Andy's Mind

A question I frequently see asked on the Internet is whether agile teams should release every iteration (or even more frequently) or not. When asked in relation to Scrum the question is usually if software should be released every sprint. With other methods – like now-popular Kanban – the question is what release frequency would be right, as the method itself doesn’t say anything about it.

I think that many times this question reveals an underlying confusion between keeping software releasable and actually releasing it to the public. I think one construct from Scrum can help clarify things here.

Scrum calls explicitly for a “shippable product increment” to be delivered at the end of each sprint. It means that the end of each&every sprint a new version of the software must be ready that passes the quality criteria (in Scrum expressed as the “definition of done“) and ideally includes all the features the team committed to delivering at the sprint’s outset. However, this new version is called “shippable” not “shipped” for a reason. This reason is that whether to release a new version of software to the users is ultimately a business decision.

This distinction is very important. Just like the shape of the backlog is a result of a whole series of business decisions made by the Product Owner the decision to go with the product to the public is a decision that can be (and quite frequently is) made by stakeholders. As such it is definitely out of the scope of any process/method managing the development itself or the team doing the development.

There are of course numerous benefits from actually releasing each and every “shippable product increment“: greater user satisfaction, faster user feedback, better discipline in the team and less risk (deploying smaller changes is less risky than deploying a big-bang, all-changing release). In fact this is what we have been always practicing, also when working on our Scrum tool.

However it has to be understood that situations (companies, products, etc.) vary and can be very different. Sometimes there are other factors involved – for example domain (in embedded software releasing every sprint may not be even possible), market influence (eg. developing a major feature and not wanting to let on before it is complete) or alignment with other elements (like when software is part of a larger project, where other things have to be in place for the new features to be used or usable).

So, a team can be agile even if what they develop is not released every iteration. They just have to work as if it was in terms of quality and effort spent testing the software to ensure that it could be deployed when needed.

Someone has asked if management or project management should come to the sprint retrospective ¹. Think the way this question was asked indicated that there was some confusion regarding sprint retrospectives versus sprint reviews which, I think, is worth clearing.

Sprint retrospective and Sprint review are two different things that shouldn’t be confused.

Sprint review is for everyone involved, especially stakeholders, to inspect where the project is and discuss how to adapt as needed. Sprint review revolves around what was built – the “shippable product increment” produced in the last sprint – and the overall product, not how it was produced.

It is good if Product Owner “represents” stakeholders, but it is even better if they come and see themselves what was accomplished, what runs etc. My advice is to welcome management of all kinds if they want to come to a sprint review, just being sure they know what the purpose of the meeting and their role in it is. Ensuring that and educating them is primarily Product Owner’s job, but of course the Scrum Master may assist him.

Sprint retrospective is primarily for the team to inspect their last sprint, concentrating less on what was done than on how it was done, and then adapt their way of work. I wouldn’t include anyone outside the team in those, besides maybe the Product Owner if he/she wants to join.

A common objection to bringing management of any kind into retrospective is that team may not be comfortable talking about their dirty laundry in front of them. It is indeed very valid – but it is also worth noting that from managements’ prospective this would be a waste of their time to, for example, listen to developers debating how to improve branching in their code repository. Even if the management knew what the heck the team is talking about this is not something they should waste their time on. Managers have lots of things to do (collectively called “bigger picture”) which no one will do for them – this is where their time will be better spent.

Having said that an overall retrospective on the project or on a longer chunk of it (like a quarter or half a year) that would include management may make sense, but it would not necessarily include all team members (if you have many teams that would make it even impossible to do). Such a retrospective would concentrate on “big picture” and could be very beneficial – if there is of course a right atmosphere within company for people to be honest enough in such a retrospective for results to be useful.

Speaking of retrospectives – definitely worth buying is “Agile Retrospectives” Esther Derby and Diana Larsen. There is not much to read there – just a couple of introductory chapters – but it is a great cookbook of various techniques to use in different phases of a retrospective based on how much time you have and what is the retrospective about. Each technique has a description of how much time to set aside for it, how to facilitate it and where its place is the overall sequence of a retrospective.

Great help, since classic “what we did well? what we didn’t do well?” etc. becomes boring pretty quickly. Anyone who facilitates retrospectives on a regular basis should have this book.

[1] – original question.

David Harvey wrote a piece on his blog entitled “The Scrum picture is wrong” where he says that the well known, canonical even, picture of Scrum loops errs by focusing too much on the product (deliverable, “potentially shippable product increment”) and forgetting that continuous improvement of the team is another important outcome of each sprint that should be shown with another loop.
[Go and read David’s piece now if you haven’t yet]

As much as I agree with David’s insight I don’t think his version of the Scrum loop should be used to “sell” Scrum outside of our community. I very much value all the focus on teams and their improvement, but it helps to understand that this is something that is important (and should be!) only for us, sitting deep in the software development community. Clients ultimately pay for products, not for our methodologies, our teams improving or our overall well being and happiness.

Let’s take the case of outsourced development, which I know firsthand. Our clients pay us for the product they get from us, but once the project is done they are gone and I don’t think my team getting better on their project is something that even crosses their minds. And why should it, anyway? Unless they would want us to extend their product or start another project with us there is no benefit there for them. What really counts is whether the product we delivered will allow them to meet their business goals, their commitments – and their bottom lines. So when I work to convince them to forget fixed bids and go with Scrum I don’t waste the attention they give me on telling them that thanks to Scrum my team will get better.

But also if you have your own, in-house teams, building your product then in the long term it is that product that counts more than the teams. Why? Well, because not only your clients don’t pay for your teams, but for the product – you also don’t really own your teams. People change jobs (one call from Google’s recruiters to your best people can make your life very hard, believe me), get sick, even die – that’s inevitable. And (as I have learned the hard way) pampering your best people won’t prevent them from quitting – so is probably not worth it. In the long run – measured in years – it counts how much value your clients get from your product, not how perfect your team has become. If your team is getting better with every sprint but your product doesn’t sell you will hit the bottom hard sooner or later. Yes, your company may be then one of those legendary places that excelled technologically but are no longer around (Commodore, SGI etc.) which may be fine for the employees – they will just change jobs – but is not fine for owners and/or investors who will have to cover the loses.

So, team getting better is a tool, a mean, to get a better product, not the other way around (unless you run a monastery – there indeed work is just a tool for monks to perfect themselves spiritually).

Just to make sure no one misunderstands me: I’m not saying we should forget about team improvement, not care about our workers’ well being, career development etc. Yes, we should do all that and more, help everyone excel in what they do, make sure as much as we can people like what they do, heck, do it with passion and care, using fully their potential. All that is important and valuable. But we should not forget that in the end it is the client paying for a product who makes it all happen. And all our methodologies, frameworks and diagrams, all our “management science”, all techniques and research is aimed at improving efficiency – and that means, sorry to be blunt, getting better products faster and cheaper.

So I think the current Scrum picture is quite right in its focus on the product. And the message is spot on. It should be still used to “sell” Scrum to managers, clients and businessmen. David’s diagram is insightful and fine, but let’s keep it within our world of process freaks, agile activists and scrum preachers.

I have started a videocast with a friend – and manager of a competing software company – Paul Klipp. We have been meeting and discussing agile software development, web applications & related stuff for some time and recently I realized we could turn this into a podcast. Paul has one of those neat Flip Mino HD cams and what was to be a podcast has turned into a videocast. So now you can listen to us discussing those topics bi-weekly under “Scrum for Success” (also available on iTunes).

The last episode covers the FOWA conference we both attended that I have blogged about last week.