
Tell the Complete Story
Bad: “We got 1M impressions!”
Good: “We reached 480K viewable users at $6.20 eCPM, earned 3.8% CTR with 45-second average dwell time, and confirmed 4× incremental ROAS through holdout testing.”
TL;DR
Impressions, CPM, CTR, and even ROAS turn into vanity metrics when stripped of context. Audit for viewability, effectiveness, and incrementality to convert empty calories into real growth fuel.
Note: These metrics and standards evolve as tech advances. Stay sharp.
