What evidence can you share that demonstrates the impact of the engagement?
This collaboration was essential—the affiliate network couldn’t have come into existence without CODEST’s work. The application served as the very heart of our business. It covered all essential operations for publishers running ad campaigns, financials, and so on.
I’m partly satisfied with their work. In the very beginning, the application ran well: it was stable, and it provided all the functionalities we needed. However, as time passed, it became a little messy, and some bad practices accumulated. During the last year of work, we had a number of problems connected with performance, stability, and optimization, which cost us a lot of time, money, and led to some stressful moments. These problems appeared after three years of continuous development on the app, including data consistency and code optimization.
How did CODEST perform from a project management standpoint?
We worked with them remotely on a daily basis. There were a lot of email and phone exchanges, and the developers acted as a remote part of our team. We also had longer remote conferences 2–3 times per week.
I was satisfied with the level of collaboration as they were always at our disposal, they answered emails and phones instantly, they had good engagement, and they addressed problems in a professional manner.
Often, they fulfilled duties that were meant for infrastructure administrators, rather than developers. There were some glitches with our hosting company, and CODEST was a great partner for us in solving these problems.
What did you find most impressive about them?
As a business manager, I’ve only worked with Agora’s internal IT team and with CODEST. Compared to my former IT department, I appreciated the quality of CODEST’s service. They’re faster, better programmers, more engaged in projects, more responsive, and so on.
Are there any areas they could improve?
They could’ve produced better-quality code. They made some mistakes designing the database structure, which became ineffective after a few years of continuous use. The quality of the code they produced wasn’t top-class, but I can’t say they’re incompetent, either. They made the application from scratch and their level of understanding into our needs and requirements was high.
This was the very first affiliate networks application that CODEST produced. It’s possible that some flows in this app were inadequate because of their lack of practice in the industry. Our own team was also new to the requirements, so we were less accurate when forming our requirements.
They could also improve by planning more before working on the design, especially around databases. This would make them more flexible and expandable. A microservices approach would make their apps less bloated and more manageable.
Do you have any advice for future clients of theirs?
Commit CODEST to produce the app. They have more experience now, and application development is in their DNA. Clients should hire developers that are not only good, but who have experience; who know the structure and requirements of a field, and who know the directions in which an application can develop. This will aid in avoiding some young-age mistakes that can then bias the ongoing development of the product.
Secondly, work closely with the team to design the core structure of an app and its database. Leaving it solely to the developers can backfire later on when the market pushes the company to steer the app in a new, unexpected direction.
Finally, let CODEST take care of server administration and configuration. Committing a third party to run the infrastructure is a bad option, as gaps emerge in the communication between the developers and the admins. CODEST is capable of running an infrastructure, so I advise picking a cloud hosting solution and bringing CODEST in to deal with it in a top-down manner.