You’re remaking four videos every time the product screen changes
You want to announce a new feature, but each channel needs a different asset. Your website needs a demo people can click through, while sales reps want a link they can send right away. Social media needs a short video, and presentations need a playable file.
If you record from scratch for every channel, a product rename or button move creates four rounds of edits too. Arcade can reduce that repetition. You first capture the product flow as an interactive demo, then reuse that same demo as the source for an AI video.
That said, this is not “publish to four channels at once with one button.” You can create several assets from one source, but the web embed, public link, and video file must each be distributed to their respective channels.
The source is an editable product flow, not a video file
When you use Arcade’s browser extension to interact with a product, clicks are recorded as screenshots and hotspots, while scrolling, typing, and dragging are recorded as short video clips. Once capture is complete, you have an interactive demo where you can edit step descriptions and hotspots.
This structure matters because it keeps the scope of revisions small. Instead of re-editing one long screen recording, you can find and fix the incorrectly captured step or description. If you need to cover apps outside the browser or move between multiple screens, you need the macOS or Windows desktop app rather than the Chrome extension.
For your first capture, choose one path to value instead of a full product tour. For example, choose only the flow in which a prospective customer sees a result, such as “create the first project → invite a teammate.” That keeps the message clear in both a social teaser and a sales demo.
Even with the same source, the asset differs across four channels
| Use case | Asset | What to do when distributing |
|---|---|---|
| Product experience | Interactive demo | Review the steps and hotspots, then publish |
| Website | iframe embed code | Insert the code in the site editor |
| Sales follow-up | Hosted Share Page link | Attach the public link to an email or message |
| Social media · presentations | MP4 or GIF | Export the file, then upload it to each channel |
Because interactive iframes generally do not work directly in email, use a Share Page link or a GIF or video file. Official guidance also says captions are not included when exporting a regular demo as video, so you need to use the social platform’s automatic captions or prepare separate captions.
Here’s how to go from sign-up to your first distribution
You need an Arcade account, a test environment with sensitive information removed, one target audience, and one product flow. Use the Chrome extension for browser-based products; for desktop products, prepare the official app and screen-recording permissions.
- Sign up for Arcade and prepare the capture tool.
For browser products, install the Chrome extension from Quick Start. For products outside the browser, install the desktop app from the download page. - Capture only one flow in New Arcade.
Use a test account to perform the actual steps. When finished, check that sensitive information such as passwords and customer names is not exposed, and that clicks and hotspots were captured on the correct steps. - Polish the descriptions and publish the interactive demo.
On the free plan, focus on reviewing and publishing step descriptions and hotspots. On Growth or above, you can also configure the Share Page CTA. To add it to a website, copy the iframe from Share → Website → Get code; for the sales team, send the Share Page link generated after publishing. - Create an AI video draft from the same demo.
From an existing demo, select Share → Generate with AI → Video, or go to New Video from Home. Set the demo you just created as the Source, then add a Brand Kit and Guidelines if needed. - State the audience and purpose specifically in Guidelines.
Your first input can say: “A social teaser for new prospects. Lead with the flow for creating the first project and inviting a teammate, and use the final scene to invite viewers to try it for free.” Review the generated sections, headers, narration, product name, and final callout; when testing another direction, use New Draft to preserve the existing edit. - Export the video file and upload it separately to each channel.
Create an MP4 or GIF from Export or Share → GIF & Video. MP4 and GIF exports are currently listed as a Growth feature on the public pricing page, so check support on your account’s Billing page before paying.
When testing for free, calculate credits first
According to the public pricing page checked on September 7, 2026, the free plan provides one published demo, one published video, and 200 AI credits per month. AI videos use 50 credits per 30 seconds, and regenerating also costs credits. Monthly credits do not roll over.
For your first experiment, it is safer to complete one flow end to end than to create several personas at once. Fix the interactive demo’s steps and copy before generating the video to reduce unnecessary regenerations. Prices and plan configurations may change, and plan names in some knowledge documents do not match the current pricing page, so confirm actual export permissions again on the Billing page.
If a demo you plan to export as video contains a form, remove it first. Also, if you enable Autoplay on a web embed, engagement Insights will not be collected. Check the form, autoplay, and caption requirements before distribution.
Success does not end with “we made four assets”
There are three completion criteria for the first test: obtain a published interactive link and web embed code from the same source, and export a fully reviewed video file. If you started on the free plan, exclude CTA editing from the completion criteria and verify only the published outputs available on your current plan.
Next, ask one test user to view more than 90% of the demo screen for over three seconds, then click, hover, or tap through to the final step. Data is processed in 10-minute intervals and may take up to three additional minutes to appear on screen. Do not assume failure if metrics are not visible immediately after the test; wait for the processing cycle, then refresh Insights.
In Arcade, completion rate is the share of users who reach the final step of a linear demo. A view is counted when more than 90% of the screen pixels are visible for at least three seconds, while a play is based on users who click, hover, or tap. That is why pageviews and completion rate differ. In Insights, check whether views, plays, completions, and step-level drop-off are separated.
In a report analyzing 14 million sessions of its own interactive demos, Arcade found that demos with AI voice had a completion rate 14% higher than those without it. Demos with 9–12 steps had a 40% completion rate, versus 22% for 1–5 steps, while 21–30-step demos also rose back to 40%.
You should not read these findings as “adding voice raises completion by 14%” or “9–12 steps is always optimal.” This was an internal vendor analysis observing Arcade customers, not a randomized experiment, and users may finish long demos when their purpose differs, such as during an extended technical evaluation. In your own demo, look first at where users dropped off and whether they reached the final callout, rather than the number of steps.
If you want to dig deeper
Video Generation You can see the actual menu for selecting an existing demo as a source and entering a Brand Kit and Guidelines to create an AI video draft. docs.arcade.software
Exports Useful for checking easy-to-miss conditions right before distribution, such as MP4 and GIF exports and caption and form limitations. docs.arcade.software
2026 Arcade Benchmarks Report See how completion rates differed by step count and the presence of AI voice across 14 million sessions. Read it alongside the limitation that this is observational data, not a causal experiment. arcade.software
