Promotion Analytics: How to Read Delivery, Survival Rate, and CPM Before You Scale

Operators who lose money on paid promotion usually did not buy wrong — they read wrong. They looked at a total view count, saw a number that felt satisfying, and moved on. The underlying data — how fast views arrived, how many held through to a meaningful point, what each thousand impressions actually cost — never got checked.

Promotion analytics exists to close that gap. It turns a delivery receipt into a diagnostic. A 50,000-view TikTok package that delivered in 18 hours tells a completely different story than the same package stretched across five days, and the dashboard numbers will show you which one you got and what to do next.

Delivery rate measures how quickly ordered volume actually posts against your content. A healthy delivery curve is not a spike on day one followed by silence — it is a consistent ramp that mirrors organic momentum. If your promotion dashboard shows 80 percent of a 100,000-view order landing in the first four hours, that front-loading can trigger platform suppression flags before organic amplification has time to compound.

The number to watch is the daily delivery percentage relative to the total order. For most mid-tier content packages, a 15–25 percent per-day ramp over four to six days produces the cleanest downstream engagement patterns. Check your delivery graph every 24 hours in the first half of a campaign, not just at end-of-run.

Pacing controls, available through the scaler surface, let you throttle volume to match a content release cadence. If you are running a three-video series with drops on Monday, Wednesday, and Friday, you do not want all ordered views front-loaded on Monday. Set the pacing to mirror the publishing schedule and the delivery rate will reflect it.

Survival rate is the percentage of delivered impressions that remain counted after a platform's own audit cycle — typically 48 to 72 hours post-delivery. A raw view number that shrinks by 30 percent three days later is not a rounding error; it is a quality signal. Low survival rate means delivered impressions did not meet the platform's retention threshold, which compounds into lower organic reach on subsequent posts.

Benchmark: a well-sourced campaign on a video platform should show a survival rate above 90 percent at the 72-hour mark. Anything below 80 percent warrants a conversation with your account manager and a delivery pause while the source quality is reviewed. Pulling a campaign mid-flight is almost always less expensive than continuing to pour budget into volume that will be removed.

You can track this inside the promotion dashboard by comparing the confirmed delivery figure at time-of-completion against the live count on the platform 72 hours later. The delta is your survival rate. Record it by campaign ID so you build a historical baseline — over five or six campaigns you will see which content formats and posting times correlate with higher retention.

Cost per thousand impressions (CPM) normalizes price across radically different package sizes and platforms so you can make apples-to-apples comparisons. A $120 package delivering 40,000 views is a $3.00 CPM. A $400 package delivering 200,000 views is a $2.00 CPM. The second is more efficient, but only if the survival rate holds — a 200,000-view order with a 70 percent survival rate is effectively a 140,000-view order, which recalculates to a $2.86 CPM and changes the decision entirely.

Build your CPM calculation to include survival rate as a multiplier. Effective CPM = (spend / (delivered impressions x survival rate)) x 1,000. That single formula will prevent you from over-ordering on platforms with chronic audit attrition, and it will surface the highest-value packages in your mix without requiring manual comparison across multiple line items.

A campaign report that only tells you what happened is half a report. The other half is the benchmark row it adds to your master file. Every campaign you run should log: platform, content type, package size, CPM, delivery duration, survival rate at 72 hours, and any organic engagement lift observed in the 48 hours after delivery completed. That file becomes your buying guide.

After eight to ten campaigns, patterns become obvious. You will see that short-form video with captions holds a higher survival rate than video without. You will see that weekend delivery on certain platforms returns a worse CPM-adjusted outcome than Tuesday through Thursday. None of that is visible without a running benchmark file — and none of it requires a data analyst to maintain.

Audience growth compounds when you use prior campaign data to pre-qualify future buys. If your benchmark file shows a consistent 91 percent survival rate on a specific package type, you can scale that package with confidence. If a new package type shows a 74 percent survival rate in its first run, you hold it at small volume until the pattern is confirmed across two more campaigns before committing meaningful budget.

Day one check: confirm delivery has started and the ramp rate is within expected range. If nothing has posted within six hours of campaign launch, flag it — do not wait for the 24-hour mark. Early delivery failures compound into missed windows when content is time-sensitive.

Day three check: pull your current delivered volume against the order total, calculate the in-flight delivery rate, and cross-reference the live platform count. This is where you catch early survival rate erosion before you are too deep into the order to redirect budget. If survival looks clean and delivery is on pace, no action needed — document the mid-campaign snapshot and move on.

Post-campaign close (72 hours after final delivery): run the full survival rate calculation, compute your effective CPM, note any organic engagement anomalies, and add the row to your benchmark file. This close-out step takes under ten minutes and is the single highest-leverage reporting habit in promotion operations.