Back to Blog
GoScript Journal

How to Choose the Right Stats for a Media Kit

·7 min read·GoScript Team /Craft

Vanity numbers versus predictive ones

Some numbers make you feel good and tell a brand nothing. Total likes, lifetime views, and follower counts that include a dead platform all fall into that bucket. They are easy to collect and easy to ignore.

Predictive stats describe what happens when someone watches. Average views per video, watch time, click-through on past sponsor links, and the share of returning viewers. These are the numbers a brand uses to guess at sales.

Pick five numbers, not fifteen

A wall of metrics hides the important ones. Choose five at most and give each a short label in plain language. If a brand has to ask what a number means, you already lost a little of their attention.

For a small channel, raw reach may be your weakest stat. Lean on engagement, comments per video, and the loyalty of your returning viewers. Small and devoted beats large and indifferent when the goal is a real sale.

Pick the stats that describe behavior, not the ones that describe how long you have been posting.

Reach, watch time, and return rate

Reach tells a brand how many people could see the message. Watch time tells them how long those people stayed. Return rate tells them whether the audience actually trusts you enough to keep coming back.

Report all three when you can, because each one covers a weakness in the others. A high reach with low watch time looks impressive until the numbers sit side by side with a quieter channel that holds attention.

Presenting a smaller channel honestly

Never inflate a number to compete with channels ten times your size. State what you have, then explain why it is worth more than the count suggests. A tight niche with high trust can convert better than a broad audience that scrolls past.

If a stat looks weak, pair it with a strength. Lower reach is fine beside strong engagement and a sample of past results. Brands can work with an honest small number. They cannot work with a number that falls apart under a check.

Key takeaways

  • Favor stats that predict behavior over stats that flatter the ego.
  • Limit the kit to five numbers and label each in plain language.
  • Report reach, watch time, and return rate together when possible.
  • State weak numbers honestly and pair them with a real strength.

Turn this into your next script.

Analyze a channel, extract its structural DNA, and generate scripts engineered from patterns that actually work.

Analyze a Channel