2026 · MVP · ABC Empleos
A jobs platform for Paraguay
ABC Empleos, from zero to a live MVP in three months.
- Problem
- Azeta, the group behind ABC Color, wanted to turn the newspaper’s reach into a jobs business: a channel between readers looking for work and local companies and recruitment agencies.
- My role
- Lead and only designer, alongside a product manager and a developer. Both sides of the product, from recruiter interviews to launch.
- Outcome
- Live in production, with real applications from launch day and strong early numbers reported by the client.
Context
ABC Empleos was two-sided from the start. Candidates would arrive from the newspaper, mostly on their phones, without having chosen a jobs platform. Recruiters needed to find the right people among the applications. Neither side is worth much without the other.
I started with recruiters: interviews with six to ten of them, from in-house HR teams and agencies, then workshops and several rounds of prototyping. Two findings shaped the product. They discard on a few must-haves before reading anything else. And their real work happens in WhatsApp, email and spreadsheets.
Decisions
01
Knockout questions rank, never hide
Recruiters add questions to an offer and set how much each one weighs. Every applicant gets a match score, and the list is sorted by it. Nobody is filtered out: recruiters wanted to make the call themselves, candidates misread questions, and in thin applicant pools hiding people costs matches.
Instead of: Hard filters that remove anyone who fails a must-have.
02
Apply first, create the account after
Traffic from a newspaper is cold, and a signup wall on a phone loses people before they see the offer. So candidates apply first, then get the option to save what they answered. By then the value is obvious: with an account, the next application shrinks to the few questions specific to that offer.
Instead of: Signing up before applying, like most job boards.
03
Fit into the tools recruiters already use
Recruiters would keep their own tools whatever we shipped. So the platform’s job was to put ranked applicants in front of them, email them when new ones arrive, and make it easy to take candidates out. They still use other tools, as we expected.
Instead of: The whole hiring workflow, messaging included, inside the platform.
04
Ship the core, learn the rest
Messaging, paid plans, company pages, job alerts and a list of tools heavy recruiters asked for all waited. We would learn which ones mattered once real offers and real applications were moving.
Instead of: A first release that covered every request from the workshops.
Results
- Real applications from launch day.
- Recruiters said the match score saved them time, and companies asked to publish more offers.
- The client reported strong early numbers and extended our engagement to keep fine-tuning the platform.
Looking back
Agree on an owner before kickoff
Many of the problems we ran into weren’t product problems. They came from not knowing who on the client side owned the platform. Next time I’d make a named owner a condition for starting.
Instrument from day one
We know the early numbers were strong because the client told us, not because we could see the funnel ourselves. Analytics belong in the MVP.