A report is only useful if the client can read it without a translator. This is the vocabulary most agency reports are built on, with the formula, the question each metric answers, and the trap to avoid when presenting it.
Volume metrics
1. Impressions
What it is: the number of times the ad was served. Answers: how much did we buy? Trap: impressions rise with budget, not with quality. Never present them alone.
2. Reach
What it is: the number of distinct people who saw the ad at least once. Answers: how many people did we get in front of? Trap: reach across platforms cannot be summed; the same person is counted on each.
3. Frequency
Formula: impressions ÷ reach. Answers: how many times did the average person see it? Trap: a frequency above 5–7 for a short campaign usually means fatigue. Show it alongside CTR by week.
4. Clicks
What it is: clicks on the ad. Answers: how many people acted? Trap: accidental clicks on mobile inflate the number. Pair with a downstream metric.
5. Completed views
What it is: video views that reached the end (or the platform’s completion threshold). Answers: did people watch, or skip?
Rate metrics
Rates must be computed from totals: total clicks divided by total impressions. Averaging the daily rate gives the wrong answer whenever daily volumes differ, which is always.
6. Click-through rate (CTR)
Formula: clicks ÷ impressions × 100. Answers: did the creative and placement earn attention? Reading it: display CTRs are small numbers; a move from 0.08% to 0.12% is a 50% improvement, not a rounding error. Show the change, not only the value.
7. View-through rate (VTR)
Formula: completed views ÷ impressions × 100 (for video). Answers: is the story holding attention? Reading it: compare by creative length and by placement, not across the whole campaign.
Efficiency metrics
8. Cost per thousand impressions (CPM)
Formula: spend ÷ impressions × 1000. Answers: what did attention cost? Trap: a low CPM on a poor placement is not a bargain. Show CPM with CTR or VTR.
9. Cost per click (CPC)
Formula: spend ÷ clicks. Answers: what did an action cost?
10. Cost per view (CPV)
Formula: spend ÷ completed views. Answers: what did a real view cost?
Outcome metrics
11. Conversions and cost per acquisition (CPA)
Formula (CPA): spend ÷ conversions. Answers: what did a result cost? Trap: make sure the client knows which event is the conversion. “Lead” and “purchase” are not interchangeable, and attribution windows differ per platform. Write the definition on the report.
12. Return on ad spend (ROAS)
Formula: revenue attributed ÷ spend. Answers: did it pay for itself? Trap: ROAS depends entirely on the attribution model. State it. A report that shows ROAS without the model invites the wrong argument.
The one that is not a metric: pacing
Pacing is where the campaign stands against its plan: days elapsed versus days in the flight, and spend to date versus budget. Clients ask about it constantly because it is the question they get asked. Show it as a simple line: “day 18 of 31, 61% of budget spent, on plan.” Screebl’s report assistant computes this from the campaign dates and budget so the answer is consistent every time someone asks.
Putting them on one page
A layout that clients read without help:
- Top row, four KPIs: one volume (impressions), one rate (CTR or VTR), one efficiency (CPM or CPC), one outcome (conversions or ROAS). Each with the change versus the previous period.
- Trend chart: the volume metric and the rate over time on a dual axis, so a spike in one is visible against the other.
- Split chart: the rate by channel or by creative, ranked.
- Table: the long tail, with all twelve if you like. Tables are for looking things up, not for reading.
- Commentary: three sentences. What happened, why, what changes next week.
When you fill a Screebl report from an upload, the suggested fill follows this shape: KPI tiles get a volume, a rate and an efficiency metric; charts group by date for trends and by a category for splits; rates are computed from totals. You can override any of it, but the default is designed around how clients actually read.
Glossary card for the client
It is worth putting this at the bottom of the first report you send a new client. Two lines per metric, in their language. Reports that define their terms get fewer “what does this mean” emails and more “what should we do” emails, which is the conversation you want.
