How does Scrum work when you have several projects?

Question

I am well versed in the benefits and processes of Scrum. I get ideas for lagging, burning schedules, iterations, using custom stories, and various other Scrum framework concepts.

With that said ... I work for a website development company that manages several projects at a time, with six team members making up the "production team."

How does Scrum work with multiple projects? You are still planning an iteration for one project in a certain time, and the whole team is working on it, and then you go to the next project with a new iteration, when this iteration is completed? Or is there a “flexible” way to manage multiple projects with their own iterations with just one team?

Answer 1

Scrum really doesn’t require you to work on one standalone product. It simply says that you need to do a lot of things (product lag), there is a certain amount of development time available at the next iteration (developed from the speed of the project), and there are elements selected by the client / business as having the highest priority from this pool of problems / tasks which will be executed at the next iteration (sprint lag).

There is no reason to believe that the product lag and the sprint lag should be from one project - even in one project there will be separate units of work that are similar to separate projects - user interface, business level, database schema, etc. The development of enterprise software in particular is such that you have several code bases that all must evolve. Scrum process - meetings, questions, burning chart, etc. - everyone works, whether it is one project or several.

Having said that, in practice it’s often very important for each iteration to have the main theme - “make a reporting module” or “the interface with the XYZ API” - so many problems arise from one project or area and at the end of the iteration you can point to a large amount work and tick off against it.

Answer 2

I think the answer depends on " which will determine the priorities of the lagging elements " (i.e. decide what needs to be done first). If this is one person, then this person is the owner of the product for your projects, and you may have one backlog for all projects for all projects or delays for one project - and you select journal items from all projects when planning Sprint. In this case, Scrum "works" perfectly.

If each project is responsible, then you are likely to encounter some problems when each responsible - more or less consciously - tries to approve its project (s). IMHO, you will need to have one Product Owner with only the authority to resolve priorities through arbitration.

One rule that must be followed in this context is to never modify the contents of Sprint during Sprint . If necessary, you can reduce the iteration to two or three weeks to reduce the risk of adding an urgent element to the current Sprint.

Answer 3

I have to disagree. I think this is a key moment for the process when the team focused on one project during the sprint. If you have specialists who cannot contribute to the entire development process (content authors, graphic people, business process analysts, etc.), I would shuffle them from the team when they can no longer contribute. Or better yet, they are trained in some of the different tasks so that they can contribute to things like testing.

Another thing to keep in mind is that running projects simultaneously kills your schedule. Consider this: to simplify, say, we have 5 projects using the same team, and starting on the same date. Each project requires 3 months of effort. In the best case scenario, you will complete the parallel script all at once, and it will take 15 months. Your speed will harden, because you can pick up only 1/5 months of effort in one sprint. At the same time, you will also have 5 demonstration meetings. In the best case scenario, you deliver your 5 projects in 15 months, and your competition will demand that they can do the same job at 3. Your maturity teams will suffer because they can only account for 20% of their workforce. You may find that in fact you cannot complete some tasks in one sprint. If you need to change the number of projects that will be processed from 5, your team will have to adjust their evaluation habits, which can damage the effectiveness of the teams. In addition, it will be difficult for your team to organize themselves when a simple reassignment of a task may require the deployment of a new development environment before work begins.

If you consistently run the same 5 projects, you must deliver the 5th project in the same 15 months, but you would inform your client that your team has such a requirement that you have a 12-month backlog and that you You can use this time to clarify the goals of the project. Or, if you have a constant backlog, you know it's time to hire another team. However, your best project ends in 3 months with a client who quickly detects improvements over the active period. You can finish this project a year earlier and can put it on your resume. Your sprint speed will stabilize during this period of time, and you may find that it reaches its step after a project or two and can achieve more in this sprint.

I think that running projects in series are one of the biggest hurdles for an organization trying to take on fights. These are important cultural changes associated with the deconstruction of the project manager role, but the benefits of the scrum process are enormous.

Keep in mind that EVERYBODY does not have to be a full member of the team. They can attract your client in the waiting room, getting ready for the development process. I keep my business analysts, network architects, and graphic designers as domain experts and join them as needed. Let them work with sprint 0. You will be surprised how attractive the work is on the look and the workflow. It is also useful to prepare your client with the understanding that when development begins in earnest, their level of participation can actually rise and that it is important for them to be available. Tell them the schedule so that they have enough time to sort things out in advance, such as vacations and holidays.

Answer 4

I was in this exact situation.

If you have one project owner, Phillip is absolutely right; One sprint with one team should work fine.

If you have more than one product owner, then ideally the business side should choose one “priority” that is assigned a work schedule. This is by far the best solution.

If (as this is probably the case), the business does not want to change the way they want to prioritize things (that would be too convenient), then you can split the team. and run two parallel sprints. However, with a team of six, I would not divide it into more than 3 teams (I would not want to split it at all, but I think that 2-3 teams will work). Dividing everything further, as Kenny suggests, and the commands of one person seem somewhat meaningless to me, since then you no longer have a team, but only individual programmers.

If you split a team, I did not find much value in combining stands, if you do not have separate sprints working on the very same code base, but then this can be an argument to combine these teams for the purpose of the sprint.

Answer 5

Recently, I have been reading a different opinion, namely that in the Agile environment, the Project concept can be counterproductive and can be eliminated. According to this line of thinking, the organization should focus on Releases . This is because projects are artificial boxes that have no value until they end. They run counter to Agile's desire to often provide shipping costs. The release is more in line with Agile because it is value-oriented and because its scope can be reduced or expanded based on what teams can deliver to the next version.

There is a potential middle ground where what was previously called the Project in your company is now repackaged as a regular Agile Theme or Feature . The advantage of this approach is that the Theme or Function can now be implemented in chunks of value, as the owner of the product considers necessary.

There is still the question that one team is working on several products , and this is a question because of legitimate concerns about knowledge in the field and possible technical skills. But products created with Agile, even with multiple products created by one team, are constantly growing. On the contrary, projects do not cost anything until they finish (and many do not).

Something to think about ...

Answer 6

As @Chris said, in a regular project there are subprojects inside. You have only one drive.

Think of the backlog of all your projects. First problem: do you prioritize tasks or projects? I prefer the lag in the project. At least to clearly define the priorities that the product owner has.

The presence of different owners of products and due to the fact that the owners of these products do not agree on how much time they should devote to each project. Someone will have to decide on how to manage priorities between projects. Note. Developers should not do this.

Here comes our project manager “S”, who will balance the resource needs of product owners and the time that team members can give. The owner of the product A fee for one month of work, but the owner of the product B always updates his project (and also pays well). There are some how you balance your team for A to have one month (developer time), and don't stop B from filling pockets.

In the case of internal software (one company, one thousand projects). Different product owners should agree on the time allotted to them. They do not live isolated, but in the same boat as you (project manager "S"). They will simplify your life in order to postpone some tasks, but at the same time you should never forget about your needs.

Ok, I'm still trying to find a better way to do this. But this is what we are pushing right now. Hope this helps.

Answer 7

Team members can share their time between scrum scenes, but much more so that team members are fully dedicated. Team members can also change from one sprint to the next, but it also reduces team productivity. Projects with larger teams are organized as multiple fights, each of which focuses on a different aspect of the product with close coordination of their efforts.

Answer 8

I think anopres was right: the best way is to avoid several projects at the same time as the fight. Do everything to convince that too much work in parallel is inefficient.

Suppose that 5 projects last about 3 months for a team with 5 people.

Approach 1: each person works on one project in a team

• 1/5 delivery speed per project gives 15 months of delivery for all projects.
• Each person is an expert, but only in his own project
• No team spirit.

Approach 2: 1 sprint for a project, switching projects

• Every sixth sprint works on a project
• Too much time between work on a project is not a regular incremental value for the project (for product lag yes), it’s easy to forget, the efforts needed to restore the context,
• The first project, delivered in about 12-13 months (subject to a 2-week sprint)

Approach 3: 5 projects in a single sprint

• It takes too much detailed breakdown of tasks to fit into the sprint
• Very little incremental build for the project
• Delivery of the first project in about 12-15 months

Approach 4: Recommended - Serialized Work

• The team is working on a separate project after the project.
• The first project was launched and sent in 3 months.
• The second project began after the third month, which was sent after the 6th month.
• ...
• The fifth project began after the 12th month, which was sent after the 15th month.
• Project-oriented team, intensive research and cooperation with clients
• The whole team has a general good knowledge of all projects.
• Lack of context switching time
• Good team collaboration is required (conflicts can slow down delivery).

As you can see, solution 4 is usually better, because projects are delivered much faster, the team works together and is effective. Other approaches include time to move away from context switching, lack of full team collaboration, very long total delivery time for all projects, etc.

What about the lag? If a team works on one project at once, it's simple - everyone will join. If there are several projects, we may need to delegate individuals to separate cleaning sessions (the full team is not involved).

It is important to convince customers that the launch of the 2nd project after 3 months will still lead to faster delivery (after the 6th month), and not to immediately start working with everyone else. This is an illusion that managers see - we start 5 projects at once, we work a lot and little by little. This is ultimately ineffective.

That’s why I don’t believe that a fight is effective for several projects in parallel, it’s very difficult to compare it with the framework and work in accordance with the rules of the fight. Sometimes it’s good that you have 2 projects so that all people are busy, but the more projects we add, the less effective the battle. Maybe kanban is an alternative to see progress and work in a team (not as much as in a Scrum team)?

Regards, Adam

Answer 9

I suggest launching one Sprint for each Project and having 1 daily meeting to launch all active sources / projects.

Answer 10

I would like to contribute. So I do it:

• The team has one product owner and one product. The owner of the product does not have to be one person, but the owner of the entity product is responsible for the backlog of the product.
• In the backlog of the product there are user stories of each project (all projects are here). Each user story has effort / story points (team responsibility) and business value (product owner responsibility).
• We have a “product care” that goes through every sprint. Before this meeting, the product owner updated the business value of the stories (if they need some kind of change for any business reason, we don’t do this and should not care) and will include some new stories. This meeting is about new stories, as well as efforts. This meeting lasts about an hour (except for the first time, about 4 hours).
• When we plan a new sprint, we open the product backlog, order ROI stories (this is business value / effort) and plan stories until time disappears.

What is it. I can say that it works very well. To do this, we use a couple of google spreadsheets (backlog for products and sprint-backlog, both with graphs, etc.), and redmine (with certain semantics) for online organization every day: time, task progress, etc. .

The problem with this approach is that I duplicate tasks in the sprint backlog and redmine mailing table. But I did not find an online tool for this completely on the Internet. I missed the backlog of the product in redmine (no other semantic works for me), the only board in jir, more stories in the taiga, etc.