The tag everyone has is the one doing nothing for you
New AI tag suggestions, built from what actually ranks: the tags the leaders in your searches carry and the field below them doesn't. Plus ten times the Pro allowance and Motion Array rankings.
You've written a Christmas track. You tag it christmas. Obviously.
So does everyone else, and that's the whole problem.
We watch a search called christmas card. Three items sit at the top, thirty-five sit behind them, and here's what separates the two groups:
| tag | top 3 | the 35 behind | lift |
|---|---|---|---|
carol |
2 | 2 | 9.6× |
eve |
2 | 9 | 2.5× |
family |
2 | 13 | 1.8× |
christmas |
2 | 31 | 0.8× |
carol is on two of the three winners and on two of the thirty-five behind them. christmas is on thirty-one of those thirty-five. It's a perfectly good tag. It's just not buying you anything — you're spending one of your thirty slots on a word the entire page already has.
That gap is what the new suggestions go looking for. Not tags that suit your track: you know those, you wrote twenty-seven of them yourself. The ones that suit it and that nobody around you thought of.
One search at a time, always
A tag that splits the field in christmas card can be dead weight in corporate uplifting. So the comparison never leaves the search it was made in, and every chip tells you which one it came from, plus your position there if we've seen your item in it.
We watch 3,727 searches and re-read them every week, of which 300 have built up enough ranking history for this measurement so far — a number that climbs every week, which is a roundabout way of saying the thing gets better while you sit there doing nothing at all.
Where they turn up
Inside the tag field, as dashed chips next to your real ones. Tap one and it's a normal tag. The cross removes it.
The strong ones have dots. One dot: the leaders use that tag twice as often as the field below. Two dots: five times. No dots means we haven't measured it yet, and the chip came from a vocabulary of 1,556 tags this platform actually uses.
Turn a suggestion down and we remember. Ask again, you'll get different words.
The boring part, which is the important part
Every candidate goes through the same tag checker your own tags do, before it ever reaches your screen.
That's less of a formality than it sounds. Two tags sharing a word root is fine; three is not — so a suggestion that looks harmless on its own can shove you over the line the moment it lands beside something you already had, which is why candidates get tested against your entire set instead of one at a time. If any were dropped, the panel says how many.
Two smaller rules, both arrived at the hard way.
Single words only. Unless it's a set phrase like acoustic guitar, a two-word tag tends to drag a second word root in behind it, and Envato's repeat rules count roots rather than tags. Your own two-word tags are your business — this governs what we hand you.
And there's a separate warning for tags that eat a root whole. Every so often a suggestion is completely legal and still spends a word permanently: take it, and nothing else you tag can ever contain that word again. Better to say so than let you discover it three tracks later.
Two modes, and the difference is deliberate
Top up fills the empty slots and leaves your own writing alone. It's the one that runs in bulk: select three hundred tracks and it works through them, quoting the cost before it starts. Short on allowance? It used to refuse outright. Now it does as many as it can and tells you how many it left for next month.
Rewrite all tags bins the set and starts over, and it lives only inside a single track's drawer. It will throw away keywords you chose for reasons no model can see. That's a call you make looking at one track, not at two hundred rows.
Either way, nothing reaches Envato. Edited tracks go yellow like any local change: review them in the table, push when you're happy, or roll them back with bulk Discard.
Pro got ten times the room
One generation, one track. Nothing else in the app burns through an allowance anywhere near that fast, and Pro's old hundred a month had been sized for prepping the odd single file rather than for grinding through a catalogue. Our biggest Pro account has around 2,000 tracks with free tag slots. At 100 a month, that's a twenty-month project.
Pro is 1,000 generations a month now. Free stays at 10, Max stays unlimited.
Motion Array rankings
Different feature, same week. My rankings under Motion Array used to work out who you were by matching your Envato artist name against MA seller names. Those are unrelated naming systems. Somebody who's Sound_Gallery on Envato might be sound-gallery-by-dmitry-taras on MA, and when the two don't line up the guess finds nobody, the page comes back blank, and nothing on it explains why.
Now you just tell us. Settings › Services, paste your producer page link, done.
Linking answers the next question straight away too: whether we hold any chart placements for you at all. A correct link to a seller we've never captured still shows an empty table, and you should know which of those two you're looking at.
If the names do happen to line up, we'll still use that — but there's now a line above the table saying so. Someone sharing your name isn't proof that they're you.