In the US and across some European countries, 50% of all new app development requests end in failure, either by not being delivered or by being delivered but not meeting the business needs, says Appian. This is the fault of either slow coding or what is known as technical debt.
Appian, which provides a combination of automation and low code development, highlighted the issue in its Future of Work Survey Report #2.
It found that business requests for new software applications are soaring globally, with an average European, Middle Eastern or Asian company generating 230 internal requests for new business software applications and major feature enhancements every year.
But half of these fail because IT departments are struggling to meet business needs, because of the slow pace of coding and issues over “the implied cost of additional rework caused by choosing an easy solution now over the right solution”, called technical debt.
Technical debt is the cost of an easy solution now over the right solution
Technical debt reflects the “lost opportunity” cost of not developing the right application needed to take advantage of a market opportunity.
The Appian report says that IT organisations spend 50% of their time coding new apps and enhancements, but lose two-fifths of their development time to technical debt.
The effects of technical debt on business are higher operational expenses and simple software enhancements taking much longer than expected. It also causes reduced performance and scalability, a longer time-to-market and incomplete customer-experience improvements.
To beat technical debt, the favourite response from the survey was to “look for new ways to accelerate application development.”
European companies are more overloaded than in the US
European companies make significantly more requests for new applications than US companies reports Appian, but all are held back by technical debt and development complexity.
The effect is that 16% of new development projects in EMEA requested by enterprises never get started, 14% of new development projects in EMEA are started but never finished and 20% of new development projects in the US and EMEA are finished but don’t meet the business need.
The low code solution
Low code is part of Appian’s solution to technical debt. It is not actual coding, but the design of a flowchart in the Appian platform that brings internal and also external enterprise applications, such as databases, into its own process.
The workflows used to be carried out manually, where employees would open company app after app, such as in a call centre, or checking a customer’s mortgage or loan potential.
An employee would draw data from one app and apply it to another in a process that was slow and open to error.
The Appian flowchart automates the process and can also detect inefficient elements in the flow to show companies where legacy systems, old methods or applications that once paved the way to business standards, are slowing them down.
Appian regional vice president Siter Ali explained: “The most complex part of any organisation is people process and data, trying to bind that and bring them together is often the biggest challenge that anyone faces.
“Data is important to make informed decisions, you need the information in front of you, you need to be able to bring data together, and often it’s in silos [where a data system in incompatible or not integrated with the other data systems], bring it together and present that so that it’s meaningful based on the context.”
Once the data is presented, people can make decisions, or automation is possible where there is a rule to drive the process to its conclusion, such as offering a customer a loan.
Appian works with enterprise robotic process automation software company Blue Prism to deliver the automation.
Ali said: “What you’ve done is bind different technologies together, but the logic still stays in Appian, your flow is still in Appian, the orchestration that is required, it allows you to evaluate these into the business rules. It’s also evaluating using third-party technology. It’s still data. Then evaluate it against any of the rules I define, such as can I automate it or do I need to get a human involved?”