How did we do ORDO - search for repair of construction crews

    Introduction (a story about the prerequisites for the birth of the idea of ​​this project and the hardships of its implementation at the initial stage)

    This article is a summary of subtotals. Our project is aimed at making the rather unpleasant process of finding performers of various kinds of repair and construction works more convenient. The idea itself arose as a result of a long and unsuccessful search for a performer of a process that was quite commonplace for summer residents - digging a well. I spent a lot of free time in search of a worthy applicant, looked at many sites of teams and companies that are engaged in the type of work I need, tried to find adequate reviews on these companies. The result of all this was a sense of wasted time, since all the brigades and companies vouched for their work, promised golden mountainsdeep and water-bearing wells, and adequate reviews of people who used the services of these organizations were never found. And then I realized that at the moment there simply is no quick and reliable way to find a proven workforce. After all, when we need to glue wallpapers, change batteries, cycle the floor, there are not so many ways to find artists:

    1. search on the Internet - in this case, you either have to overpay with large and well-known repair and construction companies, or risk money with unverified small teams that you still need to try to find;
    2. survey of friends - you can contact friends who recently made repairs and were satisfied with the result, and get the necessary contacts. And maybe it’s completely lucky if your friend is engaged in an interesting type of activity - in this case, the likelihood of resisting poorly performed work is extremely small;
    3. Well, a very old grandfather way - searching for ads at the entrance - is also a very dubious option.

    Note - Here you must immediately make a digression. At that time - about 1.5 - 2 years ago - with all my desire and a variety of search queries, I could not find sites that at least slightly implement the functionality of our project. But already in the process of developing our service, analogues were discovered, but to retreat, as they say, it was too late.

    I thought the ideal resource was reminiscent of Yandex.Market, where the user can either select the category that interests him, and look at the teams and companies with their reviews, photos of the work done and other information, or place your order and choose from those brigades that responded to his offer.
    I hatched this idea for quite some time. This was partly due to a lack of free time, but the main reason was the lack of a budget for the development of this project. The amounts necessary for the implementation of the project were beyond my means, and my knowledge was clearly not enough. Therefore, the idea for a long time remained just an idea.
    But the stars have developed so much that after a while I met a person who supported my idea and as a result of a long search we were able to find someone who had sufficient knowledge to implement the idea. And he called the amount quite adequate for our budget - absolutely free. Yes Yes. It happens in life. Only one idea was able to assemble a small but ambitious team, ready to implement all the plans and develop the service to the bitter end.
    At this point I will complete the preface. We omit all the moments of a long and tedious coordination of the basic concepts of creating a service. And without this, the preface turned out to be unexpectedly long and drawn out.

    The main part (the main stages of creating a service and the technical methods of its implementation are considered)

    It all started with the fact that as a result of the bidding, we miraculously got the domain name, which, by a lucky chance, was released just at the time of our searches.
    The second issue, which took us quite a lot of time, was the choice of the concept of searching for the necessary services. The initial idea of ​​its implementation was aimed at maximizing visualization by providing the user with a sectional view of the property, where he was asked to select the parts that need to be repaired / built. Something like this:

    Or like this:

    But for the first time we had to abandon this option due to the complexity of its implementation. It was decided to start the project with a simple search and filters.
    After some time, the first version of our site worked in test mode. It was a mixture of a bulletin board and social network with teams and companies sorted by type of repair and construction work. Through one-time targeted advertising for several days, the first hundred registered teams were formed, with the help of which the service was tested. We worked out the site, corrected bugs and typos. It is worth noting that so far we are not perfect in this regard.

    Note - Many inaccuracies and errors have been found for a long time, but since this project is only a hobby and so far cannot fully provide financial support for its participants, there is simply not enough time to correct them. But we try and constantly improve what has already been done, not forgetting to add new functionality.

    The Malomalsky functional for construction crews was implemented, but the logic of organizing our filters was still very undeveloped. Unfortunately, I could not find the final image with filters, but I am enclosing an intermediate version:

    At that stage, builders could already create their own pages with descriptions, contacts, photographs of the work performed. Users, in turn, could view all the information and leave feedback and ratings. Of course, there was no question of any rating system at that time. The three main branches - “Apartments”, “Residential Real Estate”, “Commercial Real Estate” - consisted of large sheets, cut into 20 companies, sorted in the order of registration (new companies at the top).
    But after some time, the rating was implemented. We did not begin to reinvent the wheel and come up with some ridiculously cunning formula for calculating the rating of a particular team. We used the Yandex article, which describes in detail how Yandex.Metrica rating is calculated (for which we express our deep gratitude to Yandex). The formula is as follows:


    • - assessment of the team by users;
    • - assessment of the team by administrators, that is, by us;
    • - the number of user ratings.

    At the moment, this method adequately sorts our entire database of ads, introducing a piece of justice in the fierce competition inherent in this industry.
    It should be noted here that we approach the monitoring of the process of leaving feedback and ratings by users quite carefully. When leaving a review and rating, the user needs to fill in the required field “Telephone number for communication”, by which we check the reliability of the information provided and only after that add the review text and rating to the site.

    Well, the last step that can be distinguished is to enable users to publish orders for which teams and companies will leave their feedback. To be completely honest, this function was not initially envisaged by us, but was spied on by one of the competitors, and, what can I say, it is really convenient for the end user. It was introduced recently and is in the process of testing.


    So what is the purpose of this article?

    The first and most important task is to present our project to a wide audience, consisting mainly of professionals in their field. We really hope for constructive criticism and comments. None of us can say with certainty that the path or implementation that has been chosen is extremely true. We are interested in both technical notes and tips on the implementation of the design.

    In the bottom line, we have a project that has only outlined the contours of its development. What has already been implemented to date needs to be long and carefully modified, simultaneously introducing new functionality. But, despite this, we did what we had planned for a long time. As the saying goes, a start has been made.

    Also popular now: