MoyAts
Designed an actually-good ATS for Ethiopia, then watched the subscription model ghost us like a bad Tinder date.
Built a whole system to revolutionize hiring... and the pricing strategy pulled a Houdini. Classic.
How a “Small Side Project” Turned Into a Full-Blown ATS, and My Crash Course in UX Chaos.
This project is my design origin story. You know, the “fell face-first into a flaming dumpster of complexity and came out stronger” kind. It was my first large-scale, in-house gig ... the deep end of the UX pool. No floaties. Just vibes.
It started when my cousin said,
“Hey, I’ve got some friends working on a small project. Could use your design eye.”
Me, being young and trusting:
“Sure, sounds fun!”
Plot twist: It was not small. It was a fully grown, six-foot-tall Applicant Tracking System (ATS) for Ethiopia. Think Indeed meets LinkedIn, but with the UX of a DMV website on dial-up. The stakes? Astronomical.
My knowledge? Very much in beta.
The Problem: When the Job Market Feels Like a Game of Minesweeper with Resumes Instead of Bombs ...
What we found was a flaming train-wreck of inefficiency:
You have to register before you can even look at a job posting.
Every company insists on its own app—because obviously, everyone wants 87 different ways to apply and fail.
Zero workflow flexibility. One-size-fits-none.
Recruiters were basically playing Minesweeper with a folder full of PDFs.
Bulk hiring? Only if you're into spreadsheets, stress, and screaming into the void.
Basically, we weren’t fixing a system, we were exorcising whatever demon built it.
The Reality Check: When Gmail + Excel Is Still the National Hiring Strategy
According to a 2020 labor assessment, only 12% of Ethiopian employers were using an ATS. The other 88%? Raw-dogging their recruitment process with Gmail, Excel, and misplaced hope.
So, we did what any rational person staring at this chaos would do: interviewed 5–8 candidates to get to the root of the dysfunction.
From their war stories, we extracted generalized user stories and crafted three painfully accurate personas, reflecting the lived agony of our target audience ... people aged 20 to 60, trying to survive the hiring process without burning out or blacking out.
Meet the trio:
Meron Tesfaye – A recruiter neck-deep in irrelevant CVs. Just wants a system that doesn’t make her cry before lunch.
Kaleab Eshetu – A design engineer somehow being matched with “Sales Manager” roles. Close to throwing hands (and his laptop).
Mr. Hagos Seyum – A seasoned HR pro clinging to software older than his interns. Manual entry? Hates it more than Mondays.
Conclusion? The system wasn’t “inefficient.” It was fundamentally hostile. And we were about to go full UX exorcist on it.
From Chaos to Clarity: The Not-So-Chill Game Plan That Turned Screams Into Systems
Interview. Prototype. Panic. Repeat.
We dove headfirst into deep user interviews (aka free therapy for frustrated job seekers and recruiters). Then, we did what all sane designers do: personified the chaos by building detailed personas ... because somehow the pain feels less existential when it has a LinkedIn photo.
With time not on our side and caffeine overdosing dangerously close, we skipped low-fidelity wireframes and went straight to hi-fi. Iteration wasn’t optional—it was survival.
Here’s what those glorious insights morphed into:
Centralized, customizable job posting – because no one wants to be the IT guy and the hiring manager.
A job application process that didn’t feel like a riddle from Saw.
Flexible workflows – because not all hiring processes are created by robots.
Actual communication tools (that aren’t just “We’ll circle back”).
Analytics with context – not just guilt-trippy pie charts yelling “DO BETTER.”
Turns out, when you listen to users, you build things they actually use. Wild, right?
Onboarding the Solution: Or, How I Became Besties with AI, Regret, and Improvisation
No local benchmarks. No competitors. Just vibes and prayer hands.
Since this was Ethiopia’s first real ATS (Applicant Tracking System), there wasn’t exactly a blueprint we could reverse-engineer. So we did what any resourceful designer with a deadline and no sleep would do: borrowed heavily from platforms like Catsone and Workable, then spiced them up to fit the business goals and user context.
In hindsight? Probably not one of our finest UX research moments. But when you’re launching headfirst into your first big project, you make do. Was it the most rigorous approach? Nope. Was it effective? Surprisingly, yes (in a way).
Moral of the story: when there's no path, sometimes you have to start by stealing the map… then redraw it mid-sprint. Lesson learned.
MoyAts v-1 (Catsone Wannabe)
MoyAts v-2 (Decency Redeemed)
MoyAts v-3 (Third Time is the Charm-ish)
First Usability Test (and Then the Second): When We Threw the MVP at the Wall and It Actually Stuck
Post-launch, it was time to see if this thing could walk, or at least crawl without bursting into flames.
We rolled out Version 1.0 and braced ourselves. Here’s what didn’t break (and actually worked better than expected):
✅ Reusable candidate archives (goodbye spreadsheet nightmares)
✅ End-to-end automation from sourcing to hiring
✅ AI resume parsing ... because yes, we let the robots judge for once
✅ Auto-scheduling interviews so no one had to play calendar Tetris
✅ A scoring system that let hiring managers use actual logic, not vibes.
The verdict? People used it. People liked it. No one rage-quit. And in UX terms, that’s basically a standing ovation.
Preview My Mental Breakdown …
The Afterparty Nobody Stayed For: When Beta Launched, Features Landed, and the Business Model Flatlined
Moyats made it to beta. ArifPay got onboarded. We had AI features queued up like we were launching the future of hiring. Things were happening…
But then came the ghosting.
Our shiny subscription model? It turned into a group Netflix account. One HR manager signed up, and suddenly their entire department, and their neighbor’s dog was logged in.
Where did it go wrong?
Our pricing? Too friendly. Like “I’ll cover the bill” energy.
Our assumptions? Naïve, optimistic, and clearly raised on motivational quotes.
Market research? More vibes than verified data.
Lesson learned: just because it works doesn’t mean it sells.
Lessons from the Monetization Dumpster Fire: A Beginner’s Guide to Burning Your Business Model with Confidence
So, what did we take away from watching our subscription model spontaneously combust in public? A masterclass in what not to do:
Market research? Turns out, it’s not just academic cosplay ... it’s survival.
Competitive analysis? Not just pitch deck filler. It’s how you avoid reinventing a flat tire.
User feedback > clever guesses. Every. Single. Time.
Winging it feels powerful… until your pricing model is on fire and you're holding the matches.
Next time? We're testing, pricing, and validating like rent depends on it. Because it kinda does.
Crafted with ❤️, fueled by a ton of coffee ☕ and not enough 💤... (Believe me)