The Question I Ask Before Building Anything

Skip to main content

Mary's Blog | A design engineer's journal

Mary's Blog | A design engineer's journal

Command Palette

Search for a command to run...

The Question I Ask Before Building Anything

You might need to ask yourself too

UpdatedJuly 10, 2026

•2 min read•View as Markdown

The Question I Ask Before Building Anything

Design Engineer | WTM Ambassador

Comments (1)

Join the discussion

Julian Neagu19d ago

AI makes shipping easier, but it also makes it easier to build things nobody needs. Starting with the user problem saves a lot of time.

Mary Ojo15d ago

Yes! The user problem tells you exactly what to build

More from this blog

3 UX mistakes that might be costing you users

Let's talk about usability and 3 common mistakes you might be making when building that product. Confusing 'Account Setup' with 'Onboarding': Your onboarding flow is meant to serve a primary purpose,

Jun 28, 2026·5 min read·14

3 UX mistakes that might be costing you users

Beyond Alt Texts: How AI can Help Solve the "Cognitive Gap" in Modern UX

Last year during Google's Built-In AI hackathon, I built an experimental Chrome extension called Clarity Lens. The product was inspired by thinking about the 'hidden 16%', the portion of the world's p

Jun 2, 2026·3 min read

Beyond Alt Texts: How AI can Help Solve the "Cognitive Gap" in Modern UX

That’s AI? Intelligence in Everyday Interfaces

If you've used Youtube long enough, you'd realize it gradually became better in understanding your needs. I'm sure you can relate to this - you think of something, next thing, you open Youtube and it's being advertised, or you come across it on your ...

Jan 3, 2026·3 min read·13

That’s AI? Intelligence in Everyday Interfaces

Should We Really Build Carousels Without JS?

Trade-offs, accessibility concerns, and what CSS alone can’t solve

Jan 3, 2026·6 min read·36

Should We Really Build Carousels Without JS?

M

Mary's Blog | A design engineer's journal

23 posts

Everything I know about design and engineering

Let's settle this once and for all: the progress of a product isn't measured by the number of features it has.

Why?

Because users rarely care about features. They care about getting things done. They care about outcomes.

If you were to announce any of these updates:

An AI search feature was added.
Dark mode is now available.
We've launched a new dashboard.

These sound exciting, but mainly to the person/team that built them, they don't mean much to users on their own.

What users really want is to:

Find what they're looking for faster.
Better understand their business.
Save time and complete their tasks faster and easier.

A feature is only valuable if it creates a meaningful outcome. That's why I try to build (and sell) outcomes, not features. So whenever I'm building something, I ask myself:

What outcome does this feature create? Which user goal does it help achieve?

If I can't answer these questions clearly, it's usually a sign that the feature needs more thought.

As AI makes product development even easier, the focus has changed from just writing code. The builders who stand out and get the most out of products will be those who combine product thinking, design, engineering and AI to solve real problems.

Next week, I'll share how applying this outcome-based approach increased how users engage in one of the products I'm currently building.

See you next time!

I'm Mary, a product engineer, who designs and builds usable, accessible and intelligent products

#product-thinking #product-development #ux #software-development #ai

59 views

© 2026 Mary's Blog | A design engineer's journal

Archive
Privacy
Terms