What does the collaborative ad creative on a multiplayer canvas workflow look like step by step?
The 5 steps below are the workflow teams run inside Avocado AI, from the first product photo to the exported variant set.
- 01
Open Flows and drop in the brand brief
The founder opens a new canvas, drops in the brand brief, pins three reference ads, and writes a one-line creative direction. Super Agent reads the brief and is ready to generate against it.
- 02
Designer joins and generates variants from Brand DNA
The designer joins the session, pulls in the product references saved in Brand DNA, and generates cinematic pack shots in the aspect ratios the test set needs. Cursors are live so the founder sees each generation appear.
- 03
Paid acquisition lead joins and comments on testable variants
The paid acquisition lead joins, comments on specific frames the team should test, and asks Super Agent for one more variation with a different background. Comments anchor to the exact generations they describe.
- 04
Add voice, music, and AI UGC inside the same session
Voice generation produces the voiceover. The Audio Studio produces a music track that matches the brand energy. AI UGC creators deliver the talking-head clips for the variants that need them. Everything lives next to the cinematic generations on the canvas.
- 05
Director mode finishes the cuts for every platform
The team moves the approved variants into Director mode, finishes the cuts together with music, and exports them in the aspect ratios TikTok, Reels, YouTube, and Shopify need. The weekly test set ships from the same session it was conceived in.
Which Avocado AI tools does collaborative ad creative on a multiplayer canvas use?
7 tools cover the whole collaborative ad creative on a multiplayer canvas workflow. They share one credit pool and one gallery, so nothing is exported between steps.
| Tool | Role in this workflow |
|---|---|
| Brand DNA | Products, references, visual direction, tone and rules every tool reads |
| Image workbench | Prompt or edit your own product photo, up to 4 variations per run, upscale |
| Video workbench | Text to video, animate a still, start and end frames, reference and motion control |
| UGC Studio | Creator looks, script to talking video, your product in the creator's hand |
| Audio Studio | Voiceovers, designed and cloned voices, music beds and sound effects |
| Flows | Chain steps on one canvas, run them in one click, cut a commercial in Director mode |
| Super Agent | Plans, creates on brand and publishes or schedules after your approval (Growth and Pro) |
The reason ad creative slows down is rarely the model. It is the loop. A designer generates ten variants in one tool. They drop the best three into Figma. Paid acquisition reviews them in Slack the next morning. Comments come back. The designer regenerates. Two days have gone by and the team has shipped one ad. For a DTC brand running a weekly test cadence, that loop is the bottleneck.
Flows collapses the loop. Every team member opens the same canvas. Generations land next to the brand brief, next to the reference images, next to the rejected variants from last week, next to the live comments from the founder. Decisions happen on the canvas instead of three tools deep in Slack threads.
Flows is the Figma for ad creative
The closest analogue is Figma. Designers know what it feels like to drop into a file, see the cursors of the other people in the room, leave a comment on a specific frame, and have someone else iterate next to you. Flows does the same thing for ad creative. The canvas is infinite. Cursors are live. Comments anchor to specific generations. Super Agent sits in the session and holds brand context across hours, so the team is generating off the same brand foundation rather than three different mental models.
When we last checked (2026), none of the standalone AI ad tools we benchmark against shipped a multiplayer infinite canvas. Creatify, Krea, Artlist Studio, Arcads, and MakeUGC each ran the creative session for one user at a time. They are useful as solo consoles. They are not where a team aligns on the campaign.
Who works on Flows together
A typical 7-figure DTC skincare team running a weekly test cadence looks like this in a session. The founder opens the canvas, drops in the brand brief, references three competitor ads, and writes a one-line creative direction. The designer joins, pulls in the product references saved in Brand DNA, and starts generating cinematic pack shots in different aspect ratios. The paid acquisition lead joins, comments on which variants look testable, and asks Super Agent for one more variation with a different background. Everyone sees every change as it happens. The decision to ship a five-variant test set is made on the canvas, not in Slack at 9pm.
Live comments anchor to specific generations
The single biggest difference from a Slack and Figma loop is that comments live on the canvas next to the asset they describe. The founder can say "this lighting is wrong" on the exact frame instead of on a screenshot in a thread. The designer can fix it and resolve the comment without context switching. The paid acquisition lead can comment "this is the variant I want to test on TikTok" and that comment stays next to the asset on the canvas when the variant moves into Director mode for assembly.
One credit pool, one workspace, one final cut
Collaboration breaks down when the team has to leave the workspace to finish the ad. Flows keeps everything in one session. Voice generation, voice cloning, AI music, the Audio Studio, AI UGC creators, Director mode for assembly, and exports in the aspect ratios TikTok, Reels, YouTube, and Shopify need all live next to the canvas. The whole team finishes the campaign from the same session, without exporting, re-importing, or paying separate subscriptions for the editor and the voice tool.
Brand fidelity stays consistent across the team
The risk with multiple people prompting is that the brand drifts. Different prompters write differently. Different agency partners optimize for different KPIs. Brand DNA holds your products, references, visual direction, tone and rules, and it is the persistent brand identity that every team member references. Super Agent holds the brand brief, brand voice, and brand visual rules in session memory. When the paid acquisition lead asks for "one more variation," the variation comes back on-brand because the agent is generating against the same foundation the designer used.
Marian scaled a client with this workflow
Marian (creatingadswithmarian.com) is the case study we point to for this exact pattern. Her client team used Avocado as the live canvas where the founder, designer, and Marian as paid acquisition lead aligned on the weekly creative drop. The qualitative outcome was that creative shipped faster and converted better at scale. Read the case study at creatingadswithmarian.com for the full account.
Honest tradeoff with dedicated design tools
We will not claim Flows replaces Figma for product design or replaces a full creative briefing tool like Frame.io for video review at studio scale. Figma is still the right place to design a homepage. Frame.io is still the right place to review a feature-length film with timestamped comments. What Flows does is take the lane in the middle that those tools were never built for: a live, multiplayer, AI-native canvas where ad creative is generated, reviewed, and finished in the same session. For a DTC team shipping weekly ad tests, the integration is the unlock.
Pricing and the team plan
Avocado starts at 19.99 euros per month on Intro. Flows is on every paid plan; Super Agent needs Growth or Pro. Pooled credits cover image, video, music, and voice. Commercial rights are included on every paid plan. For a small DTC team running collaborative ad creative on a weekly cadence, the single workspace replaces a Figma seat, a video editor subscription, a voice tool subscription, a music tool subscription, and the AI ad generator the team would otherwise pay for separately.
