David Norton

A member of the Gartner Blog Network

David Norton
Research Director
7 years at Gartner
25 years IT industry

In his role as research director with Gartner's application development and architecture team, David Norton supports clients by developing and delivering quality and timely research. Read Full Bio

Coverage Areas:

Will 2011 See Our Love Affair With Scrum End?

by David Norton  |  January 16, 2011  |  2 Comments

If you  talk about agile to a developer you often hear the reply “oh you mean Scrum” – the two have become synonymous.  By any measure Scrum is the most well known of the agile methods (is Scrum a method is for another blog), search results, blogs, books or surveys – its top of the agile hit parade.   There are a lot of reasons for this; first it works when done right and within the right type of organization but success’s is only part of the story.  Other methods like FDD, Crystal have had their success too but do not have the same status as Scrum.  A strong community has helped to popularize Scrum, but DSDM has a strong community however it does not come close to Scrum’s fame.

So what else is there? Well it comes down to a mix of marketing, push for certification, consultancy self-interest and let’s be honest, its cool.  Scrum is the Apple of the methods world, what I mean by that is yes it does the job but often coolness and peer status plays a role in adoption. (Just in case your thinking I am knocking Apple I just switched from PC to MacBook Pro and am very happy even with the dent in the lid after I dropped a radio on it).

The big question is does it matter how Scrum has been popularized as long as it is moving agile into the business and bringing success?   I think yes.

If a single agile method get’s a monopoly it has the potential to stifle innovation. But the chance of any one method getting a monopoly is unlikely; there are just too many variables to ever have a single silver bullet approach agile or otherwise.

My bigger concern is I am starting to see a lack of due diligence when selecting an agile approach.  Scrum is being adopted within many organizations by osmoses.  That’s fine if the organization culture is a strong fit for Scrum but if it is not we reach a crisis point, a point where Scrum fails.  Now this is true for any approach if misapplied but the hype around Scrum often means its adoption is not questioned until there is a problem.

“I don’t know why we are having problems we are using the best approach!!” I hear client’s cry.  And when I ask what evaluation process they used before adopting Scrum 90% of the time I hear “none”.  They adopted Scrum because it was already used ad-hoc within the organization, or they had some ScrumMaster certified developers, or Fred-Blogs consulting recommended it.

As agile becomes more and more poplar and is applied to bigger projects, outsourcing, package development and legacy the “Scrum effect” will start to be a real problem. In 2010 I saw a number of failures of Scrum in SOA projects and two SAP implementations.  None of these failed projects had done a good job of the method evaluation; they where driven by a real business need to deliver fast and in their haste jumped to Scrum. These failed projects would have been better suited to DSDM, FFD or Agile Modelling given the organizations types, architecture and technology.

So will 2011 see a move away from Scrum, no. But it will see more organizations run into problems with Scrum and either result in a hybrid approach or worst-case dropping agile.  It would be to easy to say a Scrum failure is a result of the organization not implementing it correctly and therefore not a failure of Scrum at all – the classic “It was not Scrum that was the problem it was the company”.  And yes many Scrum failures and issues will be down to the organization not really embracing Scrum practices but just as many will be down to selecting the wrong method.

So before you go down the Scrum road ask yourself are we adopting it for the right reason, and will it work for us? And if the answer is yes (and in most cases it will be) you have at least asked the question and looked at the options.

2 Comments »

Category: Agile Application Development IT Governance SDLC     Tags:

2 responses so far ↓

  • 1 Thomas Murphy   January 17, 2011 at 12:13 pm

    I agree completely. One key reason for Scrum adoption is also straightforward simplicity. If Agile is Scrum alone it doesn’t create nearly the need to go through the practice and behavioral changes that are required and it is a natural step from iterative planning.

    I don’t think it is a matter of Scrum being bad, just that scrum itself isn’t enough. In addition as you say, if you haven’t first figured out what is “wrong” with your current AD efforts, the change will be a pointy hair boss syndrome.

  • 2 vic williams   January 18, 2011 at 11:14 am

    Scrum by itself is useless for software development. Its just as useful for running a church or any improvement project or new product development. I see some workshops now advertising such as: 1) scrum followed by 2) Crystal etc. That makes a lot of sense, and perhaps beyond simple logic, into good teamwork producing what people really need.

    One foundation for scrum was the Nonaka et al “new new product development game”, and a lot of the concepts in that publ are still apt – the foundations are still there.