Has agile got a higher success rate than waterfall?

The first thing to recognise is that there are two very different environments for digital transformation:

1. Greenfield i.e. a new/independent service which relies very little on any other system or service
2. Brownfield i.e. established businesses with processes and systems that are running today and for which maintaining a consistent service through the transition is paramount

Greenfield is simple and frankly you should ONLY be considering an agile approach because it provides early validation that your service works and there are customers who want what you think they want.

In Brownfield environments building software and solving complex problems ALWAYS takes longer (and therefore costs more) than we think it will at the beginning because we only have a high level view of the challenge ahead. I would liken it to estimating the cost for building a house by looking at blueprints, there are too many unknown variables for a builder to accurately predict timelines and costs. Yes, you can get a high level estimate but only when the builder starts work will they understand many of the factors and constraints that drive the true cost and timeline.

Decision makers need to recognise that having an experienced IT professional say with confidence that a project will take X years and cost £Ym before they have started doesn’t mean those costs and timescales are guaranteed or fixed. With every transformation the key variables change:

• The problem to be solved is different… different goals and priorities, different users, different systems to integrate with
• The technologies used in the solution will be different or will have evolved since you last used them (cloud use and the evolution of AWS is a great example of this)
• The team delivering the implementation are likely to be different which means new interactions and collaboration styles, different skills and personnel challenges
• The office environment and project infrastructure is likely to change i.e. the tools and techniques you use to manage delivery
• The market context and external pressures are likely to have changed

Which all makes predicting timelines and costs based on past experience incredibly difficult.

But importantly decision makers must agree “What is OUR measure of success?”. Is it that a programme of work came in under budget and on time or is it some measure of the impact that programme made? Finishing a 3 year transformation is great but if there are no efficiencies or gains in customer loyalty/marketshare then was that the best investment? The real discussion is not about money, but rather it’s about value and outcomes.

There is little credible, independent research available for the failure rates of transformation programmes. Why? Because large organisations, especially those that are listed or provide a public service, are careful about their external perception and are therefore more likely to claim a failed programme delivered certain benefits before a change in strategy caused the company to “alter their approach”. In reality it doesn’t matter whether MOST waterfall/agile programmes fail but whether YOUR waterfall/agile programmes fail.

Now, if you want to measure success by outcomes then agile is going to deliver a higher chance of success when properly executed:

• Early deliverables mean early feedback from real users which has a number of benefits, it proves your theory was right and that people want what you’re delivering (it’s amazing how often it turns out that our early assumptions are actually wrong), it starts a feedback cycle meaning you can make relatively small, quick changes to improve user adoption and make solutions easier to engage with, this in turn makes your service more popular. Feedback also helps to identify new opportunities and services, potentially creating previously unconsidered revenue streams.
• The earlier you deliver benefit, the longer it is realised, rather than holding back benefits until the final “big bang” of delivery. Even better, as the programme continues to deliver incrementally benefits will rise.
• Risk is lower, change is anticipated. For waterfall programmes the cost of delivery builds until the final deliverable and changing direction half way through is expensive however, with agile there is an opportunity to re-set and change direction at every delivery point.
• Short term goals help focus teams, not only that but they have a significantly positive impact on motivation as teams deliver something so can see the result of their hard work. Working on a programme for 3 years before anyone even looks at the outputs is a huge leap of faith that most of us struggle with.

But one word of warning before you schedule an early morning “all hands” to tell your teams that agile is the new and only direction – bear in mind that agile programmes fail too. The truth is that the method you use is far less likely to influence success as much as having:

• experienced leadership,
• a single vision with measurable outcomes and goals,
• resource management processes to ensure you hire capable people and deal quickly with under-performers,
• supplier management processes so that your suppliers are delivering value and driving the transformation forward at pace,
• strong governance which enables rapid decision making and issue resolution,
• solid reporting and controls for delivery progress but also for day to day aspects such as financial management

Neil Butler

You might also be interested in these

Feb 17, 2020

BY Andy Astley

Uncategorized

Home Office: Building a Custom Cost Analytics Service for a Self-Managed Kubernetes Environment

Digital Dimensions presenting on the AWS ‘this is my architecture’ series

read more

Feb 17, 2020

BY NEIL BUTLER

Culture, Delivery, Teams

You just can’t get the staff these days

I have a theory that people want to be carpenters. What do I mean by that? When you look at why people come to work and ignore the obvious point that we all need …

read more