NotebookLM Short Video Overviews: create better 60-second videos

NotebookLM's Video Overview tool currently offers multiple formats for different jobs. Short is the compressed option: Google describes it as a way to quickly grasp key concepts from your sources in a video of roughly one minute.

That makes Short useful when the job is orientation, recall, or a fast explanation rather than comprehensive synthesis. A one-minute output can surface one central idea, but it cannot carry every caveat, secondary finding or citation trail from a dense source set.

The page's worked example is presented vertically, which makes the format naturally useful on phones. I treat that as an observed output characteristic, not as a universal product guarantee, because Google's current Short help text does not publish a fixed aspect-ratio specification.

Native Short strategy: design for one minute from the start

1. One Short = one question. Do not ask one minute to summarize an entire notebook.
2. Use a focused source selection. Reduce competing evidence before generating.
3. Write the hook into the instruction. Specify the surprising finding, misconception, tension or decision that should open the video.
4. Protect one proof point. Ask for one figure, quotation, example or source-grounded fact that earns the conclusion.
5. Protect one caveat. In research and teaching, compression should not erase the limitation that changes interpretation.
6. End with a next step. Send viewers to the paper, class discussion, full Video Overview, notebook or longer explanation.

Google documents what the Short format does, but not a complete internal production pipeline. For practical use, it is safer to think in terms of four operator decisions you can control: what sources are in scope, what single question the video must answer, what evidence must survive compression, and what the viewer should understand at the end.

1. Control the source set. A focused notebook gives the Short fewer competing narratives. If one finding matters, temporarily select only the sources needed to explain that finding.

2. Control the question. Ask for one job, not a miniature documentary: explain one result, correct one misconception, prepare one discussion, or surface one decision.

3. Protect evidence and caveats. Name the figure, example, limitation or uncertainty that must not disappear when the material is compressed.

4. Review the output against the sources. The generated video is a communication layer over your notebook, not a substitute for the underlying evidence. Regenerate when the emphasis is wrong or an important qualification disappears.

A longer Video Overview and a native Short can come from the same sources but serve very different moments. The longer format is better for building understanding across several connected points; the Short is better for making one point memorable enough to trigger the next action.

1. Build or open a focused notebook. Use the sources you actually want the video to explain. A narrower source set makes it easier to control the central idea.
2. Open Studio and choose Video Overview. Select Short as the format when it is available in your account.
3. Customize before generating. Tell NotebookLM who the audience is, what question the video should answer, which evidence matters most, and what must not be omitted.
4. Generate, then verify against the sources. Check whether the video selected the right claim, preserved the necessary caveat, and used visuals appropriately. Regenerate with tighter instructions if it did not.

Google's current Help page documents Short as a live Video Overview format, but feature visibility and plan-level generation quotas can still vary. If Short is missing from your account, check the current Video Overview help and the site's verified Studio limits table → rather than assuming a fixed rollout rule.

Stage 1 — Source engineering. NotebookLM reads your formatting, so format for the outcome you want. Convert messy tables into clean Markdown tables. Insert explicit callouts in the source text — a bracketed line like [CRITICAL DATA POINT: 43% efficiency gain] flags a figure the renderer can promote to an on-screen graphic. Rename files to short, clean titles before uploading; the layout engine often surfaces the source filename as an on-screen anchor, and final_draft_v3_marketing_brief_copy.pdf is not the caption you want on a polished clip. Q3 Marketing Report is.

Stage 2 — Persona and focal-point curation. In the instructions field, define the target audience, the exact thesis to defend, and the pacing style. Don’t let the model pick a generic voice unless that’s genuinely what you want. The more specific the persona and the focal point, the less the model has to guess — and guessing is where sixty precious seconds get wasted.

Stage 3 — Structural audio-visual control. Tell the model how to spend the minute. Allocating explicit time blocks — hook, core problem, evidence, conclusion — prevents awkward mid-sentence cutoffs and keeps the clip from front-loading fluff. If a specific chart should appear, label it clearly in the source (Figure 1: Customer Acquisition Cost) and reference it in the prompt: “feature Figure 1 during the core-metric segment.”

The most useful prompt is not the longest one. It tells NotebookLM which idea must survive compression, what evidence to keep, and which caveat cannot be dropped. Start with the template closest to your job.