Stages of game development for mobile platforms

Sergey Anankin, a producer at Pixonic, wrote an introductory article describing the stages of developing a mobile free-to-play game. With the permission of the author, we publish it on the pages App2Top.ru . In the future, Sergey plans to write a whole series of materials dedicated to the creation of mobile games.

1. Stages of project development

The diagram below shows the stages of project development with their characteristic durations and team sizes, as well as the events separating these stages.

When developing a project "from and to", the following stages can be highlighted:

• The Alpha stage. The project producer receives a task from the management (sometimes participates in its formation), develops a concept (i.e. an initial description) of a new project and gathers an initial (alpha) team to implement it;
• The "Pre-production" stage. The Alpha team identifies potential risks to the project and either eliminates them or demonstrates the ability to eliminate them in the future. The producer draws up a production plan for the project, forms a full development team. The project documentation is being written. A prototype of the project is being developed.
• The "Soft lunch" stage. The prototype of the project is being expanded by the development team to a full-fledged version of the game in accordance with the drawn up plan and project documentation. The stage ends with the release of an early version of the game for the selected market. Usually such a version contains 30-50% of the total amount of game content and differs from previous technical versions in its high stability and low number of errors.
• The "Hard lunch" stage. Statistics are collected on the early version of the game, taking into account these statistics, the game is brought to a later version, released to all selected markets. A marketing strategy is selected and a plan is formed to further improve the game.
• The "Support" stage. A support team is being formed (as a rule, it is part of the development team) to release consecutive updates taking into account statistics. The producer continues to work on the project as a consultant, shifting the main tasks to the game designer of the support team.

2. The Alpha stage

• The beginning of the stage. The producer receives from the management (or helps to form) the primary task — that is, a set of conditions for the start of the project and a list of requirements for the project. The conditions are usually: the desired scale of the project (for example, small or vice versa — large-scale long-term construction), the required development time (for example, no more than six months), financing conditions (for example, no more than $ 300 thousand for everything), pre-selected technologies and team properties (for example, the company wants to do a project on Unity due to the availability of its own adaptations of the engine and programmers experienced in working with Unity). The requirements are the project's indicators in terms of retention, attraction and monetization of incoming traffic (for example, the game must provide 50% of the players' input on the seventh day of the game and earn at least one hundred thousand dollars per day).

Often, however, a project does not start with a clearly defined task, but with a producer's suggestion. So, starting the development of one of our projects — the Renaissance game — we first formed the idea in a few words: a tactical mid-core survival shooter in a post-apocalypse setting. At the alpha stage, the management helped us to form the concept of the project taking into account the wishes of the company.

• The goal of the stage. The producer needs to receive and approve the project concept and the composition of the alpha team from the management.
• The course of the stage. The producer forms the concept of the project — that is, any description that gives information about how, under given conditions, the project will meet the requirements presented to it. The concept usually includes a description of the target audience, the setting and genre of the game, key (possibly unique) gameplay features, development techniques, tools and technological solutions, sales markets, supported platforms, deadlines and costs of implementation, requirements for the development team. One of the key success factors at this stage is the analysis of the current market situation — tracking existing trends, reviewing the most successful projects, choosing new development tools for mastering, etc.

Concept approval is almost always an iterative process. To approve the concept of the Renaissance game (due to the planned scale of the project), we needed a dozen meetings with the involvement of most of the company's producers, and not only from the mobile direction. Following the results of each meeting, the concept underwent changes, sometimes significant.

• The end of the stage. Once the concept is approved, an alpha team is selected - that is, a set of key employees who will form the basis of the future development team. Usually it is the producer himself, the art director, the technical director, the leading game designer. The producer and the alpha team move on to the next stage.

A good practice, in my opinion, is some work by the alpha team already at this stage. Key players can participate in the development of the concept along with the producer. For the Caribbia project, by the time of the final presentation of the concept, we had prepared not only the initial design of the project, but also examples of art, and also solved several technical issues with the help of the future technical director of the project.

• The duration of the stage. As a rule, the elaboration of the first version of the concept takes one or two weeks. If the concept does not meet the requirements of the management, an iterative improvement of the concept is possible. Each subsequent iteration takes a week, and the total number of iterations (including the first one) is 3-4. Thus, the Alpha stage takes from two to five weeks.

When determining the duration of this stage, it seems to me that we should proceed from the scale and complexity (in any sense) of the proposed project. If the project is performed in a fundamentally new genre or has a unique gameplay, then this also leaves an imprint on the time of development and approval of the concept. So it took us two months to approve the Renaissance concept, while the Lost it Baliboo project concept was approved in less than a week.

3. The "Pre-production" stage

• The beginning of the stage. The producer and the alpha team identify the most significant potential risks for the project. Risks can be generated by the game (will the gameplay be interesting? does the graphics match the target audience? will the game go on the selected device?), management (will we have time to make the described game in the allotted time? how much money will we spend?) or the team (will the team cope with the deadlines, will it be able to exhaust all the risks?).

The set of identified risks strongly depends on the project. For example, in our current Lost in Baliboo project, the main risk to minimize was the quality of graphics, since the mechanics and technical solutions of the project inherited from its predecessor Robinson Island.

• The purpose of the stage is to minimize the risks found by designing a prototype of the game, developing and detailing project documentation, and drawing up a development plan for the following stages.
• The course of the stage. The risks posed to the game are minimized by creating a prototype (an initial version of the game containing examples of functionality, graphics, or other risky aspects). Management risks are minimized by drawing up a production plan (i.e., a description of the development process with reference to time) and requirements for the development team (size, structure, properties of participants). The team's risks are minimized by choosing technologies that have already been mastered, as well as teams with experience in releasing similar projects or using these technologies.

Prototyping can also be an iterative process if the project requires it. In Lost in Baliboo (a project with minimal risks), the prototype was demonstrated and approved only at the end of the stage, while in the currently released Walking War Robots, gameplay and controls were demonstrated to the entire office every week to make sure that the design solutions were successful.

• The end of the stage. When the prototype is demonstrated to management and approved as solving the task of minimizing risks, when the production plan is approved, a development team is formed and the next stage is launched.
• The duration of the stage is determined depending on the number of identified risks. How to properly pre-production takes no more than one sixth of the total project development period. For example, if the worldwide release of a product is scheduled for the end of the sixth month of development, then pre-production will last one month.

Robinson Island, for example, had a two-week pre-production, and its prototype demonstrated, for the most part, the team's ability to port the existing social version to mobile platforms. This deadline was just the sixth part of the project, which was planned to be launched on the market in three months.

4. The "Soft lunch" stage

• The beginning of the stage. The producer and the development team resolve the remaining issues on development methods, workflow organization, tools, etc. Then the development process begins according to the approved production plan.

Speaking of development according to the production plan, we can highlight the approach that is gaining momentum in the West called Global Oriented Roadmap, which we use in the Renaissance and Caribbean projects under development. In this regard, the development stages (soft launch and hard launch) are divided into sub-stages. For each of these sub-stages, the date of the demonstration, the key features implemented at the stage, and the goal of the stage formulated in one sentence are determined. Such a roadmap allows you to keep the development process "flexible" in detail, have a local goal for the team, and also makes the development process easy for the company's management to control.

• The goal of the stage is to develop the first full-fledged version of the game, which will be released to a small initial audience in order to collect statistics and see the potential of the project.

A counter-example is the Robinson project, which did not pass the soft-launch stage and was released to a wide audience immediately with an extensive advertising campaign (see hard launch). This paradigm was determined in advance by the marketing department of the campaign due to factors such as cheapness and tight project development deadlines.

• The course of the stage. The team is working on getting the so-called soft-launch version — the first version of the game that will be available to a selected set of players (as a rule, this set is a certain segment of a certain market). There is usually a list of requirements that the company imposes on such a version. These requirements are formulated as certain statistical indicators that the game must achieve within a certain time after its release: retention (i.e., the ability of the game to delay the player in the game for 1, 2, etc. days), conversion (i.e., the proportion of players who make payments within the game), average check (the size of the average payment) and the average frequency of payments. All these parameters are key in calculating the so—called life time value - the average earnings per player for the entire time he is in the game.
• The end of the stage. When the statistical picture of the project is already visible after the launch, the management decides on the long-term fate of the project. In the worst case, this is the closure of the project as having no potential, and at best, the transition to the next stage.

The release of the game to the market is the milestone of the soft launch stage, but not its formal end. The project may go through several iterations of improvements before a decision can be made on its development (i.e., on moving to the next stage). The Soccer Tactics project underwent a number of updates (in total, in two months) after an early release, before its release to a wide audience took place.

• Duration of the stage: depending on the criteria for open beta, the stage can take from 40 to 80 percent of the time remaining from the end of pre-production to the world release. For example, for a project with a total duration of six months out of the remaining five months after pre-production, a soft lunch can take from two to four months. If the company wants to release a very early version in order to start collecting statistics as early as possible, this will be the case with two months, and if statistics are supposed to be collected at the stage of an almost finished game, this will be the case with four months.

As always, there are options. As described above, the Robinson project did not have this stage at all. The revenues from the first advertising campaigns showed the need to increase advertising turnover without modifying the project.

5. The "Hard Lunch" stage

• The beginning of the stage. When the statistical picture of the soft-launch version meets the set criteria, the development team begins to prepare a more complete version of the game for launch in all possible markets. The list of additions and changes to the project consists of both pre-planned works and those identified by the results of statistics collection.

For example, the initial statistics of the soft launch version of the Soccer Tactics project showed problems not at all in the places where they were expected by the team. Because of this, after the soft launch, the list of proposed improvements to the hard launch project was significantly reformulated.

• The goal of the stage is to prepare a version of the game for a worldwide release, i.e. release to all planned markets and audiences.
• The course of the stage. All changes are usually made to the game gradually, as long as it continues to be available for download to the soft lunch audience. Each such change changes the project statistics, which may affect the task list.

The Soccer Tactics project was developed with a clear account of the statistical picture of the soft lunch version. So, due to the large outflow of players in the early stages of the game (which was reflected in the so—called funnels - showing when players leave the game), the team had to curtail all other work and focus solely on the edits of the first game session, the tutorial and the initial userflow.

• The end of the stage: When the statistical picture of the current version of the game is satisfactory, a so-called hard lunch takes place — that is, the launch of a full-fledged version of the game for all selected markets. The statistical picture of the project then determines the fate of the project. As a rule, the "hard lunch" stage goes through several identical cycles, at each of which the current statistics determine the front of work for the development team and influence the choice of marketing strategy.

Here it is necessary to understand that soft-launch and hard-launch are not black and white. The size of the advertising budget is determined (among other things) the current performance of the game and does not have to be minimum or maximum. Robinson Island started the release with a small advertising budget, but showed such a good return that the budget instantly increased many times (which is why we are essentially talking only about the hard launch of this project).

• Duration of the stage: the remaining time until the world release. In the case of an early soft lunch, this is 60 percent of the time between the end of pre-production and the worldwide release, and in the case of a late soft lunch, 20 percent.

The percentage of soft- and hard-launch durations is determined by the project and the experience of the team and the company. Lost in Baliboo, due to the minimal risks and its small scale, was released after three months of development, while the Dwarves Tale project soft lunch was held 9 months later. At the same time, Baliboo came out with minimal functionality, and Dwarves (despite the failure of the soft lunch version) contained most of the game mechanics and content.

6. The "Support" stage

• The beginning of the stage. If the project statistics stabilize at a satisfactory level, and the project ceases to require significant improvements and extensions, it is given "for support". Before that, a support team is selected, which is often part of the development team.

When developing Dwarves Tale, the fate of the project was decided precisely by the cost of its support. The life cycle of the project turned out to be short due to the large size of the team required for support. At the same time, support for the Baliboo project is relatively cheap, because it is carried out by a small team.

• The goal of the stage is to maintain an acceptable level of income from the project with minimal means until such a level becomes objectively unattainable.
• The course of the stage. The support team iteratively improves the game with small additions (these are either content extensions, bug fixes, or functional innovations). Such iterations with improvements can continue as long as the support of the game is justified from the point of view of management.

As a rule, we release content updates during the support phase. For Baliboo and Robinson, these are primarily quest packs, new islands to explore, and new characters. New mechanics are rarely introduced and only in order to fix the problems of the project, which statistics show.

• The end of the stage. If further maintenance of the project becomes unjustified for any reason, the project is considered closed. At the same time, it can remain accessible to the end user, if it is appropriate and does not require the participation of a support team.
• The duration of the stage is largely determined by the properties of the project. An acceptable period of reasonable project support is one year, whereas two to three years can be assessed as a very good result.

Solely the company's income from the project determines the duration of the support phase. So Robinson has been around for more than two years, since its support is relatively cheap, and revenues with a large margin cover the costs.

From the editorial office App2Top.ru : Other articles by Pixonic can be read in the companies section on our website.