To the ClickUp Brain team,

I want to share my experience using Brain to set up my workspace, because I think it's exactly the kind of real-world case your team needs to hear about.

Some context: I'm a solo business owner launching a commercial fleet fuel delivery company in New Jersey. I'm not a technical person, and I have a demanding full-time job, so I can only give my business 30 to 60 minutes a day. I used Brain to build a 432-task workspace — 17 lists, custom fields, six filtered views, and a dashboard. It works now, and I'm glad I have it. But getting there took roughly nine sessions spread across two full weeks, and it should have taken one or two. Those two weeks were essentially all the business-building time I had, and most of it went to fighting the tool instead of building my company.

Here's specifically where Brain fell short.

First, Brain gave me confident instructions that were simply wrong. It told me to "Select All" statuses on views that had intentional filters, which broke three views I then had to spend a session restoring. It told me that references to a deleted custom field would automatically re-point to the replacement field; they didn't, and two views broke silently. As a beginner, confident-but-wrong advice is worse than no advice, because I have no way to tell the difference.

Second, Brain contradicted itself and worked from stale information. Its own audit told me duplicate views needed deleting; days later it insisted "there are no duplicates" while my screen showed two pairs of them. More than once it gave me instructions based on an old screenshot, for a screen I was no longer on — including telling me to delete a view that, by that point, was the only good copy. If I had followed that instruction blindly, I would have destroyed working configuration.

Third, Brain cannot show me where anything is. It can read my screenshots, but it can't point, highlight, or circle. Instructions like "click the funnel icon in the top right" mean nothing when the icon isn't where it's described — and several times it wasn't there at all, because the interface varies by card type and view. I ended up needing a second AI service just to translate Brain's instructions into actual locations on my screen. Your customers should not need two AIs to use one product.

Fourth, Brain's capabilities are inconsistent in ways that aren't communicated up front. It can create views and fields, but it can't delete or rename them, can't build dashboards, can't reorder the sidebar, and couldn't pull view IDs when we needed to identify duplicates. Every task turned into "Brain does half, I do the other half blind."

Fifth, the platform itself repeatedly fought me, and Brain didn't warn me about any of it. The CSV importer created 17 duplicate empty lists and ignored my Space-level statuses. The "Show closed tasks" setting doesn't persist on the Unlimited plan, so we had to invent a filter workaround just to keep completed tasks visible — a basic requirement for anyone running a checklist. Board views silently fail when grouping by a multi-select field, showing a cryptic "Private Field" warning instead of a plain explanation. Dashboard cards hide their filter settings in different places depending on card type. Each of these cost me real time, and each was discovered by breaking something first.

In fairness, here is what Brain did well: it owned its mistakes when I challenged it, its CSV import structure was accurate on the first try, and when I demanded item-by-item PASS/FAIL verification, its audits were thorough and honest. The underlying capability is real.

My suggestions, from the user's side of the screen: Brain should verify against the user's current screen before giving instructions rather than relying on stale screenshots; it should say "I'm not sure" instead of guessing with confidence; it should know and disclose its own limitations before starting a task, not midway through; and it badly needs some way to visually indicate where things are on screen — even a simple highlighted screenshot would have saved me hours.

I'm writing this because I stuck it out, but I'm certain most non-technical users in my position would have quit and cancelled. There's a great product in here. Right now it requires too much persistence to reach it.

Thank you for reading.

Richard Ravaioli

Founder, Deliver'd Mobile Fuel & Energy LLC