New ways of delivering software are necessary to compete in today’s dynamic marketplace
The recent past has brought dramatic advances in technology that have eliminated entry barriers for new businesses, revolutionized our expectations as consumers, employees and citizens and continue to improve efficiency and automation all around us. We believe this trend is going to not just continue, but accelerate.Many businesses are seeing the writing on the wall: To compete in this environment they must innovate relentlessly, bring new technologies to market faster, with more flexibility and higher quality. That calls for a new, fundamentally different way of delivering software that is foreign, yet achievable, not just for start-ups but for established companies as well. We call it liquid delivery, and we break it down into four critical phrases:
Ideate—Generate new ideas to solve specific problems by applying technology innovation.
Try—Experiment with new ideas by prototyping new products and experiences.
Build—Engineer and release products quickly at the highest-level of quality while infusing automation into existing processes.
Advance—Improve products in the marketplace through continuous feedback and ongoing enhancement.
|Liquid delivery is entirely fluid.|
Unlike traditional software delivery, the liquid approach enables constant tinkering. If you learn something in the design phase, you can go back to the ideation phase and adjust the concept accordingly. Or, if a concept is not working for your users during any phase, you can go back and ideate something different. The starting point should always be a specific business problem, but the process is never truly finished upon development of a solution. In other words, anything can be improved upon at any point.
Each phase is of critical importance to the others. During ideation, the objective is to generate ideas to solve a specific problem, or define new opportunities for growth. It is oriented to the customer experience or business objective, rather than to a specific technology innovation.
The experimentation phase offers a chance to try out ideas and gain experience with a select group of customers to learn how they respond and how the design needs to adjust. This phase can even be executed using low-fidelity resources such as wood replicas of products—or cardboard cutouts.
The build phase is a chance to re-ideate, based on user feedback. Speed is crucial. This is one reason why DevOps is gaining prominence as part of software design and testing—it applies automation to eliminate manual build/test/deploy processes, saving huge amounts of time.
The next (but not final) phase, advancement, offers still more opportunity to improve a product. With a liquid delivery model, just because it’s in the market, doesn’t mean it can’t be tweaked. Automation plays an important role during this phase as well, because it is instrumental in scaling technology quickly and efficiently.
But that’s not all
Beyond these four phases, companies need to rethink culture and technology itself. For example, companies should scrutinize whether organizational silos are hindering speed and accountability for quality and business outcomes. Instead, they should look at their products holistically, embedding an end-to-end ownership culture from idea through to customer experience.
Successful liquid delivery requires a great deal more interaction between users and developers than most companies typically plan during software design and deployment. New methods and tools are needed to operate in a more iterative way and to manage collaboration between teams.
Finally, new technology innovations may be required, such as flexible, lightweight architectures that keep the development and release process nimble, microservices platforms that allow rapid assembly of applications, and prescriptive architecture principles separating components consistently and clearly.
While many see the investment in iteration and rigor through liquid delivery as incremental cost, it is actually a hedge against having to redo work later, so it can lead to cost savings in addition to flexibility. In other words, going liquid may seem higher in terms of cost at the estimate stage, but in reality the comparable waterfall method will be more costly in the long run due to hidden rebuild costs required after project completion to address the rapidly changing needs of the business. Daunting as it may seem, going liquid is absolutely do-able.
Max Furmanov leads Accenture Liquid Studios globally.