Mode breakdown
A useful test, not a novelty demo
Ronny Hermosa’s video is built around a practical problem many creators and small teams recognize: Canva work can start as a simple campaign idea and quickly turn into a stack of repetitive tasks. You need templates. You need copy adapted for the offer. You need the right files organized. Then you need the posts scheduled. Hermosa’s experiment with Claude Cowork and the Canva connector asks a straightforward question: how much of that chain can be handed off without losing control of the final result?
The short answer is that Claude can do a meaningful amount of the heavy lifting. In the workflow Hermosa shows, it can update a template-based campaign, help break a multi-page Canva document into separate single-page designs, and even move through scheduling. But the video also makes clear that this is not a fully hands-off system. The strongest result comes from pairing Claude’s speed with a human eye for layout, visual consistency, and final checks.
Start with the connector, then give Claude real context
One of the most practical parts of the video comes before Claude does any creative work at all. Hermosa shows that the Canva connector has to be active, and he points viewers to the settings area where it can be connected if it is not already available. That may sound basic, but it is the kind of setup detail that often decides whether a workflow feels smooth or frustrating.
He also makes a point of using Claude inside a project that already contains business context, instructions, and memory. That matters because his prompt is intentionally short. Instead of explaining the whole membership campaign from scratch, he can ask Claude to update a specific Canva design and rely on the project context to fill in the rest. For readers trying this kind of workflow themselves, that is one of the clearest lessons in the video: AI works better when the tool is not guessing what your brand, offer, or tone should be.
Hermosa’s campaign example is a launch for a Canva community, and he uses a set of templates to build out multiple pieces of content around it. The workflow is not about inventing a design system from nothing. It is about taking existing visual assets and letting Claude adapt them for a specific campaign.
Where Claude actually saves time inside Canva
The first clear time-saver is the template adaptation step. Hermosa loads multiple templates into a single Canva document, gives the file a campaign name, places it in a folder, and then asks Claude to update the design for the offer he wants to promote. Claude returns questions about the offer, the call to action, and whether there is a time frame or code involved. Once those are answered, it starts modifying the template content.
What stands out here is not that Claude replaces design judgment. It does not. Instead, it handles the early content pass that usually eats time: swapping campaign copy, aligning the message with the offer, and preparing a draft that is close enough to review rather than start from zero. Hermosa notes that the copy lands well overall, even though he still makes manual adjustments to text boxes and alignment.
That balance is important. The video is strongest when it treats AI as a draft assistant rather than a finish line. Claude handles the repetitive and text-heavy parts, but the creator still decides whether the visual arrangement needs to change. Hermosa even points out that the tool flags certain image choices, which reinforces the idea that the workflow is part automation and part review.
Why the file-splitting step matters more than it seems
The most useful workflow insight in the video is not the initial design update. It is the reason Hermosa has Claude split one Canva file into several separate designs.
He demonstrates a limitation in Canva’s scheduling flow: if you try to schedule different pages from the same multi-page document as separate posts, the planner can become awkward to manage when the file is edited later. In his example, one scheduled post does not behave the way he expects once the design is changed again. That leads him to a practical conclusion: if a campaign is meant to become multiple posts, it is cleaner to break the original document into one-page files before scheduling.
Claude helps with that too. Hermosa asks it to open the design and create separate Canva documents, one page per design, while keeping everything else the same. He then asks it to move those new files into a folder for organization. The point is less about the novelty of having an AI perform the split and more about the workflow discipline it encourages. The campaign becomes easier to manage when each post lives in its own file.
For anyone using Canva for content calendars, that is a real operational takeaway. Multi-page designs are convenient for drafting, but they can create friction when each page needs its own publication path. Hermosa shows a way around that problem without rebuilding the campaign from scratch.
Scheduling is possible, but supervision still matters
The final stretch of the video is the most revealing. Hermosa asks Claude to schedule all of the new documents, and the tool opens a browser, requests access, and begins working through the content planner. It is not a one-click magic trick. Claude has to inspect the browser, interact with Canva, and handle each post individually. There is visible back-and-forth before it gets the process right.
That hesitation is part of the value of the video. It shows what current automation looks like when it is actually doing work inside a live account rather than just describing a workflow on paper. Claude can create captions, assign dates, and move through the planner, but it is still acting like an assisted operator rather than an autonomous content manager.
Hermosa’s final result is encouraging, though not flawless. He ends with five scheduled posts and captions tied to the campaign. At the same time, one scheduled design does not match what he had manually adjusted earlier, which is a good reminder that automation can carry a workflow far without removing the need for final review. The practical takeaway is not to trust everything blindly, but to treat Claude as a fast production layer that still needs a human sign-off.
What this workflow is really good for
The best use case here is not one-off design inspiration. It is repeatable campaign production. If you regularly build social content from templates, juggle multiple versions of a post, or spend too much time reorganizing files before publishing, this Claude and Canva workflow points to a real efficiency gain.
Hermosa’s approach works best when three things are already in place: a clear campaign goal, reusable templates, and enough project context for Claude to understand the offer. Without those, the tool would need far more explanation, and the workflow would slow down. With them, Claude can handle the parts of Canva work that are tedious rather than creative: adapting copy, duplicating assets, splitting posts, and moving through scheduling steps.
That makes the video useful in a very Mode Digital Media kind of way. It is not about treating AI as a replacement for design work. It is about identifying the boring, repetitive parts of the process and seeing how far the tool can go before judgment and polish need to take over.
Recommended next
Products & tools
One Mode Digital Media product and one relevant affiliate recommendation selected for this page.
Perpetually Online
A collegiate-style internet culture design for people who practically live on the web. “Perpetually Online” turns constant connectivity into a tongue-in-cheek badge of honor for creators, gamers, streamers, social-media regulars, and professional tab keepers.
Artlist
Music, footage and creative assets for video-first creators.