Vision Score vs Craft Score: What Each One Measures

TL;DR: Wave Vision gives every analyzed video two scores. The Vision Score (1 to 100) rates distribution: how far the video actually travelled, driven mostly by its views. The Craft Score (1 to 100) rates execution: your hook, script, pacing, call to action and edit, with views left out entirely. Use the Vision Score to see what's travelling in your niche. Use the Craft Score to judge your own videos and decide what to fix next.

You made a video you're actually proud of. Tight script, clean edit, a hook you rewrote four times. It got 312 views.

Then you run it through an analyzer and it comes back with a 41. Ouch. Does that mean it's a bad video?

Not necessarily. And that's the whole reason the Vision Score vs Craft Score split exists. One number can't honestly answer two different questions. "How far did this go?" and "How well was this made?" sound related, but on short-form platforms they come apart all the time.

So Wave Vision answers them separately. The Vision Score tells you about distribution. The Craft Score tells you about execution. When you read them side by side, a low number stops being a verdict on you and starts being a clue about where the problem actually is.

Here's what goes into each score, what each one rewards, and when you should care about one more than the other.

What Is a Vision Score?

The Vision Score is a 1 to 100 rating of how far a video actually travelled. Views set most of it, on a scale where each tenfold jump in views moves you up a tier. Engagement rate and script quality then nudge the result up or down by a few points. It describes what happened, not what will happen.

In the analyzer, the Vision Score card is labelled "distribution" and described as "How far this video actually travelled." That's an accurate summary.

Three things go into it:

• Reach. The view count, on a log scale. Going from 1,000 to 10,000 views moves the score about as much as going from 10,000 to 100,000. This is the biggest input by far.
• Engagement. Likes, comments and shares divided by views. If you want context on what a healthy rate looks like, our guide to a good TikTok engagement rate covers the benchmarks.
• Craft. Hook, content, CTA and pacing, read from the transcript. These adjust the score but don't drive it.

We built it "outcome-first" on purpose. An older version leaned heavily on craft, and it produced a silly result: videos with millions of views scored around 50, because going viral pushed engagement rate down. Now an A+ has to be earned with real reach. Videos around a million views usually land at 90 or above. A small video with a perfect script can't get there.

That reflects how platforms work. Instagram has said it shows each eligible Reel to a small group first, then expands it to bigger groups only if the early viewers engage. Reach is something the platform hands out in rounds. The Vision Score just keeps track of how many rounds you won.

Two fairness rules are built in. If we can't read a video's likes or comments, engagement goes neutral instead of counting as zero. And if the video isn't in English, the script parts are left out without any penalty, so nobody scores lower because of the language they speak.

What Does the Craft Score Measure?

The Craft Score is a 1 to 100 rating of how well a video is made, completely independent of how many people saw it. It looks at six parts of execution: your hook, content, call to action, speaking pace, opening visual and cut rhythm. A video with 200 views can score 90 here.

In the app, the Craft Score card is labelled "execution" and described as "How well it was made, yours to control." That last part matters. You can't control whether the platform picks your video for a bigger test group. You can control your first line and your first cut.

The Craft Score isn't a chatbot's opinion, either. It's calculated from two things we can actually measure: your transcript, and the timing of the visual changes in your edit. Run the same video twice and you get the same score.

The Six Parts of a Craft Score, in Plain English

Each part is scored out of 10 and shown in its own tile. The hook carries the most weight of any single part, because it decides whether anyone watches the rest. The two visual parts together make up a big share, because they cover the half of a video that a transcript can't see.

1. Hook. We read the opening lines of your transcript and look for reasons to keep watching. That means openers like a question, a pattern interrupt ("Wait, stop doing this"), a specific promise ("I'll show you exactly how") or a curiosity gap. A flat "Hey guys, so today..." scores low. TikTok's own ad guidance says to land your hook within the first 6 seconds and state your main point in the first 3. If your hook score is low, start with our guide to why viewers drop off in the first 3 seconds.

2. Opening visual. How quickly the first visual change arrives: a cut, a new angle, a movement. A change in the first second or two scores best, and a static opening that holds for several seconds scores lowest. People decide fast. Research Facebook cited found people could recall feed content after just a quarter of a second of exposure.

3. Cut rhythm. How often the picture changes, measured as visual changes per 10 seconds. This one has a sweet spot rather than a "more is better" slope. A frozen talking head and a strobing, chaotic edit both lose points, for opposite reasons. If your talking-head videos score low here, try these pattern interrupts that work without fancy editing.

4. Content. Whether the script has a clear shape. Problem-and-solution, step-by-step tutorials and numbered lists score well. Scripts with real depth earn a bit more, and thin, surface-level ones lose a bit.

5. Pacing. Your speaking speed in words per minute. A natural, conversational pace scores best, and very slow or very rushed delivery scores lower. For reference, everyday American English conversation runs about 150 words per minute.

6. CTA. Whether you ask the viewer to do something, and how well. Having no call to action scores low. A specific ask that names a benefit ("Follow for a daily editing tip") scores higher than a generic "follow me." So does a CTA that isn't crammed into the last second.

One honest note: the cutoffs for "good" in each part are our product judgment. They're not platform rules, and no platform publishes rules like these.

What happens when we can't see a part. Sometimes we can't score everything. The script parts are English-only today, and the visual parts need us to detect real cuts in the edit. When a part can't be read, it shows a dash and "not scored" instead of a zero. The Craft Score is then rebuilt from the parts we did see. If that's less than about half the full picture, the score is marked provisional and shows no letter grade. A grade is the most confident thing we say about a video, so we don't print an "A" we didn't fully earn the right to give.

Why Are There Two Scores Instead of One?

Because reach and quality are different things, and blending them hurts the creators who most need feedback. A single number either rewards views, which tells a small creator nothing they can act on, or rewards craft, which makes viral videos look mediocre. Two scores keep both questions honest.

Think about what a small account sees with a single blended score. Every video comes back low, because the views are low. The feedback says "you failed" without saying where. That's discouraging, and it's not even accurate.

Platform reach is also lumpy by nature. Researchers at Oxford found that on platforms with strong recommendation systems, like Instagram and YouTube, outcomes are far more concentrated at the top than on platforms without them. When rewards are that skewed, a well-made video can easily land quietly. If your score only tracked results, it would mostly track luck and timing.

This is also why we're skeptical of single-number tools that promise to forecast virality. We wrote more about that in our breakdown of whether viral score predictors are accurate.

Reading the Two Together: Four Combinations

In the full report, the two scores sit next to each other under the heading "Reach vs Craft," with the line "Two different questions: how far this went, and how well it was made." The interesting part is the gap between them.

High Craft, low Vision. The video is well made and didn't travel. This is the most common result for growing accounts, and it's good news in disguise. Your execution is working. Look at topic, timing and the audience you're reaching before you rip up your hook. Our guide on telling a hook problem from a topic problem walks through that diagnosis.

Low Craft, low Vision. Both are weak. Start with the lowest Craft tile, since that's the part you control. Fix one thing, post again, and compare.

Low Craft, high Vision. The video travelled despite rough execution. Usually the topic, the timing or a trend carried it. Don't copy the execution. Ask what the topic did, and whether a better-made version could go further.

High Craft, high Vision. Well made, and it travelled. That's a format worth repeating. If you have a few of these, look for what they share; our guide to finding your best-performing content patterns shows how.

When Should You Look at Vision Score vs Craft Score?

Look at the Craft Score when you're judging your own videos, especially small ones, and deciding what to fix next. Look at the Vision Score when you want to know what actually travelled, such as when you're studying other creators' videos in your niche. Most of the time, the useful answer is in the gap between them.

Use the Craft Score when:

• You have a small or new account and every Vision Score is low by default.
• You want to know which part of a video to fix next. The lowest tile is your next experiment.
• You're tracking whether your execution is improving from one video to the next, without luck muddying the picture.

Use the Vision Score when:

• You're studying a competitor's or a peer's video and want to know if it really travelled before you learn from it.
• You're picking which videos to break down, because studying a video that didn't travel teaches you less.
• You want a quick, consistent read on distribution. It's the same scale in the free analyzer and inside the app, so a 90 means the same thing everywhere.

One caution: the Vision Score reads raw views and doesn't adjust for follower count. A 20,000-view video from a 2,000-follower account and from a 2-million-follower account get the same reach input. Keep account size in mind when you compare.

Want to see both scores on a real video? Paste any public TikTok, Instagram Reel or YouTube link into the free video analyzer. No account needed. It works on other creators' videos too, and the full report shows the Vision Score and Craft Score side by side.

What Neither Score Can Tell You

Neither score forecasts the future. The Vision Score describes reach that already happened. The Craft Score describes execution, which improves your odds but guarantees nothing.

Platforms weigh signals we can't see from a video file. TikTok says its For You feed considers things like what each viewer likes, shares, follows and watches, along with captions, sounds and device settings. Instagram's Adam Mosseri has named watch time, likes and sends as its top ranking signals. Those depend on who saw your video, and that changes every time.

So use your platform analytics alongside both scores. On YouTube Shorts, for example, the "viewed vs swiped away" figure shows how many people stayed past the opening, and Buffer calls it one of the most important metrics in Shorts analytics. If your hook tile is low and your swipe-away rate is high, those two are telling you the same story. For more on that metric family, read our explainer on hook rate and why it controls your reach.

One more distinction. If you connect your social accounts, your dashboard also shows an account-level Vision Score. That's a different roll-up across your posts (engagement, growth, consistency and share rate). This article is about the two per-video scores in the analyzer.

The Bottom Line

The Vision Score tells you how far a video went. The Craft Score tells you how well it was made. You need both, and you need them separate.

If you're a growing creator, the Craft Score is the one to watch week to week. It's the part you control, and it moves when you get better, even before your views catch up. Then let the Vision Score show you which videos in your niche are worth studying.

A full analysis gives you both numbers, each part of craft scored out of 10, and specific fixes tied to moments in your video. Start free, no card needed, and your analyses stay in your library so you can watch your Craft Score climb.