4.5-5 out of 5 for the highest-rated page in testing, and BeyondCampus launched in the U.S. three months after kickoff.
Four pages, two markets, and a brand that was not in the scope.
The company Korea's developers are hired through, looking for a second market.
Grepp is a South Korean technology company that sits across the whole life cycle of a software engineering career. Founded in Seoul in 2014, it runs Programmers, Korea's first combined developer education, assessment and recruitment service, alongside Monito, an online proctoring product used for coding tests and certification exams. More than 420,000 developers had taken a coding test on the platform by late 2021, and the company's assessment business serves well over a thousand Korean employers, from the country's largest conglomerates to its best known technology startups.
It is an engineering company, and it shows in both directions. Grepp was founded and is run by people who taught technology themselves, and the platform behind BeyondCampus, a web IDE, coding tests, code review, Q&A and live video classrooms, was built in-house. The front end was the thin part. The version that existed when WANDR arrived was essentially a grid of cards a user could click into to see what a course contained, with live coding, cohorts and skill verification still on the roadmap rather than on the screen.
In late 2021 Grepp set its sights on the United States, and BeyondCampus was the product it was taking there. The company had already tried and failed to hire a product design team in the U.S. Samuel Lim, Grepp's co-CEO, reached out to WANDR directly in December 2021. What Grepp needed was not a translated website. It needed a front end built for a different audience, a different culture, and a considerably more crowded market.
Strong engineering, a crowded market, and nothing on screen to prove either.
This was not a hard design problem, and framing it as one would be dishonest. It was a clear one, with five constraints that had to be respected at once.
Grepp's engineering was genuinely differentiated: a browser-based IDE, code review, Q&A and live classrooms sat behind the product. None of that was legible from the outside. The existing interface showed cards and course contents, which is what every competitor showed, so the one thing BeyondCampus had that the others did not was invisible at exactly the moment a visitor was deciding.
The U.S. online education market is crowded and unforgiving. BeyondCampus would be landing beside Udemy, Coursera, Codecademy, Skillshare and Maven, every one of them with an established audience, a known name, and a marketing budget Grepp could not match. A new entrant with no brand recognition has one chance to give a reason to believe, and it happens above the fold.
The target users, computer science students and junior developers, are unusually discerning consumers of software. They are digital natives, they research before they buy, they judge interface quality instinctively, and they leave in seconds when a value proposition is not immediately legible. One participant summed up the barrier: “When it comes to online learning, it is very hard for me to click on the ‘pay’ button. I am always weary if the course will be any good.” A generic education site would not survive this audience.
BeyondCampus was built on community and cohort-based learning, and Grepp expected that to be the hook. The research said otherwise. Every interviewee recognised community immediately, and every one of them treated it as a given rather than a delighter, because Stack Overflow and Reddit already supply it for free. Cohorts also carried a cost these users felt keenly: senior developers had no room in their schedules for a fixed timetable at all.
The plan was for Grepp's internal team in Korea to supply brand assets for BeyondCampus. They were delayed, and what did exist was built for Korean expectations rather than American ones. With a launch date fixed and a timeline already tight after the failed U.S. hiring attempt, the project needed a visual identity that did not exist yet, and needed it without moving the date.
Drag the handle to see how the landing experience changed, from a page that listed what the platform could do into one that translates every feature into the outcome a developer actually wants.
The research round was small on purpose, because the timeline left no room for a full cycle of interview rounds. WANDR interviewed Grepp's stakeholders first, then ran six in-depth interviews across three deliberately different career stages, computer science students, junior developers with nought to two years of experience, and senior developers with five years or more, to find out where the real ideal customer profile sat. Every session was a scripted, conversational Zoom interview about how these developers actually find, judge and buy learning, run alongside secondary research into cohort-based learning outcomes and a section-by-section teardown of the platforms they already use.
Senior developers liked the idea and had no room for it. “Live courses provide the discipline, but with a full time job and a new born, there is no way I can make time for it right now,” one told us. Students and junior developers would commit to a cohort, but only for genuinely complex topics; for a problem at work they go to Google or Stack Overflow first. That single finding moved seniors out of the primary audience and pointed the whole site at students, junior developers and career switchers.
We took 23 sites apart across three groups: cohort-based course platforms including Maven, Section4, Reforge and Bloom Institute of Technology, MOOCs including Coursera, Udemy, edX and Codecademy, and course-creation tools including Disco, Teachfloor and Airtribe. Mapping the landing page structures showed the pattern: a hero, a feature showcase and a footer on 23 of 23 sites, testimonials and a closing call to action on 18, a logo cloud on 16, featured courses on 14. Maven won the team's vote, and it won it on one thing, translating features into benefits.
Self-paced online courses are completed by as few as 3% of the people who start them, while cohort-based courses routinely clear 90%, and the research is consistent about why: 85% of participants in one study said being part of an online community helped them learn, and 94% said cohort discussions increased their interest in the material. That gap became the narrative spine of the page, because it is the one claim a skeptical developer cannot argue with.
Final-year and freshly graduated CS students who want the latest and greatest to launch a career with. Price sensitive, portfolio driven, and looking for a project big enough to show off. “The courses I enjoy the most are the ones where every assignment is part of a bigger project I can show off to my friends.”
Nought to two years in, learning for advancement or mid-pivot into engineering. Will commit to a cohort for a complex topic if the time cost is bounded and mentors are reachable. “If it only lasts a couple of hours and I have access to mentors and can ask questions in that period, that is great.”
Five or more years of experience, interested in the model and unable to fit it in. Prefers documentation and self-paced video that can be sped up, and researched the strongest option for a specific use case rather than following a recommendation.
Translate every feature into a benefit. The original site listed what the platform could do. The new one sells what a developer gets, on the model of the iPod launching as a thousand songs in your pocket rather than a gigabyte of storage.
Lead with the completion gap. The 3%-versus-90% difference between self-paced and cohort-based courses is the single hardest fact in the category, and it belongs where doubt peaks rather than buried in a features section.
Design like a developer would. Handwritten accents, hand-drawn arrows and small references to coding culture ran through the pages, because credibility with this audience is earned in the details. “There is a lot of small references to the coding culture and things we use in tech, that is very nice,” one tester said.
Over-inform on the course page. Everywhere else less is more, but on the page where a developer decides to pay, there is no such thing as too much detail: curriculum, structure, time per section, instructor, and outcome.
Grepp came in believing community was the reason to choose BeyondCampus. Every single interviewee recognised the community theme, and every one of them treated it as the price of entry rather than a reason to switch, because free communities already do that job. The thing that actually moved people was narrower and more concrete: writing code alongside other people, getting it reviewed, and being unstuck by a mentor in the moment. Community is a category; co-programming is a mechanism. In a market this crowded, the mechanism is what a skeptical buyer can evaluate, and the category is just what everyone claims. It also happened to be the thing Grepp's engineers had actually built and the competition had not. The research did not change the product. It changed what the product said about itself.
The engagement closed with the four core pages designed, tested and documented as a design system, a U.S. brand direction that had not existed at kickoff, and a research base Grepp's team could keep building against. Grepp announced the public launch of BeyondCampus on 13 June 2022, three months after kickoff, and carried WANDR into discussions for the subsequent UI and development phases.
They have a lot of experience in strategically and systematically doing design work.
[They organize] everything in a set of pipelines, which allows them to work very quickly.
