Ideas are like bellybuttons

6 min readJul 30, 2020


Idea and a button

Most of our waking moments, our minds run around creating little sparks of our awesome ideas! From a little paper clip to taking us to mars — someone’s spark has got us here.

Ideas in themselves, however, are not enough, they need to be coupled with the right execution, timing, and in case of commercial products a good market fit. Without the the optimal execution no matter how great an idea, it would not realize its full potential — and this is what separates great executors from rest of couch ideators.

Here are a shortlist of few ideas I was involved with over the past decade, where my execution wasn’t up to the mark, and someone else did a kickass job at building a similar product.

Almost a blockbuster JS framework

A long time ago, I was working on a central UX team, and we often faced challenges around building common reusable controls. Typically the control needed to include some markup and accompanying javascript. Any team consuming our controls had to include markup in their pages, ensure they were loaded and available, and then include javascript which will initialize the controls to render content. Ensuring that markup and javascript were always in sync, and avoiding errors were challenging.

To reduce dependencies, we moved on to pure JS controls. However, that increased challenges for us, wherein we either had to copy markup as string — making creating markup very difficult, or create every element in javascript and DOM-append them, which made it difficult to get a handle on complete visual hierarchy.

I came up with an idea, that allowed us to continue building controls as we did writing HTML and TypeScript (or JavaScript). We parsed it through a custom build task, to transpile it to pure javascript controls. Each control inherited from a base Control class, and must implement a single method load. The base control class took care of rendering and handling DOM interactions, and managed dependencies. Controls were allowed to nest, and a parent may include a combination of native html markup and other controls. The base control, took care of rendering children in their DOM order.

Control Markup (HTML) + Control Logic (TypeScript) passed through a Build Task Magic, created a pure JavaScript control.
Control Markup (HTML) + Control Logic (TypeScript) passed through a Build Task Magic, created a pure JavaScript control.

Life was bliss — for a while — however soon enough, our priorities changed, and I moved on to some other project. My Awesome Blockbuster JS Framework was lost into oblivion. Later some smart folks at Facebook released ReactJS, and the world of JS frameworks was never the same again.

Anonymous corporate messaging

There was a time, not so long ago, when anonymous messaging caught on like wildfire. There were half a dozen apps that allowed messaging anonymously to people around us. Most of them turned into cesspool, and to some degree of relief they are all gone now — including the one we built.

Taking a cue from the first few apps, our small team thought of doing an experiment within the corporate environment. Employees of an organization would be able to message each other with complete anonymity still knowing that each one of them were a verified colleague.

We quickly build a prototype for internal testing. About 1500 employees were let in to try it out, that created endless amusement for us. We added more features — channels, topics, hashtags, avatars, search and more around usability of the app. We also introduced org administration and creating verified conversations, if a person chooses to.

The anonymous messaging app to talk to your colleagues.

All that said and done, we really didn’t think of a concrete go to market strategy. Being anonymous has its charm to it, however at the time all anonymous apps were wrecking havoc across their user base. Rampant abuse, bullying, threats, and very little moderation and control, led all such apps to become a social nightmare. We saw this in the message streams flowing through our app as well, despite of the fact that everyone participating was a colleague who we may closely work with. To quote — give someone a mask and they show their true face — would be a pretty accurate description.

Eventually, we gave up, and shuttered the service within its limited run of a half year. We planned and rejected a host of moderation, abuse prevention and active control to org admins, but nothing at the time seemed good enough to justify the risk of putting out yet another anonymous platform. In hindsight, though, some of the fears might have been misplaced, as evidenced by Blind that launched anonymous corporate messaging about the time we shut down, and quite recently introduced AMA with verified individuals!

Slack Trello hybrid

You love staying updated with your team on Slack, and you love managing and staying on top of your tasks on Trello, then you’d love them together in a single app. Yeah, that was our pitch when we built the hybrid — one tab apart.

Within a couple of weeks of burning midnight oil we came up with a converged UI for messaging and managing tasks, and built a seamless integration across those two services. We could create tasks about what we were talking about, and then talk about tasks that we have created. We built notifications left and right, and integrated with most services that Slack integrated with.

Communicate and maintain the project timeline in a seamless integrated app.

It was a cheeky idea, and very hard for us to see where we could take it. People we tested it with, were all amused, but didn’t — at the time — see much value in such integration. As a result, we decided to move on and focus our attention onto other experimental stuff. Quite recently a relatively new service has raised millions to build a service exactly like that — here’s wishing good luck to

Managing your Prescriptions

At one time, this was the favorite thought exercise we did with potential senior hires. “How do we make it easier for old age people to refill, and follow schedule for their prescriptions”. As we started defining the MVP we realized that taking prescriptions is a difficult problem for mostly everyone. Overall any system should meet these basic requirements:

  1. Ability to sort the prescription based on the medicines and their schedule.
  2. Dispense the right prescription at the right time.
  3. Automate refills with pharmacies

We threw technology to solve it — a high tech dispenser that used modularly packed prescription medicines, with embedded code that would allow the dispenser to sort and dispense the correct doses at the right time. The machine would connect with phone or other ambient smart device for notifications.

A magic machine to dispense your medicines — right dosage at the right time.

Thankfully, this was just a thought exercise with an over the top plan which, luckily for us, we didn’t really execute on. Years later, this problem was solved with a much simpler and impeccably crafted user experience to deliver prescriptions by PillPack — which packed individual dosages with timestamps, in order of their consumption that simplified the sorting and dosage problems for everyone with a very simple delivery.

Ideas are like bellybuttons everyone has them — what we have less is the right execution to bring them to reality.




Mostly stories documenting the lived experiences, stoking my self-conscious, introverted, sarcastic and borderline racist, middle-aged personality.