Software concept: how digital transformation succeeds
Inefficient processes in the company have been identified and the will to change is there. The answer quickly becomes: a software solution is needed. Ideally, you would also like to start development right away, but how can you transform the process in a structured way? A solid concept is needed.
In this case study, we therefore take a look at the concept sprint of a software development, which we carried out for the video game company Ubisoft.
"SIDESTREAM developed a concept for us to map an existing Excel-based process in a modern web application. For this purpose, a concept sprint was carried out with the inclusion of all departments involved in the process, the good progress of which was decisive for a high acceptance of the new process. The team's way of working is very pleasant and goal-oriented."
The starting point: dynamic data in rigid tables
The video game industry is a fast-moving field. There are new games, updates and promotions on a regular basis. All of this requires careful planning and precise evaluation based on a reliable data foundation. Ubisoft's previous process for planning and accounting for sales campaigns is heavily Excel-based. This has allowed the digital implementation of the process to grow with it over the years. This is because Excel is very flexible. Thus, the diverse requirements of the users could be covered most of the time.
In the long run, however, the implementation of the process branches out. The collaboration of different users on many files, spreadsheets and macros is confusing and leads to a number of disadvantages:
- Employees spend a lot of time manually maintaining data, e.g. sales data is entered and kept up to date. This requires a lot of resources and is prone to errors.
- Data can quickly become inconsistent if controlling is possible on different aggregation levels by "copying together" and modifying different Excel files. This effect is compounded when different departments use the files.
- Excel spreadsheets make reliable real-time integrations with external data sources (e.g. Amazon Seller Central) difficult. Delayed availability of data makes it difficult to perform timely analyses of the current state and, consequently, target planning.
The requirements: Concept for a user-friendly and practical application
After a few preliminary discussions, it was clear that the solution to the problem fell within SIDESTREAM's area of specialization. In the development of solution concepts and the development of a new tool, we relied on a "democratic" process. In short: Not the managers alone decide what will be built, but all involved departments should be heard. After all, only when we hear about problems and wishes first-hand, a sensible solution can be designed with a high level of acceptance. The day-to-day benefits of the solution are the focus at all times.
To this end, we have used our proven three-stage process for concept sprints:
1. preparation and planning.
First, we discussed with Ubisoft's process owners and established initial problem areas and solution hypotheses for the process. Despite the management summary, it is important to get deep into the users' heads and obtain as much detailed information as possible. The density of information can be daunting at first. One should not rush into the programming phase if a very good result is to be achieved. Therefore, in the second phase it is necessary to delve deeper into the process in order to be able to design a meaningful solution.
2. information acquisition
In order to get deeper information about the process, we held a remote workshop day with the employees of Ubisoft. This served to understand details and contexts of the process. We conducted interviews with 1-2 stakeholders at a time to dig deep into the existing processes and tools to understand individual needs and desires. In this case, we conducted interviews with sales, key account management, controlling and internal IT departments and uncovered a wide variety of problems and wishes. For example:
- Better reporting,
- smart auto-completions,
- partial automations such as campaign duplications,
- granular access rights and change logs.
3. elaboration and presentation
Using the information gained, we worked out the concept. In this phase, too, it was important for us to remain in close contact with the specialist departments in order to clarify questions, discuss and iteratively improve elaborated preliminary versions. The concept sprint is concluded with a handover and presentation of the deliverables.
For Ubisoft, we worked on the following deliverables:
An essential deliverable is a wireframe that shows how the application will look later and how various challenges will be solved using good user experience. Already during the concept sprint, we use the wireframe to play through the various work steps with the respective employees, thus uncovering gaps in the solution (without having written a line of code!) and increasing acceptance for the future solution. For the Ubisoft wireframe we used the tool Whimsical. It can be used to iterate quickly and is well suited for table- and data-heavy applications.
Target Software Architecture:
Where wireframes uncover and solve various functional blockers, designing a software architecture ensures that technical blockers are also uncovered and solved early on. It is essential here that requirements from the various abstraction levels (interviews, UI/UX, technical implementation) are communicated smoothly. This is the only way to create a customized software architecture that is as simple as possible (and thus minimizes effort) and at the same time future-proof and expandable.
The software architecture for Ubisoft consists of a number of microservices: a web app, a reverse proxy, a dedicated user management and a backend including database. The software architecture was implemented using the draw.io tool.
Based on the wireframes and the target software architecture, a development plan can be created as the final deliverable of the concept sprint. Despite agile collaboration, it is important to have a preliminary cost and implementation plan. This increases planning security and enables an ROI calculation of the development project. For the development plan, the developed specifications are divided into several packages, each of which is provided with effort estimates.
The first phase forms the basic construction and in this case was estimated to take three 2-week sprints: at the end there is a functioning application with the most important functions and a stable foundation for further functions. In addition, we have already set up plans for further phases and sprints. Our credo is: Best Value first: All features are weighed up in advance according to their benefit/effort ratio (value). Features with the best value are implemented in a prioritized manner. This approach maximizes the business value of the application.
For the Ubisoft development plan, we used the Productplan tool, which has a wide range of relevant features, as it was developed specifically for development plans and product roadmaps.
A carefully and practically created concept is the cornerstone for the development of a solution-oriented application. For this purpose, it is crucial to enter into an exchange with all process participants. This means that both users and participating departments are involved in the planning. In addition to the pure gathering of information, the focus of the planning is on the presentation and play-through of the target application. In this way, decisive features and workflows can be worked out at an early stage.
After a concept sprint, you are left with a sharpened understanding, a customized implementation plan and an estimate of the effort required. Now you can calmly decide whether, how and with whom to start the development phase.