Sunday, 12 June 2016

Agile 101 book: Scrum’s advantages

Scrum pros and cons: blue pill, red pill
Scrum pros and cons: blue pill, red pill


I don’t want to become one of those Agile evangelists, preaching everywhere about how good and modern it is to be agile. But now that we’ve discussed the disadvantages of Scrum, we should look at Scrum’s advantages. I’m going to focus on one of its brightest features: regular deliveries. Thanks to these deliveries, Scrum will make the following possible:

Clients can start using their products

The client can start using the product even if not all the elements have been built. If 20% of the new features have been developed—which are the features that will be used 80% of the time according to the Pareto principle—users could start enjoying the product.

With the feedback of users after trying out one of the deliveries, we might realize that some of those features are far more important than 80% of the remaining tasks in the Product Backlog. Somebody could say “but the draught of the new regulation no longer requires a signed approval form” or “actually what we need is a button that would cancel the procedure” (we actually received these two comments in real life).

We can decide where we’re going

Businesses change, needs change, regulations change. And what was key when the project was signed might not be so important six months later. The client can pursue new goals, what to do in each new sprint and what we should focus on in our next delivery.

Divide and conquer

Colossal tasks require colossal efforts. If we must deliver just a part of that task every two weeks, our load will be lighter. Smaller, more manageable tasks make us feel that our job is easier. In each delivery, we have the feeling of having advanced one more step towards the final goal.

Fewer surprises

When we see our product grow, little by little, we all get an idea of what we are doing with it and if it is going to be useful or not. Also, we will know pretty accurately at what speed things are being delivered and how long it will take to finish up. Should we correct our course? We would know that in weeks.

Deliver what the client needs

One of the basic principles of Scrum (and also one of its main challenges) is accepting that clients can change their minds about what is or isn’t necessary. With Scrum, we strive towards providing a flexible response, admitting that clients themselves may not have put their finger exactly on the problem, and that often, as the project advances, they may come to realize what they truly need. That is why, among other things, Scrum is considered an Agile methodology.

The list of requirements and features created at the beginning is open and can be modified at any time. It contains rough estimations of the amount of effort that each feature requires. Before each iteration, or sprint, a group of these requirements is set as the goal for that period. Two or three weeks later, requirements may have shifted and the new goal might be in another direction, rather than the one that was set a few sprints back.

It is a new way of working, and a bit hard to adopt, for both the client and the provider. There are other, apparently easier, ways. We can always go back to the traditional formula. First we analyse the problem for months. When we’ve determined what should be built, we start developing it for a few months more. When we’re finished, we cross our fingers and deliver our final product to the client.

After this step, who hasn’t heard sentences like: “but this is not what I wanted, there’s this thing missing here…” or “no, it wasn’t like this, you misunderstood me” or “yeah, it’s fine, but I’m going to call the Director, actually he’s the one who has to sign it off” After months of work, of stress and rushing around to deliver by the deadlines, clients have not received what they need and we need to work even more to try and patch up the proposed solution.

How Scrum helps to build big projects

Revisiting Raúl Hernández post “Learning how to build cathedrals” I started reflecting on that well-known story about building cathedrals. In the story, two workers are chipping stone to build a cathedral and somebody asks them what are they doing. One complains about how hard and never-ending it feels to build the wall he is working on while the other one simply replies “I’m building a cathedral.”

I have wondered if it is possible to be agile when you are building something so big. I think that the answer is yes, you can. At least I can think of a few advantages of being agile when you are immersed in a project that is the size of a cathedral:

With a framework such a Scrum, you’ll always bear in mind the final goal and the features it should have. In the Sprint Backlog you will have a series of requirements such as the creating a ground plan, raising the vault, decorating doors and windows, and many other things that will give shape to the final cathedral.

But once the goal has been defined, we have to start chipping stone. At the beginning of each iteration, the team will choose which tasks can be started from the list of things that are pending construction (and that has been prioritized by our client). We will define a smaller goal for the next few weeks, so we’ll have a new goal, a much closer goal that will help us not to despair at the distance that still separates us from the final goal. It is the moment to pick up our hammer and our chisel.

When our pending tasks are divided into small groups to be solved in fixed amounts of time, they are timeboxed. This is a divide and conquer strategy. Plan your most immediate goal and what you will do to solve it without getting overwhelmed with all the work still pending. A Burndown Chart will also help you see how the amount of pending work is diminishing and that, however slow it may seem, you are making progress towards your final goal: the cathedral.


You can find texts like this and many other about how to manage agile projects in my book Agile 101: Practical Project Management (available on Amazon).


Translation by Begoña Martínez. You can also find her on her LinkedIn profile. Proofreading by David Nesbitt.

Sunday, 5 June 2016

Agile 101: Scrum pros & cons





“There is nothing more difficult and dangerous, or more doubtful of success, than an attempt to introduce a new order of things.”
Machiavelli

Disadvantages of using Scrum

How many technologies have quickly popped up and then vanished even quicker? They all promised to be revolutionary, the new silver bullet that will end every problem in the software industry. When we get as much hype as has been lavished on Scrum and Agile, our first reaction is to raise a critical brow.
If you are adopting Scrum in your project or company, it is best for you to be informed beforehand about its drawbacks:

The team might be tempted to follow the shortest route

When we have things to deliver every two weeks, and in the latest deliveries the projected functionality wasn’t ready and the speed of the team does not indicate that we will finish the next one on time, we have a problem. We can succumb to the temptation of finishing the pending tasks in any way we can and leave a “technical debt” behind us. Everything has the appearance of going well, we create more features, we start new things, and the client is happy because deadlines are being met.

But every time you leave a small hole of a deficit behind you, I can guarantee that you will stumble on it again and make it bigger. If new things keep caving in this way, the “debt” will become a pit. Sooner or later you’ll have to stop everything and fill in the pit—pay back that loan and with extra interest because of that late payment. The project doesn’t finish when we were so close to the end, and the burndown chart appears to have a horizontal asymptote at 0 (sorry, Calculus I).

Do you need to have delivery deadlines way in advance?

This is one of the main criticisms that Scrum normally receives. In a way, it is logical that Scrum cannot give you those deadlines. At the beginning of the project, you cannot predict when you’re going to finish, if you’re making it easy to change the thing being built as time goes by. But, do you prefer a product that you know “for certain” will be finished in 12 months, but is built over the ideas and opinions you had a year earlier? Maybe you prefer a product that could be finished in a similar period of time, but that you have been able to steer towards your real needs, and you’ve been able to use and test before the final delivery deadline.

Stress!

We cannot sprint for our whole lives. We deliver something and the next deadline is just two weeks away. Then another one, and another one. If we have to run like this for many miles, we will sprint really hard for the first deadline, but we will walk for the last one —and that, with luck. It is something we should bear in mind during planning, from the start.

Is your team able to self-organize?

One of Scrum’s principles is that the team members must make their own decisions and self-organize. Also, one of the requirements is to avoid having teams that focus just on specific tasks, like analysis, testing, design, documentation, development etc., but rather to make sure that every member of the same team can carry out all of these tasks without depending on external teams or members. Can we always have a team like this? What happens if we don’t have one, or if we lack a key component?


You can find texts like this and many other about how to manage agile projects in my book Agile 101: Practical Project Management (available on Amazon).


Translation by Begoña Martínez. You can also find her on her LinkedIn profile. Proofreading by David Nesbitt.