How did you select this vendor?
It was a bid process. Based on a previous experience in developing web applications, technology stack, and experience in developing property market products we have chosen few software houses and ask them for offers.
Future Mind provided us with the best offer and had the most experience in the required tech stack. We liked their approach to development and project management. Had good communication during our meetings.
Describe the project and the services they provided in detail.
Property Group Sp. z o.o. provided a website prototype - including a business concept, UX / UI in the form of hi-fi mockups and style guide, as well as requirements for the tech stack updated during the project. The Product Owner and, if necessary, dedicated tech, design and SEO specialists took part in the project from the Client's side.
We decided to use Python, along with its framework Django, for the backend layer. This was mostly due to the fact that Python works well for Rapid Application Development, which gives the developers a lot of possibilities when it comes to prototyping, taking advantage of ready-to-use components, and speeding up the development process.
Indeed - the first effects were visible almost immediately. At the same time, Django is a powerful web framework supported by a large and active community, not to mention substantial documentation. It also makes it easier and faster to build advanced web apps with less code. When it comes to the frontend, we chose React.js.
The decision was influenced by the our great experience with both SEO and React itself, which proved to work better than i.e Angular in previous projects. These technologies also formed part of the our’s tech stack, which in turn minimises the costs of maintaining and developing the product further, making it easier to collaborate when using common programming languages.
Since primary and secondary markets are quite different, we needed a solution that would help us categorise the offers any way we want it, while keeping the search for listings as efficient as possible. That’s why we decided to take advantage of the search engine Elasticsearch. This largely influenced the speed of the searches and thus, the user experience.
One of the other major challenges for GetHome, however, was the existence of duplicates when it came to the listings (which affected the credibility of both the offers and the whole portal). Therefore, finding the way to track those duplicates was one of the our main priorities during the development of GetHome.pl.
Developing such a precise mechanism required taking into account multiple attributes of the properties and handling a lot of exceptions.
The listings of the same properties are now “merged” and displayed within one view, which is why the user doesn’t find the same house or apartment multiple times. As we have an in-house SEO team, we were also fully aware of Google Search which is one of the most effective channels for acquiring new users.
That is why proper optimisation of the whole website for search engines was the main requirement when developing it. We kept that in mind from the very beginning, assessing the impact that any changes and improvements might have on search results.
What was the team composition?
In the beginning we had two back-end developers (Python/Django) and two front-end developers (React.js) as well as Project Manager assigned from Future Mind. They were backed-up by some of our top developers as consultants when needed in discussions.
After launching MVP we’ve work with one back-end developer, one front-end developer and Project Manager. Few months ago our internal team took over the project but we are still cooperating with two back-end developers from Future Mind.