HeyGen teaches the lesson. Cutroom makes the ad that sells it

Three properties a course library needs and a recorded human cannot supply.

Consistency across a year. Module three and module thirty-eight look like the same person on the same day. A student notices drift and reads it as a course abandoned halfway.

Correction without reshooting. A price moves. A feature is renamed. You change the line and re-render rather than rebuilding a set.

Length with no ceiling. A twelve minute lesson is normal. Onboarding, walkthroughs, module intros, the whole curriculum.

Language reach on top of that, if you sell across borders. Their tiers are published on their own site and they move, so read them there.

Keep all of it for the course. Then look at the one video that sits outside the course, where a stranger has not agreed to watch anything. That is the ad, and it is the video Cutroom builds.

A student pressed play. They paid. They set aside the evening. They want the information. A steady presenter talking straight through is exactly correct for that person.

A stranger scrolling did not choose anything. They are two seconds from a video of a dog and they owe you nothing.

So the ad has requirements the lesson never had. The first sentence has to stop a thumb. The picture has to change before attention drifts, roughly every three to five seconds. The captions have to carry the message alone, because the sound is off. The pace has to move when the argument moves.

One unbroken shot of a face saying good sentences fails every one of them. That is true of filmed footage too. It is not a rendering problem and better rendering does not fix it.

This is why a lesson clip run as an ad performs the way it does. The clip is not bad. It answers a question the viewer never asked.

The fix is not a longer script or a nicer presenter. It is a cut, and that is the thing Cutroom hands back.

Inside the course, who is speaking barely matters. The value is the curriculum, and a consistent presenter delivering it well is a feature.

At the top of the funnel it is the whole product. Somebody deciding whether to hand over money for your knowledge is deciding whether to trust you specifically. There is no way to make that decision about a presenter you rented.

So the ad wants the twenty seconds where you say the thing you always say on a call, in the room you always say it in. The pause before the number. The half sentence you start again. The detail too boring to invent.

That is the one asset a competitor with a bigger budget cannot buy off you. It costs twenty minutes with a phone.

There is a route if you will not go on camera. The avatar module renders a lip-synced presenter from a photo you supply and a voice-over, at 320 credits per rendered minute. Generated voice-over is 70 a minute.

The photo is yours, or somebody who gave written permission. That render goes through the same batch as any recorded take and comes back cut, covered and captioned.

Record one continuous take of up to three minutes and upload it. A batch is 100 credits and covers the transcript, the director pass and the b-roll search.

What comes back is a finished 9:16 MP4. Cut, covered with footage, captioned from one of six packs, paced.

You change it on the words. Highlight a phrase and a clip lands over exactly those words. Delete a line you do not want and the cut rebuilds around the gap. Change the pace and it re-cuts. Emphasise a word so it pops in the captions. Write the opening headline, or pick one it drafted.

Coverage is capped by the compiler rather than by advice. 40 percent of the runtime at chill pace, 45 at normal, 52 at fast. Even at the top of that range you are on screen for about half the ad, which keeps it reading as a person rather than a montage.

Eight ads from four takes is 400 credits of batches and 80 of exports. 480 credits against the 2,500 Basic carries at $39.99 a month, or 5,000 on Premium at $79.99.

The trial is 7 days and 300 credits with no card. That is two takes and four finished thirty second ads before any money is involved.

Do not start with a camera. Start with the five questions people ask you before they buy. You already know them, because you have answered them in a hundred emails.

Each of those is one ad. Record them as five separate continuous takes, phone at eye height near a window, under three minutes each. Most people finish inside half an hour.

Say the question out loud first and answer it cold. That is the hook. A lesson clip can never have one, because a lesson earns its opening from the module before it.

Batch each take. Read the draft looking for the three seconds you would change. Highlight different phrases where the footage missed, then set the pace.

Export and run one against your best current creative in the same ad set. Same audience, same offer, same budget, four days.

Keep the course on the tool that renders forty matching modules. Keep the ads on the tool that returns a cut. Treating those as one purchase is how course creators end up doing both jobs badly.