First experience return

My business directory website (if you haven't read how I built it, please go here) named Gougle.nc (yes, I know) was welcomed as an oddity in New Caledonia. The web app name is close to Google. It was sold as a search engine. However, it worked more like an open-source directory, with a sophisticated search engine made for searching its internal data only. People were sometimes expecting a world wide web search engine but were stuck with business pages only. We had to get everyone involved to complete all information besides our initial scraped data.

Our users were excited about this product, but somehow our design stood out a bit too much, especially the page editing part, so we decided to go through a complete UX review with a specialist.

We found out that we were selling our product as a search engine, but it didn't look like what people think of search engines. Since the release of Gougle.nc, pretty much all competitors have designed their products with a similar user experience. You don't want to reinvent the wheel, and it's better to just serve something that everyone expects.

Our first mockup

Here are the main differences :

• We used to welcome our users with pre-made search results, usually pages with the highest score in Algolia, just like ads. We use a design similar to Google, Baidu, or Yandex, with a search bar calling for action and some general information coming from news websites.
• The design could also help the initial loading performance since I didn't have to load search results before showing up on the page. This would benefit both the user experience and our SEO since this page is much easier to index for Google.
• One trick I planned to do, was to pre-load the next page so that anytime the user triggers a query, he would immediately see the search result page.

• Once you land up on the search page, after your initial request, you are, once again, provided with a similar experience as most popular search engines.
• Compared to what we had before, we significantly reduced the space used by a single result, especially the logo part that was considered huge before.
• Also, you can notice a different system design. I used and abused material design from Google in our initial version, suitable for internal apps but maybe not as good for a public-facing search engine.
• You can also see that this time we chose to disclose which results are sponsored and emphasized on the interface
• We are also moving the search result map out from the single result page up to the search page.

Summing up

Taking a week off to get an external point of view from an experienced UX designer was the most significant decision we have made so far. I realized that I've developed and designed a website for almost a year without overthinking users' needs and expectations, staying in what we call the ivory tower.

In the following weeks, I started refactoring our UI according to the mockups and feedback we got... until I wondered why shouldn't we develop our own Google instead of simply acting like it?

Follow the story here.