The conflict this framework solves
Two systems grade the same video and they want opposite things. YouTube rewards suspense, because suspense keeps people watching. AI search rewards an immediate answer, because an immediate answer is what it can quote.
Pick one master and you lose the other. Script only for retention and the crawlers find nothing worth citing. Script only for extraction and the human leaves in ten seconds.
The five parts below are ordered to satisfy both. The answer goes first, for the machine. The reason to keep watching goes second, for the person.
Setup: name the one person before you write a line
Everything in this framework is written for one person with one problem. Fill these in first. If you skip this step, every blank below gets filled with a guess.
- [Target Audience]: who exactly is this for? Job title, company size, industry. Example: "VP of Marketing at a 50 to 200 person B2B SaaS company."
- [Specific Pain Point]: the one problem they would say out loud, in their words, not yours. Example: "We publish videos and nobody watches past 30 seconds."
- Desired outcome: what their world looks like after they act on this video.
- The one next step: the single action you want them to take. One. Not three.
- Target run time: 4 to 7 minutes is the working range for most B2B explainers.
The rule of the framework: if a line in your script does not serve [Target Audience], or does not move them off [Specific Pain Point], cut it.
Part 1. AI BLUF: answer in the first 20 seconds
Timestamp: 0:00 to 0:20.
Visual strategy: talking head, tight framing, no logo animation. Burn a plain-text lower third with the exact question being answered, so AI search and human skimmers can both parse it.
Fill in the blanks:
"If you are a [Target Audience] dealing with [Specific Pain Point], the short answer is: [one-sentence answer]. Here is how it works, and here is the part most people get wrong."
This block is what the crawlers read. It is the part that gets lifted into an answer when someone asks your buyer question. Deliver it before the story, not after.
Part 2. Hook pivot: earn the attention you just borrowed
Timestamp: 0:20 to 0:50.
Visual strategy: cut to a screen recording, chart, or physical prop that proves the stakes. This is the first visual change of the video, so make it earn the cut.
Fill in the blanks:
"Most [Target Audience] try to fix this by [the common wrong approach]. That fails because [real reason it fails]. We tested this across [number or context] and found [surprising finding]. Which means [reframe of the problem]."
Agreement is cheap. The video earns attention at the moment it tells the viewer something they did not already believe.
Part 3. The step-by-step: the payload
Timestamp: 0:50 to 4:30.
Visual strategy: on-screen numbered steps that persist in the corner. B-roll or screen share per step. A new visual every 15 to 20 seconds so retention does not sag in the middle.
Fill in the blanks:
"Step 1: [action]. Do this because [reason]. Step 2: [action]. Watch out for [common mistake]. Step 3: [action]. You will know it worked when [observable signal]. Here is what that looked like for [example customer or scenario]."
Every step needs a reason attached, not just an instruction. A step without a reason is a to-do list, and nobody cites a to-do list.
Part 4. Conversion bridge: one ask, not three
Timestamp: 4:30 to 5:15.
Visual strategy: return to talking head, then a clean full-frame card with the single offer. No competing buttons, no crowded end card yet.
Fill in the blanks:
"If you want to do this yourself, everything you need is above. If you would rather have it done with you, [the one next step]. That is a good fit if [qualifier]. It is not a fit if [disqualifier], and I would rather tell you that now."
Do not trap the viewer on the platform. Give them a trackable, low-friction asset, something a buyer will actually take, and link it in the description. That link is the bridge between the video and your CRM. Without it, you have no way to tell whether the video did anything.
Part 5. The loop: hand off instead of ending
Timestamp: 5:15 to 5:45.
Visual strategy: end screen with exactly one video thumbnail, held long enough to click. Point at it verbally before the card appears.
Fill in the blanks:
"The question this always raises next is [the natural follow-on question]. I answered that one here. Start with [the specific thing to try first], and when [observable signal] shows up, you will know the framework is working."
A video that ends is a video that stops working. A video that hands off keeps compounding.
Before you record: the five-point check
- Read the AI BLUF out loud. If it takes more than 20 seconds, it is not a BLUF yet.
- Every step in Part 3 has a reason attached, not just an instruction.
- The conversion bridge names one action. Count them. If it is more than one, cut.
- The loop points at a real video that exists.
- Nothing in the script would confuse someone who is not [Target Audience], because you never wrote for them in the first place.
What happens after you record
The script is the source code, not the finished asset. One recording should produce several.
- Correct the transcript by hand, then build a blog post from it.
- Pull the newsletter insight out of the same material.
- Cut five micro-clips for social.
- Embed the video on your own landing page, and send your email and LinkedIn traffic there instead of to the platform.
Why does the AI answer go before the hook?
Because both human skimmers and AI answer engines reward content that leads with the answer instead of burying it after a two-minute intro. Put the summary first and the crawler finds a quotable claim in the first 20 seconds. Put the hook first and it has to dig through story, and most of the time it will not bother.
How long should a B2B video be?
4 to 7 minutes is the working range for most B2B explainers, and the timestamps in this framework are built for that length. The run time matters less than the order. Compress or expand Part 3, which is the payload, and leave the other four parts roughly where they are.
Is a video with 30 views a failure?
Not in B2B. A highly technical video with 30 views is a success if it drives five decision-maker demos. Cross-reference viewers against your CRM instead of grading on view count. That is the whole reason Part 4 insists on a trackable asset.