Why I Care More About How Software Feels Than What It Does

The Feature That Technically Worked

A few months ago, we finished building a new form flow. Everything functioned. Fields validated. Data saved. No bugs.

I opened it, filled in two fields, and said: "This doesn't feel right."

The form worked. It just felt like homework. Too many fields visible at once, a save button too far from where my eyes went, labels that made me think instead of just act.

We changed it. Trimmed what wasn't essential, put the primary action where your thumb already was. Same data got captured. But now it felt fast instead of tedious.

If software makes you pause and think about the interface, the interface is wrong.

"It Works" Is a Low Bar

Most software works. The question I care about is: does it feel like something you'd actually want to open?

Some tools have incredible feature sets — dashboards, automations, integrations with everything. I avoid opening them. They feel heavy. Every interaction takes a beat too long.

Then there are tools I open without thinking. They don't necessarily do more. They just feel lighter. Like the software is working with me instead of waiting for me to figure it out.

That's the gap I obsess over. Not "can it do X" but "does doing X feel effortless."

Things We Changed Because They Didn't Feel Right

How Kit responds to vague requests. Early on, if you told Kit something ambiguous — like "check on the Meridian thing" — it would ask you to clarify with a formal prompt. Correct behavior. But it felt robotic. We rewrote those responses to sound like a person trying to help, not a system that needs exact input. Same clarification, completely different feeling.

The morning briefing. First version packed everything into one screen — follow-ups, deals, meetings, tasks. Every piece was useful. But opening it felt overwhelming. We broke it into sections with clear priority ordering. Now it feels like a colleague giving you a quick rundown instead of a dashboard screaming at you.

Form flows for logging activities. Logging a call required selecting the contact, then the deal, then the activity type, then adding notes. Four steps. We collapsed it into one conversational input. Say what happened and Kit handles the rest.

Why This Matters

Here's what actually happens: a founder signs up for a CRM. They have fifteen minutes between meetings. They open it. If it feels easy, they log their deal and move on. If it feels heavy, they close the tab and tell themselves they'll come back later. They won't.

CRM adoption doesn't fail because of missing features. It fails because of accumulated friction. A form that takes one beat too long. A response from the AI that sounds mechanical instead of helpful. None of these are bugs. All of them are reasons someone stops opening your product.

Every feature we ship goes through a simple gut check: would I want to use this at 4pm on a Thursday when I'm tired and distracted?

We'd rather ship fewer features that feel right than more features that technically work. Every time.

Software should feel like it's on your side. That's the whole philosophy.