How multicardz Breaks 2D Limits with a Multidimensional Data UI

In 2017 I was trying to use Trello to manage product/feature/release/dev tasks planning after having tried all of the other offerings. There were a lot of things I liked about it, simple, fast to use, no imposed structure. But as much as I wanted to, I could not use it because product management is multidimensional, sometimes you want to see dev tasks by feature, other times by release or by month, or by completed, etc. All of the products locked me into a two-dimensional structure, so I just kept waiting and waiting for someone to make the right tool until I got tired of waiting this year and wrote it myself. Along the way, it transformed into something much more amazing than I thought it would be. I started chasing performance. For anyone who knows what Lighthouse scores are, I got to 100 performance, 96 accessibility (it does not like my earth tones), 100 best practices, and 100 SEO. Then I decided I wanted to use it to match my emails to my spreadsheets, so I engineered a solution that would allow me to do any operation over the entire data set in less than a half a second round trip to the browser, even for over 1 million cards (I have a LOT of archived email). I also came up with an entirely novel UI/UX that is like no-code but for search queries. It allows you to discover complex data relationships in only three drag and drops. Drag. Drop. Discover. The whole process has been very rewarding, and I am very proud of what the product has turned into.

The most important tactic was the most obvious one, get user feedback. I released and listened to all of the feedback on all of the ways that the product did not work. This came down to two things, not intuitive on first glance, and lack of tutorials to explain how it worked. This gave me early input with time to fix things while there were still no paying customers on the site. I am just this week releasing the fixed version. What flopped was the innovative UI/UX. People just found it too complicated and non-intuitive. I was in love with its "simple elegance," but I had to substantially change it so that it was written for other people, not for myself. I also had to add AI so that people could just explain what they wanted and have the app set itself up for them. This turned out to be easier than I thought except for one thing, guarding against abuse. That is an arms race that might never be over.

Apart from the complex UI and lack of AI, the biggest complaints were about lack of support, such as tutorial videos. So I'm now building tutorial videos and in-app help to fix exactly that. AI has been a tremendous help. Because I had built extensive UI testing in Playwright, I was able to use those tests to have AI build the tutorial much faster and better than if I had to record each one myself. I also used text-to-voice to narrate the videos, but I will change that as soon as I can. I think there is something superior in a real human voice. Finally, having AI write the scripts is a mixed success. I have to completely rewrite each one myself, but it is easier than starting from a blank page. All in all, I am lucky to have had such success; the complaints were few and very specifically targeted.

Everyone just accepts "2D prison," nobody knows that there is a way to break out of it. We do not have to live this way. multicardz is the proof. People need to be able to see different dimensions of the same thing in every business, real estate agents, lawyers, media managers. Almost nobody can see all of the info they need in a rigid two-dimensional grid. Real estate agents need to see their inventory by many dimensions, listing readiness, time on market, sales stage, open house dates, property type, target buyer, and many, many more. Media managers need to see assets by readiness, campaign, target market, success, cost, and a host of other factors. Lawyers need to see by client, court date, strategy, readiness, etc. Everyone needs to see a calendar view. How do I know this? Because I have shown it to real estate agents, a lawyer, and some media managers. The pattern is always the same when I demo, about 5 minutes of total confusion, followed by a big AHA! and enthusiasm for all of the potential.

Working hard to get as many users as possible. To get there, we have a very generous free tier with unlimited collaborators and guaranteed grandfathered pricing for early adopters. But since this is my first consumer product, I have a lot to learn. I will leverage my X presence as much as I can, but to be honest, the opportunity to be listed on SaaS Browser really came at just the right moment for me. I have high hopes for it as a channel for multicardz to be discovered.

Every card operation (searching, filtering, moving) takes less than a second even with over a million cards. The website has an auditable proof of this for the techies out there.

This is my first consumer product after 40 years of building large systems for enterprise use. It was a liberating experience; I finally got to build everything the "right way" with no compromises. The results are amazing. I have rigorous testing; every line of code is personally written and rock solid. What has been most challenging is that I was creating a product for myself at first. And although my needs were the same as many other people's, my level of "tech savvy" was much higher. I really had to work hard to make the product something that anyone and everyone would want to use.

I made the UI too powerful and too complex at first. The real problem was that I built it for myself, and I am very tech-savvy, so what felt simple and elegant to me was a brick wall to everyone else. I was so in love with the novel design that I couldn't see it. It took watching real people use it, confused and staring at the screen for minutes, before it finally sank in. I lost three months that way, chasing power and elegance when I should have been chasing clarity. The lesson was simple but it hurt, build for the person using the product, not the person building it. Now I only add power once the simple path already works.