Quick guide to business agility


The group of software experts that first defined agile development did so in a publication called the Agile Manifesto, its title perhaps echoing Marx and Engels’ Communist Manifesto.

Thirteen years on, the revolution which begun in a Utah hotel is gathering pace. Major commercial software developers, including Microsoft and Google, employ tens of thousands of programmers working to agile development principles.

Agile development is about breaking down a software project into a series of steps, called sprints, each of which result in usable code. Sprints lasting a few weeks include rapid testing and correction of code with the close involvement of the eventual user of the software.

Techniques for promoting teamwork are at the heart of agile development. Daily meetings, chaired by so-called scrum masters and sometimes conducted standing up, are held to review progress and set targets. Progress is often charted on large wall displays called radiators, which all team members can see.

It is all very different from the traditional waterfall approach in which an entire project is planned and timetabled from the outset and executed in a sequential manner.

“In the UK, waterfall is still the predominant method for software development, but agile is coming up fast,” says Gordon Morrison, director of technology for government at TechUK, the trade body for high-technology companies.

“In waterfall you are fixing the scope and varying the time; with agile you are varying the scope and fixing the time. You use agile when you want to develop software, but you are not 100 per cent sure of the outcome.”

Nowhere in the UK has agile development been embraced as enthusiastically as in central government and with good reason. For years billions of pounds have been squandered on IT projects that were never completed.

Earlier this year, the Government Digital Service demonstrated 25 online services that had been developed using agile techniques and celebrated the end of “big IT”.

However, some civil servants remain to be convinced. HM Revenue and Customs has been criticised by the National Audit Office for failing to adopt agile techniques in its development of the Universal Credit tax system.


The ideas behind agile software development have influenced the management of a wide variety of engineering, IT and product-development projects.

Agile managers adopt a flexible and interactive style of management that is particularly suited to small- scale projects or elements of larger work programmes, which are complicated and when outcomes are difficult to predict.

As in software development, agile managers are people focused. They assemble teams of experts, customers and suppliers to work incrementally, testing and refining their work in a process of continuous improvement.

Collaborative working is to the fore with regular meetings to review progress, discuss problems and find solutions.

“This approach can be leveraged effectively for non-software products and for project management in general, especially in areas of innovation and uncertainty,” says Jean-Loup Richet, research fellow at the French business school ESSEC.

“The end result is a product or project that best meets current customer needs, and is delivered with minimal costs, waste and time, enabling companies to achieve bottom-line gains earlier than via traditional approaches.”

There are strong similarities between agile management techniques and methodologies such as lean, kanban and six sigma, which have been applied with great success by manufacturing companies.

Unipart Group, for example, has built its manufacturing and logistics business round a methodology called the Unipart Way.

“The Unipart Way, our proprietary interpretation of agile, is the fundamental element of our competitive edge,” says Frank Nigriello, Unipart’s director of corporate affairs. “It is the philosophy that allows us to do more for our customers, and to engage all our people in innovation and problem-solving.”

Close study of picking, packaging and packing techniques has considerably cut the time it takes to dispatch orders in logistics.

Warehouses run by Unipart sport communications centres were performance statistics, targets and lists of problems to be sorted are displayed on boards. Staff attend daily meetings at the centres, and also take part in quality circles and receive instruction at onsite training centres.

And it works. For example, Unipart has been able to cut the time to diagnose faults in Sky set-top boxes by 90 per cent.


Price comparison website CompareTheMarket.com cut the time it took to develop new services a thousand-fold by adopting a more agile approach to data.

Instead of having to wait for information to be downloaded from a data warehouse, business managers can access it immediately, enabling the company to reduce the time it takes to add new features to a site from three weeks to 30 minutes.

Database management has been a conservative business with data owners understandably reluctant to do anything that might jeopardise the accuracy and security of their data, the lifeblood of their enterprises.

However, circumstances are forcing their hand. The volume of data generated by businesses is growing at 40 per cent a year, according to research company IDC. The company forecasts that with the amount of data doubling every two years, databases will be bulging with 44 trillion gigabytes of data by 2020.

That growth is being driven by an upsurge in online activity and new types of data gathered from mobile devices and sensors linked to the internet of things.

This data tends to be unstructured and sometimes inaccurate, but often needs to be dealt with in real time. Companies have had to find more agile ways of handling the data.

A first step is to reduce duplicated data and to pack it more tightly using techniques such as virtualisation and data compression. The parallel and in-memory processing techniques associated with Hadoop and NoSQL database software can relieve conventional data warehouses of the burden of handling large datasets.

Chief information officers (CIOs) are increasingly turning to the cloud, transferring the chore of storing and managing information to service companies for all but the most business-critical processes.

Moving to the cloud also allows them to be much more responsive to changing circumstances, turning facilities on and off as required. In addition, analytics as a service enables CIOs to mine information and develop rapid prototypes of new services.

“As many organisations in the industry move towards a model of continuous delivery, our database models must be designed to support the change,” concludes database expert Dan Wood. “There really is no intrinsic difference between application code and databases.”