Creating a project management system at Yandex Tolstoy Camp

    Between the development of mobile solutions for large customers, ideas of new services constantly come to mind. In order not to delay their implementation in a long box, we decided to give ourselves an ass and passed the selection at Yandex Tolstoy Startup Camp 2014.

    We are working on creating a new project management system. The main goal is to develop a tool that will help small studios to more accurately evaluate projects, reduce the risks of projects coming out of time and budget, and more fully see the picture of what is happening.

    If you have heard about the Lean Startup approach, you should also know about Customer Development - an approach that makes startups not sit and silently saw a brilliant product for six months, but go out to their potential customers right away, look for their real problems and solve them. So, since Habr users are our potential target audience, we will test our hypotheses about problems for you. We will regularly post on how we are developing, which hypotheses were confirmed and which were refuted, how the vision of our product changed during the cast, and how much money we saved compared to the usual approach (write code, b❚❚❚❚❚!) , well, talk about Lean Startup and Customer Development.

    So, if you are a project manager or CTO, want to take part in an interview and express your opinion on what your project management system should look like, or just want to see how customer development works in a startup from the inside - welcome to cat.

    Many, having heard the phrase “new project management system”, will probably think - how does it differ from existing tools like JIRA, Redmine, Asana, Megaplan and many others. From our point of view, all these systems are good, but they can hardly be called “project management systems”, since they cover only a few of the processes that occur in the project (examples of such processes are task assessment, development of a hierarchical structure of work, control and management the cost of the project; there are others as well. Task trackers, as their name implies, cover task management in part for several other processes - reporting, quality control).

    We want to create a system that will cover a lot more processes. What for? I hope that the answer to this question will become clear to both you and us =) from our future publications.


    Consider a small studio that develops custom or mobile or web applications. Usually, each project of such a studio is divided into several steps:
    1. collection of requirements from the customer, initial assessment and sale;
    2. after signing the contract - the phase of analysis and design development. By the end of this phase, the customer receives a more or less accurate work breakdown indicating the components of the program being developed, supported by use cases, with a price affixed opposite each element. The format of this document is usually a Word or Excel file:
      Component / FunctionalityDesignDevelopmentTesting
      start page100500200

      The project manager knocks out the necessary resources for his project, understands the relationship between tasks, therefore, in addition to the document with estimates, the customer can be provided with a project plan in the form of a Gantt chart, compiled, for example, in Microsoft Project.
    3. After final approvals, the work on the project begins. The project manager starts tasks in the task tracker (for example, JIRA); the first sprint starts if the team works on scrum; Happy workdays are coming.

    ... and dangers

    In our opinion, such projects lurk the following dangers:
    • At the assessment stage, no risk assessment is performed. More precisely, each participant in the assessment process multiplies the figure that first came to mind by 2 (and an experienced appraiser - by 4), but usually no one knows what the risks are and whether they need to be increased / decreased for a given project.
    • This approach makes it difficult to negotiate with the customer, because it is difficult for sales to understand how much he can reduce the initial price before the project ceases to be profitable. As a result, the company offers a price higher than its competitors, and loses a potential customer.
    • Artifacts created as a result of evaluation rarely get into the task tracker in the same form in which they were presented to the customer. This makes it difficult to calculate actuals for each feature being developed (screen, use case) separately and does not make it clear whether we can afford to spend more time on it or not.
    • PM has to regularly and manually synchronize information in the bug tracker, project plan, reports that he sends to company management and customers. Moreover, it often happens that the items in Excel on which he reports to the customer have nothing to do with the list of tasks in the tracker (see paragraph above), and preparing each such report requires a lot of effort to understand what percentage of readiness is another part of the functional.
    • PM doesn’t have an understanding of how many risks were really taken into account when evaluating and after reaching which figure the fact the project works at a loss (well, or simply reduces the margin of the project)

    Do you face similar problems? How do you solve them? We will be glad to see your feedback on this topic in the comments to the post. Also, please spend a few minutes and fill out the survey .

    We will be especially grateful to those who agree to spend 30-40 minutes of their time on Skype interviews. To do this, simply register on the site and check the "contact me" box - and we will come after you. If you live and work in Moscow, we will be happy to meet with you personally to find out your opinion.

    Also popular now: