Tag Archives: stories

Agile: Scoring

Once you have your stories defined and you are ready to “score” them, there are some things you need to keep in mind.

What scale are you going to use for scoring?

I use 1,2,3,5,8,11, and 13. Why? Well, that is just what we used when I started doing Agile, and it seems to work :) Fibonacci + 11. I know I probably should change it so it kind of goes along with the more mainstream numbers that some 3rd party tools and other people recommend. Maybe, someday :)

1 is the lowest, 13 the highest.

At first, the team has no idea what a 1 or 2 or 8 or anything is. It takes time. Over time the team gets a gut feel for what a 1 is, or a 3, or whatever.

Don’t equate points to time! No, a 1 isn’t an hour, or a day, or X. It is a 1.

What are points?

Well, I try to explain it as time, effort, complexity, outside factorness, craziness, etc, etc. You kind of have to bundle all those things up and say, yea, that is a 2. It is more than a 1, less than a 3. Each team is going to have a different meaning for what a 3 is.

During your sprint planning, what you want to do is give everybody a pack of “planning poker” cards so they can score as you go through stories. (There are also web apps for virtual teams, iPhone apps instead of using planning poker cards). As you get done reading the story, and answering any questions, the team should “score” – show their cards. If they are all the same – great! You have a winner.

If some are different, then you discuss or come to an agreement on why some do’ers think it should be higher or lower, and get a score at the end. Example? 6 devs, 4 say it is a 2, 1 says a 3 and 1 says 1, well, most say it is 2, do the other 2 devs agree? if so, you can mark it a 2. If not, you might discuss why. The one dev saying it is a 3 might say, “what about this an that” and then everyone realizes, oh yeah, it is more than 2, and you go with a 3.

What happens if you see 13’s?

What I try to do is say – if you score it a 13, then the story is too big, you should break it up into smaller stories. With that logic, you should never have a 13 on your board.

The biggest thing to remember is that the team is going to have to take a few sprints to get used to the scoring, and understand how stories fit into points. You ARE going to over and under score – it is just reality. But don’t adjust the scores during the sprint. Things just tend to work themselves out over time. What you can do is capture actuals at your retrospective and compare to what the original scores were. The team might get a better idea on what they should be scoring things then. Another thing to try is a scoring guide. After some time, you can say “every story we do that involves making a new button on a form is usually a 2” so then if you have a story like that, and the team scores it a 8, or they don’t know what to score it, you can look back at your guide and say -hey, we usually score these 2’s, so is a 2 ok?

Who scores?

This is a good question, who should score? My opinion? Just the do’ers. The scrum master shouldn’t score, and product owner/analysts shouldn’t score. Just the do’ers.

What do you do with the points/scores?

You track your daily burndown, and your “velocity” per sprint. More on that later.

Just remember, think of stories as being scored by points. Or bananas. Or stars. Just not hours :)

Advertisements

Agile: Backlog

The product backlog. Not a ton to say on this right now, but here are some thoughts.

  • The current sprint is not backlog
    Seem’s obvious. But the backlog is the backlog. The current is sprint is what you are focused on and working on.
  • The next sprint is backlog
    yes, the next sprint, on your board, is still backlog. It is somewhat solidified, but it is in flux until when you start the next sprint as your current sprint.
  • Your backlog should be semi-full
    having nothing in your backlog means.. well, you have no backlog, and that isn’t good. You should have enough stories for 1-2 sprints. Having no backlog is like living paycheck to paycheck.
  • You should prune old stories
    Older stories that have been just living in your backlog, should get pruned. Either they are higher priority and should get done, or they aren’t worth doing. They might resurface later as something but they shouldn’t be carried along for months/years if they aren’t important enough to do.
  • Priority matters
    The product owner should make sure they have the backlog prioritized. An backlog that isn’t prioritized is just a stack of cards.
  • Split Backlog into buckets
    I like to split backlog into buckets. Technical vs Product, maybe Features vs Bugs, some way to differentiate things.
  • Try to get as much as you can scored
    You might want to try to score as many stories in the backlog as you can so you are always ready. Of course there is a balance. You want a full backlog, you don’t want 5000 items, and you don’t want to take 3 days scoring things. Score enough so that if you do get ahead you can take stories from the backlog to work on. Try to avoid scoring outside of sprint planning sessions.

The backlog is one of the most important parts of your Agile process. Without a backlog you don’t have anything to focus on as you move forward through iterations. Try to think of it as a well, a well you don’t want to dry up, but you don’t want to overflow either.

My preference is to have a product backlog and technical backlog. That ensures that technical debt isn’t ignored, but separates concerns as far as features that you might want to implement.


Agile: Stories

In Agile, there is the concept of “stories” or “User Stories”. What are stories? Well, maybe it is best to look at what they aren’t:

  • They aren’t projects
    Stories are smaller than projects, by a long shot. A group of related stories is called an “Epic” but even then an “Epic” isn’t a project. But again, it all depends on how you or your organization defines a “project”. An agile team can focus on one project throughout their entire lifetime. Another might work on multiple projects, it all depends on how things are set up. But stories aren’t projects. Projects are made up of stories.
  • They aren’t tasks
    Stories should be bigger than tasks. Each story should really take a few to many tasks to complete. You can get into the mindset of breaking up everything into tasks, but then you might be going to granular. You want your stories to be small enough to get completed within your sprint, but also big enough to not take 10 minutes to complete.
  • They aren’t direct requests
    What do I mean by direct requests? A direct request from and end user or customer doesn’t directly relate to a story, it usually ends up being multiple stories. In the case where the request is just one story, it probably should be reworked to fit the mold of a story.
  • They aren’t features
    They might be feature requests, but like “direct” requests, usually a feature request ends up being multiple stories, or if the request is small enough to be one story, it needs to be reworked to become a story.
  • They aren’t bugs
    same as direct requests and features.
  • They aren’t day to day operational actions
    Some people might get into the mode of “well, we do this every day, so it should be a story”. No… it shouldn’t. It should be what you do day to day. Your velocity will get adjusted by your do’ers doing things day to day, but you shouldn’t storyize day to day things. Would you create a story for your commute? For reading email? For lunch? The only thing you might want to storyize is an architecture type meeting, where the deliverable is a bunch of stories. I like the idea but I’m not 100% sold on it yet.

Well, so there is a short list of things stories aren’t. So what are they?

As a user/group/xyz, I would like new feature/bug fixed/thing done, for this reason/business value.

I would say the most key part of that story format is the end. The business value. Without that you are spinning your wheels, you don’t know why you are doing something.

What about things that go along with a story?

You might write the # of a record in your virtual system to easily correlate. Depending on your process, you might hand write cards (I personally like this method). Or you might print them. What about other notes? As a developer or do’er you might write things on the back of the card for when you pass to QA or the validation of the story, or just to keep things in the open. If you use a virtual system, you might have extra notes or background info so the do’er can actually do it.

I like to have a small title, and then the story.

Example,

As a sales executive, I would like to be able to see my sales and goals by product and year so I can better analyze my commissions.

A title might be “Adding goals/sales to cube for sales”

I also like to have a “component” when working on a team with multiple mini projects that might be getting done. “Sales Cube” might be a good component for the story above.

What else do stories have? Priorities and Scores. More on that later..


Agile: The Story Board

The story board…

If you have used Agile for Project Management or Software Development, and used Scrum, you probably have heard of the “Story Board”.

Trek BI Agile Story Board

Funny thing though, I have talked to a few people that have came through and they saw the boards we have and it was “oohh I have never seen one in real life”.. but yet they have run or been involved with scrum before.

So what is the deal with the “board”?

  • The Board is a focal/gathering point
    Every day at your standup/scrums, you should bring the board in and use it as a tool. Also at the sprint planning. Think of it as another team member. Talk to it. Talk in front of it, analyze it, adjust it, show it off, explain things on it to people
  • The Board shows tangible evidence of results
    You should hang your velocity and burndown charts on the board, you should have a nice full current and next sprint that changes every day as people do things and new items get prioritized. Your backlog should be filling up. Communicate your iteration dates and maybe other key info, but most important, if someone walks up, they can see your burndown.
  • The Board is a talking point
    People around the office start asking “What is that thing?” People in other teams want to know what is up. Other people might come up and ask where things are at, or where their “story” might be. You can use the board to start discussions on agile and also about your project
  • The Board let’s your focus
    As a developer or “do’er” for your project, you can focus on just the current sprint. You can focus on the cards on the board. You see slightly what might be coming in next sprint, but you are focused. As a product owner or team lead, you can see progress and focus on the current 2 weeks.
  • The Board makes it tough for distributed teams
    If your team has some members off site part of whole of the time, then it is harder for them to actually “use” the board. They need a virtual board as well. They need virtual stories or some way to update progress.
  • The Board adds a little overhead for Scrum Masters
    With a online system or virtual system, the Scrum master can update things there, and so can the developers. With a board, and especially with a board + virtual system, the scrum master has to make sure things are in sync. What should they focus on? The board or the virtual system? Both. But since the board is visible to EVERYONE who might walk by, they need to make sure it is up to date and in working order.

My thoughts: The board *should* exist, you *should* use a board.

But, what if you run distributed teams? Then.. if majority of them are not on site, then use a virtual board. If it is a mix, use a physical and virtual board.

But what if we don’t have room for a board? Then make room.

But a board costs money. Not really. I made one out of things lying around the office. Minimal cost. You can expense story cards, etc.

But xyz.. well, I am sure there is an answer.

If anything, start with a board, and cards (another blog post on that) for at least one year, and then look at going virtual, *if it makes sense*. I still think the use of the physical board is a key part of the process, but like with anything Agile, the process always differs by team and project.

You can always roll the board somewhere to demo. You can show the board to anyone. Any worker, any executive, a group of many people, or small. You might not be able to do that with a virtual system: you might be limited by number of accounts or other system limitations.

The board might be a “relic” of agile, but I think it lends itself to running the process in a way many can understand, without having to learn any software, and also lets your team communicate progress and results easily. Use a board? Just do it.