I was very interested to read this post about agile colocation on Artem Marchenko's Agile Software Development blog. The comments are also very interesting.

As always, I don't think the colocation debate is black and white...

I've read plenty of articles and blog posts about the merits of colocation. I've also read lots about how colocation is an essential ingredient of agile development. And I've read some very good counter-arguments explaining how agile can work - and is working - with teams that are not located together and even working across borders.

I think the debate, as with most things depends on your circumstances.

In our case we have business teams in multiple locations. So is it best to locate the associated development teams in the same building as the business units, or in a central development group colocated with their technical peers and other people of a like mind?

There is certainly a case for colocation with either group, but they can't be in two places at once so we have to choose!

In situations like this, I think it comes down to which group of people is most valuable to collaborate with most frequently.

If you have an established product or business in a BAU (Business-As-Usual) situation, but a development team whose development practices are not well established, I would suggest that ongoing product development is best placed near other, more established development teams, where good development practices and experience can be shared amongst peers.

If, on the other hand, you have an established development team with good development practices, performing well and delivering what's expected, but a product or business unit that is not in a stable situation (e.g. new product development, competitive threat, major problems, etc), I would suggest that colocation with the Product Owners is much more important. In this case, active user involvement is imperative, in order to provide extra visibility, get very regular feedback and customer insight, and be able to respond in Sprints accordingly.

In any event, I do think distance adds to the compromise, wherever you are permanently located. I've made comments before about this in relation to Agile India. If the opportunity for face-to-face communication is virtually nil, then I think the compromise is a really big one.

Kelly.

P.S. Click one of the icons below to join the growing community of people reading this blog by RSS or by email...

read more

I was doing a bit of research on Google searches and I noticed that a large volume of searches for agile development are coming specifically from Bangalore in India.

And this got me thinking about agile development offshore.

Of course I realise that India has a large number of software developers, as a result of the trend for offshore development in the last 10 years or so.

But in my agile travels, I've come across very little research, success stories or case studies about agile development in India, and particularly in an offshore development model where the development team is in India and the client is not.

I can see clearly how an offshore development team in India might operate internally using agile principles and practices. But I'm very unsure about how it would work effectively in terms of (a) active user involvement in the development process, (b) defining lightweight requirements piecemeal throughout development, (c) integrated testing, etc.

Perhaps the challenges are no different than for a distributed team? But when I think of the added complexity of a distributed team, the added complexity of client/supplier relations, the added complexity of distance (preventing regular face-to-face communication), and timezones inhibiting other forms of communication, I worry.

Add all that to the usual complexity of software and humans, and I wonder how high the barriers to success really are with this model.

Martin Fowler from ThoughtWorks has written an interesting article about 'Using An Agile Software Process With Offshore Development'.

If anyone has experience of agile offshore, positive or negative, in India or elsewhere - or if anyone has any interesting reference materials about the subject, it would be great if they could share via the comments on this blog post...

read more