Late to the web
The top-of-mind printed classified ads of Brazil had a very weak presence on the web. I was then invited to kickstart a proper transfer to the web with excellence. Getting late to the market, a characteristic of the Globo group, we had to compete with some well-established competitors. So, over the already existent (old) systems, we would have to add all of the regular features of a modern classified ad for the web, such as maps, photos/videos, message system between buyer and seller, as well as digital handshakes.
Our differential was that every ad online was new and fresh, as we detected that one big flaw of the competitors was their lack of control over what was sold or not.
Besides the regular responsive design of a whole front-end system for the ads, we had to face all differences between the verticals (Real Estate, Cars, Business/Jobs, and Services/Products) on the back-end. I know it is a cliché, but this was a very clear example of the iceberg metaphor. Although the front-end itself was huge, the back-end was like 4 to 5 times larger in terms of developing and designing. Not counting the many service touch-points to cover like phone ads and handwritten ads — as O Globo newspaper still had some walk-in shops at the time.
For the Real State vertical, we had to create a plugin to the system so we could import all the data from the realtors, as they already had photos and lots of additional information of the items they were selling in their own systems.
For cars segment, there was a company that had practically the monopoly of the car dealerships databases, so we had to create a direct flow with them to populate our car ads. And for the other two, we had to create a token system to link the physical/analogous ad to the seller login, so they can enrich their ad once online. And we even had to create all the back administration of this whole service system, with some other very intrinsic layers such as payment monitoring, CRM, sales commission, and an internal dashboard personalized for each persona logged in. And, of course, we had to create mobile apps for the buyers, for both iOS and Android.
Within the firsts months, I made some researches with users of the classifieds, benchmarking, and started to plan what we were going to need to start developing it, such as dimensioning the team, positions, and tasks to the backlog with the web PO. I did all the wireframes for the web client side, co-created the responsive differences with the front-enders and supervised it, and did the same with the back-end admin, sharing it with other UX designer in my team and then compounding it with the whole system. I also created the mobile apps within a little squad formed by me as the UX/PO, a visual designer and a senior developer, in a very agile workflow — later known as ‘braided design’, surrounded by stakeholder of all kind such as data scientists, product managers, and executive directors, to mention some. The project was a great success at the time, reaching the desired mark of 1M (Brazillian Reais) of monthly revenue by the middle of the 2nd year.
The project took so many effort and time that by the mark of 2 years, we felt that we had so many changes, so many new features were added from the initials wireframes, that a front-end redesign was needed. So we build the new one with Angular in about 3 months, making it so much faster, and ranking the site way better, especially abiding the Google mobile indexing rules that was just put in practice, raising us above the initial competitors.