Social Jet - Marketing platform

In 2020, a serial entrepreneur approached us and proposed that we take part in a very special project - a marketing platform. The project's distinctive feature was its critical, practically extreme complexity, so professional help was necessary. The project did not just have existing developments, it was already working and generating money. At the same time, it was a system written by professional engineers, but since it was originally a startup that did not guarantee success, all efforts were focused on quickly releasing functionality, and the project architecture as such did not exist in a formal and documented form. At the start, this was not a problem, but the further things went, the more the owners invested in quality rather than speed of feature release (the key money-generating features were already known). The lack of standard architectural practices and deployment principles was beginning to noticeably slow down the writing of new code, which would inevitably have led to an insurmountable ripple effect and, probably, the need to rewrite everything from 0. It is important to note that today this is no longer the system it was at the start of our cooperation; the project has made a huge leap in quality and is a professional innovative platform.

SocialJet is the largest analytics and advertising platform for marketing and business. More precisely - a Content Distribution and Native Advertising Platform.

To describe the essence briefly - if you sell any product, the platform automates for you the search for and placement of advertising with various bloggers and opinion leaders.

On the other hand, if you create your own content and earn from advertising, the platform can allow you to focus on your creativity rather than searching for advertising clients and sales.

Those who have dealt with influence marketing know perfectly well how much time is spent on operational work of this kind and how many nuances there are. The platform will take over all the routine; you only need to enter the data.

The database contains more than 300,000 content sources; it can be used for filtering, selections, and launching advertising campaigns.

It is important to note that despite the convenient interface, the platform provides a huge number of settings and configurations - which is rarely compatible. And technically, it consists of extremely complex subsystems.

One of the biggest complexities is the pricing system; the system calculates the cost of advertising publication based on various factors, automatically publishes this advertising with bloggers (and also removes it automatically, because some tariffs involve a temporary post). And it does this across all popular social networks and messengers.

And, adding even more complexity, the system does this in real time, around the clock, without breaks. And regular system updates must not stop its operation, because we are talking about enormous budgets of a huge number of users. One small failure - and the company has major legal problems that will need to be resolved through manual recalculations and wagonloads of apologies to advertisers and content makers. In other words - responsibility of the highest order.

Achieving such requirements is quite difficult. The system must be resistant to changes and updates, to separation, to high traffic and load scaling, to failures on the server and in individual services.

Another challenge of this system was that the platform had a long history of code writing, and some employees were no longer working on the project, while some could consult for only a couple of hours a week, so it was necessary to read and understand tons of code, and piece together hundreds of thousands of components of the internal structure of subsystems.

We studied the system, provided consultations on further changes within the framework of modern standards and project goals, communicated with the team that developed other related modules of the system, spoke with one of the old developers and, of course, regularly kept in touch with the investors, of whom there were several. But more than anyone else - with Maxim, who in particular performed the role of product owner.

A flattering review from the client about the cooperation can be viewed on our landing page.