Descript finishes the episode. Cutroom finishes the ad

Editing by text rather than by waveform is the correct idea and Descript proved it at scale. Once you have worked that way, dragging a clip to a time feels like arithmetic somebody else should be doing.

It is strong at the jobs a founder actually has. An hour of podcast. A screen recording of the product. A recorded call turned into something publishable. Filler words removed in one pass. Two speakers on separate tracks.

Their plan tiers are published on their own site and they change, so read them there on the day.

This is not a page about a worse tool. It is a page about one job inside your week that a long-form editor was never built to finish.

That job is the paid ad. Thirty seconds, vertical, sound off, competing with a friend's holiday photos for the third second of somebody's attention.

It looks like a smaller version of the podcast edit. It is a different build entirely, and Cutroom hands it back finished.

Long-form editing is subtraction. You start with a good hour and take out the bad minutes. The structure was already there, because you had an hour to establish it.

An ad has no hour. It builds a structure from nothing in the first two seconds. Every part of that structure is a decision the transcript alone does not make for you.

A hook that earns second three. Cutaways on the exact phrases that need proof, and not on the ones that do not. Captions carrying the message, because most of the audience has the sound off. A pace that shifts when the argument shifts.

In a long-form workflow you make each of those by hand after the trim, one at a time. That is forty minutes per ad on a good day.

Twelve ads a month at forty minutes is eight hours. For a founder, eight hours is not a line item. It is the reason there were four ads instead of twelve.

The gap is not quality. You can build a good ad in a long-form editor. The gap is that you build it from scratch, every time.

You keep the interface you like and lose the assembly. Upload one talking-head take of up to three minutes. Spend 100 credits on a batch. The cut comes back built.

Trimmed. Covered with footage on the phrases that earn it. Captioned from one of six packs. Paced. A 9:16 MP4, not a project to finish.

Then you direct it. Highlight a phrase and a clip lands over exactly those words. Delete a line and the cut rebuilds around the gap. Change the pace and the machine re-cuts the whole thing. Emphasise a word so it pops in the captions. Fix a mis-heard word and the captions follow.

There is no timeline underneath and that is deliberate. Give the machine every time calculation and it rebuilds in seconds. Keep any of it yourself and it cannot.

The cap on cutaway coverage is enforced rather than suggested. 40 percent of the runtime at chill pace, 45 at normal, 52 at fast. That is why a take of you talking still reads as you talking.

Six ads from three takes is 300 credits of batches and 60 of exports. 360 credits, inside the 2,500 that Basic carries at $39.99 a month.

Founder-led ads work because a specific person stands behind a specific claim. Swap that person for a rented one and you have spent the only advantage you had over a company with a bigger budget.

So the input is your own take. Your pause before you say the price. The small laugh at the line you think is silly. The fridge you can hear in the background. Those are the details nobody would invent, and viewers read them in about a second.

The avatar module exists for the scripts you will not film twice. A photo you supply plus a voice-over becomes a lip-synced presenter at 320 credits per rendered minute. Generated voice-over is 70 credits a minute.

The same rule holds underneath. The face is yours, or somebody who gave written permission. There is no roster of strangers, because a stranger is what founder-led advertising is trying to avoid.

That render is a source take rather than an ad. Hand it to the same batch for 100 credits and it comes back cut, covered and captioned like any other.

Music is 20 credits a track if you want a bed under the voice. Describe the style, leave the lyrics empty, and an instrumental comes back.

The habit that changes is when you record, not how you edit.

Pick three angles you already know work. You have said them on a call and watched somebody's face change. Record each as one continuous pass, phone at eye height, near a window, under three minutes.

Do not script it word for word. Read the hook cold, say the thing, let the ending run long. The extra sentences are options for later, since deleting a line is one gesture.

Batch all three the same morning. Three hundred credits. Read each draft with one question. Which three seconds would I change.

Export at 20 credits per output minute. Put the first one in the same ad set as your best current creative. Same audience, same offer, same budget, one variable, four days.

The trial covers that entire week at 300 credits with no card. Twenty minutes on camera and 110 credits gets you a real answer about your own account.