Loading ...
Sorry, an error occurred while loading the content.

Starting out - research, learning, preparation or real user stories?

Expand Messages
  • a.bobrovskiy
    Hello, Scrum Community! We are just starting out a new project with a new team. Our challenge is that we need to build a complex JavaScript-based user-friendly
    Message 1 of 4 , Apr 28, 2011
    • 0 Attachment
      Hello, Scrum Community!

      We are just starting out a new project with a new team. Our challenge is that we need to build a complex JavaScript-based user-friendly interactive drag-n-drop kind of an application. We really do not have good experience building that complex applications in JavaScript.

      As a tech lead, I want to take some time for research and experimentation with various frameworks and libraries, doing some toy-projects on various frameworks and approaches. In this case I can create product backlog items for each research or experiment goal and supply it to the product owner with an explanation that these items should be highest priority because we need it. Is this the way it is usually done? Or may be scrum teams do it differently?

      Thank you for any help!

      Artem B.
    • Koti Reddy
      Hi Artem, In your case you can spend some time in researching the best possibilities for you as you have mentioned. Also as you mentioned having a plan for
      Message 2 of 4 , Apr 28, 2011
      • 0 Attachment
        Hi Artem,

        In your case you can spend some time in researching the best possibilities for you as you have mentioned. Also as you mentioned having a plan for each research on different framework make sense and you can claim each one as a story point.

        Also i suggest you can call all those sprints as "Research Sprints " or "non functional sprints" till you conclude which one to actually implement in actual feature delivery sprints. Why i am calling them as Research Sprints or non functional sprints is we are not actually shipping them to the end user but required as part of your work. The credit and accountability everything can be claimed for those sprints as well but different naming conventions for the easiness of the stakeholders.

        The above comments are truly from experience and understanding of scrum process. If anyone has different opinions, please go ahead and provide more inputs on top of this so that we all can mutually beneficial.


        --- In scrumdevelopment@yahoogroups.com, "a.bobrovskiy" <a.bobrovskij@...> wrote:
        >
        > Hello, Scrum Community!
        >
        > We are just starting out a new project with a new team. Our challenge is that we need to build a complex JavaScript-based user-friendly interactive drag-n-drop kind of an application. We really do not have good experience building that complex applications in JavaScript.
        >
        > As a tech lead, I want to take some time for research and experimentation with various frameworks and libraries, doing some toy-projects on various frameworks and approaches. In this case I can create product backlog items for each research or experiment goal and supply it to the product owner with an explanation that these items should be highest priority because we need it. Is this the way it is usually done? Or may be scrum teams do it differently?
        >
        > Thank you for any help!
        >
        > Artem B.
        >
      • Steve
        Hello Artem A couple of points: Does your R&D have to be paid for out of the project budget? Seems to me that your team will be gaining expertise that is
        Message 3 of 4 , Apr 28, 2011
        • 0 Attachment
          Hello Artem

          A couple of points:

          Does your R&D have to be paid for out of the 'project' budget? Seems to me that your team will be gaining expertise that is valid beyond the scope of the immediate project and would ideally be paid for out of a general development budget; that is if one exists.

          However, how ever it is paid for, that is obviously a time implication; is the PO looking for an early release date that with R&D time would make getting a viable release impossible?

          Can you not do what is required with current skills?

          OK, so you have decided that R&D is a goer (the PO agrees); in Scrum many would call this time 'Sprint 0', a special type of Sprint that does not deliver a potentially releasable part solution.

          You can use PBI but the PO does not prioritise them, the Team does or you can jsut use a Sprint Backlog with the Team decided R&D tasks; I have to say that I prefer the latter, keeps the PB consistent with 'proper' Stories.

          A thought about if the R&D can be paid for by 'someone' other than the business project, you could set up a completely seperate project to do the R&D before you begining the business project.

          Hope that is of some help

          --- In scrumdevelopment@yahoogroups.com, "a.bobrovskiy" <a.bobrovskij@...> wrote:
          >
          > Hello, Scrum Community!
          >
          > We are just starting out a new project with a new team. Our challenge is that we need to build a complex JavaScript-based user-friendly interactive drag-n-drop kind of an application. We really do not have good experience building that complex applications in JavaScript.
          >
          > As a tech lead, I want to take some time for research and experimentation with various frameworks and libraries, doing some toy-projects on various frameworks and approaches. In this case I can create product backlog items for each research or experiment goal and supply it to the product owner with an explanation that these items should be highest priority because we need it. Is this the way it is usually done? Or may be scrum teams do it differently?
          >
          > Thank you for any help!
          >
          > Artem B.
          >
        • avi_a@mapa.co.il
          Hi All. For me these R&D activities would be named categorized as a spike and placed in the RBL like any other story. Usually we ll have most of them during
          Message 4 of 4 , Apr 30, 2011
          • 0 Attachment
            Hi All.

            For me these R&D activities would be named categorized as a "spike" and placed in the RBL like any other story.
            Usually we'll have most of them during sprint0, but we might decide to conduct some spikes throughout the release, scheduling them in a sprint just like any other story.

            As a PO I have no problem accepting them and justifying them to stakeholders - for me, each sprint should increase both value and knowledge (eliminating risk). So if there is a story that the team can't estimate because they don't know enough about the technical issues involved in it, I would like them to conduct an R&D spike which purpose is to enable them to estimate other stories which have business value to me so I can incorporate them into the release plan.

            Also if a certain story might be implemented in different technical ways, I'd like the team to conduct a spike on the possible solutions so I can weigh out the possible costs and benefits.

            To stakeholders I'd say that these are R&D activitie necessary in order to asses the feasibility and cost of a feature, and that it will be time-boxed and require some budget. This is usually accepted by the stakeholders if it is within acceptable limits, which of course it should be.

            Also, as a PO I can decide which spikes I want and when I need them, depending on the value of the stories they would shed light on and on the risk they will be eliminating for me in terms of planning.

            Some things to note though:
            * The spikes are usually decided upon collaboratively with the team, although the final decision of if and when to conduct them is the PO's, since he is ultimately responsible and held accountable for the total ROI.

            * Spikes are estimated and timeboxed by the team for sprint-planning purposes, usually in hours/days.

            * On the other hand, Spikes are *NOT* estimated in storypoints and are *NOT* counted in the release burndown nor in the team's velocity.
            This makes sense, since if a team is busy during half a sprint conducting spikes it means they are doing less work delivering features and their overall velocity and ability to deliver features drops. This should be visible and taken into consideration when deciding if and when to incorporate spikes.

            HTH,
            - Avi


            --- In scrumdevelopment@yahoogroups.com, "Steve" <steve@...> wrote:
            >
            > Hello Artem
            >
            > A couple of points:
            >
            > Does your R&D have to be paid for out of the 'project' budget? Seems to me that your team will be gaining expertise that is valid beyond the scope of the immediate project and would ideally be paid for out of a general development budget; that is if one exists.
            >
            > However, how ever it is paid for, that is obviously a time implication; is the PO looking for an early release date that with R&D time would make getting a viable release impossible?
            >
            > Can you not do what is required with current skills?
            >
            > OK, so you have decided that R&D is a goer (the PO agrees); in Scrum many would call this time 'Sprint 0', a special type of Sprint that does not deliver a potentially releasable part solution.
            >
            > You can use PBI but the PO does not prioritise them, the Team does or you can jsut use a Sprint Backlog with the Team decided R&D tasks; I have to say that I prefer the latter, keeps the PB consistent with 'proper' Stories.
            >
            > A thought about if the R&D can be paid for by 'someone' other than the business project, you could set up a completely seperate project to do the R&D before you begining the business project.
            >
            > Hope that is of some help
            >
            > --- In scrumdevelopment@yahoogroups.com, "a.bobrovskiy" <a.bobrovskij@> wrote:
            > >
            > > Hello, Scrum Community!
            > >
            > > We are just starting out a new project with a new team. Our challenge is that we need to build a complex JavaScript-based user-friendly interactive drag-n-drop kind of an application. We really do not have good experience building that complex applications in JavaScript.
            > >
            > > As a tech lead, I want to take some time for research and experimentation with various frameworks and libraries, doing some toy-projects on various frameworks and approaches. In this case I can create product backlog items for each research or experiment goal and supply it to the product owner with an explanation that these items should be highest priority because we need it. Is this the way it is usually done? Or may be scrum teams do it differently?
            > >
            > > Thank you for any help!
            > >
            > > Artem B.
            > >
            >
          Your message has been successfully submitted and would be delivered to recipients shortly.