I built a paid newsletter that curates real, relevant remote job opportunities for professionals, evolving through three iterations until it became a highly focused product for tech roles.
Iterations:
- Remote job board (one-time payment)
- Tu Próximo Rol, premium newsletter (general roles)
- Tu Próximo Rol Tech, premium newsletter for technology, data, and product roles
Problem solved: Helping people find the right job opportunity at the right moment, in a market where information is fragmented, low-quality, and time-sensitive.
Results
- +75 users paid for these iterations
- I shared more than 6800 jobs between April and December 2025
- +2500 jobs in 2024
1. Context: why this product existed alongside others
This product was intentionally different from my other offerings.
At different times, it coexisted with:
- The membership (community and belonging)
- The remote work program (skill-building and transformation)
This product was designed around a very specific job to be done: helping professionals discreetly monitor relevant opportunities until they are ready to act. Each iteration narrowed the job to reduce noise and increase relevance.
I wanted to build something that:
- People could buy quickly, without a long onboarding process
- Delivered value immediately
- Was tied to a moment of urgency (job search)
- Could be sustained with a repeatable system
This also solved an internal constraint: outcomes in career programs take time. A job discovery product could create faster “wins” because it puts opportunities in front of people weekly.
2. The core problem: job information is broken
The job search problem looks simple on the surface. “Just search online.” In practice, job discovery is broken because:
1) The information is not unified
Candidates waste time searching many places, repeatedly. Remote roles live across:
- Company career pages
- Multiple ATS pages
- LinkedIn listings
- Niche communities
- Country-specific boards
- Lists and aggregators with unclear sources
2) The match is hyper-granular
A job is not just “marketing” or “product”. Two people both “looking for remote product jobs” can have totally different definitions of what counts.
Real filters include:
- Seniority
- Region eligibility (global, LATAM, country-specific)
- Time zone overlap
- Contract type
- Language requirements
- Industry constraints
3) Remote roles expire fast:
Listings close quickly. Some are visible for days, not weeks. Freshness is not optional.
3. Insight: curation beats aggregation
I realized the product was:
- a trusted filter
- a consistent weekly habit
- a reduction of noise
- a shortcut to the few sources that reliably produce quality listings
Over time, I built a personal system:
- My private database of 959 trusted sources
- Notes on how many jobs I found in each
- Last time I checked the site
- Which sources are worth checking weekly
- I removed the sources that were misleading or too broad
I also decided on strict quality constraints:
- No agency listings
- Direct company links
- Clear labeling when a role had regional restrictions
- Always include the date of discovery
This shifted job discovery from “search” to “signal delivery
4. Iteration 1: the paid remote job board
The first iteration was a one-time paid product.
- Built in Airtable
- Well categorized
- Direct links to companies
- Updated manually
Why this made sense at the time
It was the fastest way to:
- package work I was already doing
- validate willingness to pay for curation
- ship without a complex tech build
What worked
- People paid because the database saved them time
- The value of a curated list was obvious
It was easy to communicate: “Here is the list”
The failure mode was structural:
1) Link decay: Jobs expire and ATS links change. The database degraded daily.
2) Maintenance cost
To keep quality high, I needed to:
– revisit links
– remove closed jobs
It was operationally expensive.
3) Value wasn’t personalized
A directory gives breadth. But relevance is personal. People would browse and still feel: “I can’t find the one that fits me.”
What I learned
– Static libraries don’t work well for opportunities that decay.
– This product required push distribution and constant freshness.
Framework: Build–Measure–Learn
Each iteration surfaced new constraints (link decay, personalization cost, role distribution), which informed the next build cycle.
5. Iteration 2: premium newsletter (Tu Próximo Rol)
The second iteration solved expiration and timing.
A newsletter solved the two core issues: Freshness and Delivery. Instead of asking people to go search, I delivered fresh links to them weekly.
The promise became: “I search for jobs for you.”
Every week:
- curate 150–200 listings
- verify they are direct company roles
- label restrictions when possible
- categorize them
- format and send
This required both research discipline and operational consistency.
This was high value but high cost:
- searching and verification takes time
- categorization takes time
- formatting three email versions takes time
- links and formatting errors become expensive at scale
Framework: Problem–Solution Fit
Early versions validated willingness to pay for curated jobs, but revealed misalignment between the breadth of access and perceived relevance. Iteration focused on tightening the problem before scaling the distribution.
Segmentation design
When someone subscribed, the welcome email included a short survey. Users selected their track.
Initially, I shipped 3 tracks:
- Creative: marketing, design, comms
- Tech: product, data, development
- Business and ops: finance, support, PM, account roles
Each week, I produced three versions of the same product.
What I discovered about the market
1) The market repeats roles
Most companies hire for common role families. “Unique” jobs are rare. This limits novelty.
2) Remote work in Spanish is extremely limited. The majority of quality remote roles require English. So a Spanish-language product has a structural ceiling.
3) Junior remote roles are rare. Remote work usually demands autonomy. This filters out junior candidates.
This created a mismatch between who wanted the product and who could realistically use it.
Learning: Segmentation helps, but it doesn’t fully solve hyper-specificity. To increase relevance, the product neededa narrower scope.
User feedback
I constantly kept running surveys, and these were the answers:
- Suitability of roles: Some users found the available options interesting or useful, while others noted a lack of opportunities in specific fields such as Architecture/Interior Design, Product Marketing, SAP B1 Junior Remote consulting, or Project Management for non-tech/architecture engineering companies.
- Mismatch with profile/experience: One respondent found roles close to their profile but needs clarity on translating their skills for remote or corporate options and is awaiting a response regarding a potential fit in Human Resources.
- Barriers to application: Some suitable roles required a bilingual level of language proficiency or required in-office attendance once a week, making them unfeasible for the respondent.
- Vacancy distribution: There are significantly more vacancies for UI/UX specializations compared to Graphic Design or Illustration roles.
6. Iteration 3: Paid newsletter Tu Próximo Rol Tech
The third iteration came from a simple observation. Roughly half of all quality remote roles I found were:
- technology
- product
- data
That meant:
- higher supply density
- more consistent demand
- more repeatable sourcing
So, I made a product decision: create a dedicated version for tech roles only, with the promise of at least 100 curated tech roles per week.
This did two things: strengthened differentiation and set an operational output target that enforced consistency.
How the product architecture changed
Instead of embedding links inside emails (which created formatting errors and heavy effort), I shifted to Airtable-based delivery.
I built two tables: the Tech roles table and the General roles table
Each role included:
- company
- role title
- main area and sub-area
- seniority
- link
This changed the production workflow:
- less formatting fragility
- less copy-paste failure
- easier categorization and reuse
Distribution logic
- Tech premium subscribers received the tech version
- General premium subscribers still received the general version
- Remote work program students received the broader version as part of the program value
This allowed one sourcing engine to serve multiple customer segments without duplicating work.
Narrowing the scope increased perceived value because:
- Relevance went up
- Users could see themselves in the listings more often
- The product became easier to describe and justify
List example
7. Operational excellence: building a sourcing engine
This product improved dramatically because the operation improved.
Time reduction
– Early workflow: 4+ hours to curate and assemble weekly jobs
– Current workflow: about 90 minutes
That efficiency gain came from systems
Source database as a product component
I tracked:
- When I last reviewed each source
- How many relevant jobs did it produce?
- How often has it produced roles that fit my audience
- Whether it drifted into low-quality listings
This turned sourcing into a measurable pipeline.
Tooling and ATS pattern recognition
I introduced:
- scraping tools (Simple Scraper)
- AI-assisted categorization
- SEO tools for scraping company career sites (Screaming Frog)
I realized most companies fall into a small set of publishing patterns:
- custom career pages
- Notion pages
- common ATS tools (Lever, Greenhouse, Ashby, BambooHR, etc.)
Once you understand these patterns, you can:
- search faster
- extract faster
- verify faster
This is where the work shifts from “manual browsing” to a repeatable system.
8. Business model insight: incentives define information quality
A major insight was that job boards are often low-quality by design.
If a directory earns money from ads, affiliates, and SEO traffic, then the incentive is to get more pages, more clicks. Accuracy becomes optional.
My model flipped the incentive. If users pay me, then accuracy matters, freshness matters, and relevance matters
People were not paying for “more jobs”. They were paying for better signal, fewer dead ends, less wasted tim,e and trust.
9. User acquisition: building demand week after week
Getting users for this product was a continuous system.
Framework: AARRR (Acquisition, Activation, Retention)
The product relied on continuous organic acquisition, fast activation through immediate job delivery, and retention through weekly habit formation.
Because the value was time-sensitive and recurring, I had to generate demand every single week, for almost two years.
I used two main acquisition engines, both fully organic.
Pain-driven short-form content (Instagram + automation)
The first channel was short-form video.
I recorded a series of videos focused on very specific job-search pains, not on the product itself.
Each video started with a different hook, for example:
- Wanting to change jobs without your manager finding out
- Knowing you want to leave, but not being ready yet
- Starting a job search three or six months in advance
- Feeling late to opportunities because you find them too late
- Being overwhelmed by low-quality job boards
These videos were designed to make people think:
“This is exactly what I’m going through.”
Distribution mechanics:
- Published on Instagram
- Call to action was always to comment
- I asked for their email in the replies
- Used ManyChat to capture emails automatically
- Sent a follow-up email sequence explaining the product and its value
This worked because:
- The pain was immediate and emotional
- The ask was low-friction
- The product was positioned as a solution to a private, uncomfortable problem
Scarcity and transparency on LinkedIn and Instagram
The second engine was direct signaling.
Every week, I publicly announced:
- That I was about to send a new email
- How many jobs it would include (e.g. “150 remote roles”)
- When it would be sent (e.g. “in 12 hours”)
- That it was only for subscribers
Example positioning: “I’m about to send a newsletter with 150 remote jobs.
It goes out in 12 hours. Only subscribers get it.”
This did two things:
- Made the value concrete
- Created urgency around timing, not discounts
This approach worked especially well on LinkedIn, where:
- People were already thinking about career moves
- Public signaling felt credible
- Social proof built over time through repetition
Promotional experiments and pricing incentives
I also tested limited-time incentives, such as:
- One free month during Cyber Monday / Black Friday
- Short promotional windows tied to specific moments
These experiments helped:
- Reduce friction for first-time buyers
- Increase volume during specific weeks
- Validate that people needed to experience the product to understand its value
Consistency as a growth lever
The most important growth factor was consistency.
For nearly two years:
- I curated and sent the newsletter weekly
- I promoted it weekly
- I talked about the pain weekly
- I showed what was going out before it went out
There was no single viral moment. Growth came from repetition, trust, showing up, and delivering every week without breaks
Designing acquisition around timing, privacy, and relevance was just as important as building the curation engine itself.
10. Extra: Data report
I was analyzing so much data that I had unique information about the market that’s hard to get. I did this data report analyzing patterns from 45 real job descriptions.
11. Key learnings
- Freshness can be a product feature
- Curation scales better than aggregation when incentives align
- Narrowing the scope improves relevance and perceived value
- Delivery matters: push beats pull in job discovery
- Operational excellence compounds over time
- The market sets structural limits (language, junior roles)
12. Conclusion
Tu Próximo Rol became a real product when I stopped thinking of it as “a list of jobs” and started treating it as a curation engine.
The product evolved by reacting to real constraints:
- expiration
- relevance
- language and seniority ceilings
- operational cost
By narrowing the scope and building infrastructure, I turned it into a system that delivers trust weekly and improves over time, and creates real leverage from consistent execution.
Thanks for reading!