VIDZY / FREE MARKETING TOOL
Free Video Performance Analyzer
Find where viewers drop out of your video funnel—and what to investigate next.
Your numbers. Transparent calculations. A clearer next step.
01 / YOUR VIDEO
Start with what you know
Use one video and date range. Only video starts is required. Leave missing numbers blank; enter 0 for a measured zero.
Check your counting basis
Retention checkpoints must count the same starting population, with each later stage a subset of earlier stages. Use matching CTA impressions and clicks. Replay-weighted charts or counts from different date ranges may not fit this model. Visitors, leads and conversions are not automatically subsets of one another.
02 / YOUR NEXT STEP
Turn counts into context.
See the retention you can measure, the actions viewers take and the questions worth investigating.
- Rates calculated from your available data
- Largest observed viewer drop
- Relevant guides for your next investigation
No score. No assumptions about missing numbers.
02 / YOUR RESULTS
Your video, in context
Viewer retention
Only supplied checkpoints. Loss is relative to the preceding supplied stage, not video starts.
Actions & outcomes
These events can occur during or outside playback. They are not stages after completion.
03 / INVESTIGATE NEXT
What the numbers suggest inspecting
Observations describe counts. Investigations are hypotheses to check, not proven causes. Missing checkpoints can conceal larger drops; interval lengths may differ.
FROM OBSERVATION TO INVESTIGATION
Bring the next question to your video.
Vidzy offers video analytics, drop-off analysis, timed CTAs and timed email opt-ins. Use these capabilities to investigate viewer behavior and test a clearer next step.
READ YOUR RESULTS WITH CONFIDENCE
What video performance means
Video performance connects opportunity, attention, action and outcomes. Total views tell you how often playback began. Retention tells you how much of that starting population reached a checkpoint. Neither number alone explains whether the video helped a viewer take a useful next step.
Which metrics matter?
Start with your goal. Play rate relates starts to visitors; retention and completion describe continued watching. CTA CTR relates clicks to impressions. Leads and final conversions describe outcomes, provided attribution matches your video and reporting window.
Exposure is different from engagement
CTA exposure describes opportunities to see an action. CTA CTR describes clicks relative to those opportunities. Few impressions can reflect timing, eligibility or tracking definitions. A click count alone cannot distinguish these explanations.
A funnel needs connected populations
Retention checkpoints can form a nested sequence. A lead may arrive before completion, and a conversion may come from another visit. Confirm attribution before calling a difference between counts a loss. Repeated events can produce some ratios above 100%.
There is no universal “good” rate
Audience, video length, intent, placement and measurement definitions change what a rate means. Compare similar videos or your own earlier results. This analyzer applies arithmetic and visible rules; it does not apply industry averages or predict an uplift.
Common questions
Can I use this with only video starts?
Yes. You will see your supplied count and guidance on which data to collect next. A retention checkpoint or completions enables a first viewer-loss comparison. Optional fields remain optional.
Why is a rate missing when I entered zero?
A measured zero is valid. Zero divided by a positive denominator is 0%. A zero denominator makes the rate undefined, so it is omitted. A blank field means the number is unknown, not zero.
Does this analyze a video file or use AI?
No. Enter counts from your reporting tools. The analyzer uses deterministic formulas and diagnostic rules. It cannot establish why a viewer left, watch your video or predict future performance.
How is the largest observed viewer drop chosen?
For each pair of successive supplied retention checkpoints, subtract the later count from the earlier count, then divide by the earlier count. The largest positive percentage loss is highlighted; ties are reported. Missing checkpoints are not estimated, and wider intervals are not treated as evidence of an early, middle or late location.