Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

Is fail fast the next trend in app development?

Fail fast -- the idea of embracing failure as an option -- is set to be the latest business trend of 2015, according to recent predictions. Are you ready?

One way to guarantee being a trendsetter is to name the trends. Software AG did just that in its recent integration...

outlook for 2015. With his eye on the Internet of Things (IoT), integration and how they need to work together, Software AG's Senior Vice President of Products Mark Herring discussed elements of the list in an exclusive interview with TechTarget.

With his penchant for publishing annual predictions, Herring is a regular contributor to Software AG's blog series Reality Check. Prior to joining Software AG in 2012, Herring served as vice president for software infrastructure & MySQL at Sun Microsystems, technical support engineer at Oracle and programmer at Standard Bank Group.

It's not often a software company declares 'failure is an option' as a major trend.

Mark Herring: "Fail fast and adapt" is how businesses can differentiate themselves. The perfect example is Google. They are supporters of Agile methodology -- it's great. Put it out quickly, but if it fails, get rid of it quickly. Google Reader was out there, then they took it away. If you fail fast, you can get feedback from your customer quicker and make changes to improve or replace the product.

The expectation of failure would not seem to be a viable, long-term career strategy.

Herring: Of course. Not everything can be allowed to fail. But if you try to build the perfect app for the next 20 years, that's not likely to be right. IT hasn't learned that. IT has to learn to be nimble. You're much better off to put something out quickly, get customer feedback, then keep or throw out. It's not the number of application deployments by which you measure success, but rather the revenue they generate and their overall impact to the top-line business.

You've identified 'Spend, baby, spend' as another key 2015 trend, noting that IoT and software as a service (SaaS) are on track to drive a 33% increase in corporate spending on integration technology, with IoT supplanting 2014's SaaS dominance. What aspects of the IT budget are going to suffer as a result?

Herring: What loses out from a budget perspective is architecture. [IT will] put a fence around that and instead work on the next generation of apps with different designs and requirements. Beacon technology, for speaking to customers as they enter a store, is one. We have thousands of new apps and millions of devices being connected. This is new technology that developers have to learn. SaaS vendors are going to have to integrate with a business' back-end systems and these devices. IoT is far more pervasive than we realize. It's all about integration of these different platforms, apps and devices.

Positive impact on the fiscal bottom line can be achieved in two ways, by increasing revenue or by cutting costs. Is one giving way to the other?

Herring: It used to be about using IT to help reduce costs, but cost reduction is not the biggest driver anymore. Now, it's about increasing the number of customers to drive revenue. Getting more and more customers into the store or onto the website is the key, not getting leaner and meaner. That means closer integration of the customer database with the real-time, in-store customer experience.

It's not the number of application deployments by which you measure success, but rather the revenue they generate and their overall impact to the top-line business.
Mark Herring,senior vice president of products, Software AG

You're positioning IoT as the 'integration of things,' and as the biggest driver for integration technology since the advent of the Internet. A key trend you identify is that smart integration techniques will evolve, forcing changes in the way real-time streaming and bursty traffic are handled.

Herring: We're forgetting that as the number of IoT 'things' climbs, the need for integration soars. Devices or beacons or smart machines are of value only if you can tie their knowledge together. If, for example, I'm an IoT vendor with a smart vending machine that reports it's low on a particular item, it does no good for the product stocker to show up with inventory only to find that a new machine has just been installed by the service department. Different teams that are on different cycles have to integrate their disparate systems and communications.

IoT can be characterized as billions of devices that generate miniscule data packets, such as a light switch that says, 'I'm on and I'm dimmed to 50%.' How does traffic that consists of many small packets differ from systems and applications that generate large data streams, and what are the development implications?

Herring: This is an important point that's often overlooked. They require completely different design paradigms. A device like a light switch that sends sporadic information might send nothing for many hours, whereas if your refrigerator overheats, it's suddenly going to start transmitting lots of packets. Our premise at Software AG is that design points differ and that our integration push, along with packages and services, has to take these into account.

What does the integration trend mean for legacy CIOs, whose image of IT is having everything in an on-site data center?

Herring: Legacy CIOs are afraid of the new SaaS world, and of being forced to work with it. What they're being asked to do is take new SaaS solutions -- Salesforce.com for one -- and get it to work with the company's existing legacy customer database. Those CIOs may not have been included in the discussion to subscribe to Salesforce, but they are being forced to make it work and integrate with legacy data. It's going to be a problem for some.

Next Steps

Going beyond fail fast

How is fail fast working for you?

When "failing fast" works better as "experimentation"

This was last published in April 2015

Dig Deeper on Mobile cloud computing application strategy

PRO+

Content

Find more PRO+ content and other member only offers, here.

Join the conversation

8 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

Do you think that more new businesses should embrace fail fast as an option?
Cancel
The businesses should understand what it means. It's a process and culture change. On the other hand, life critical, national security, financial and legal sectors better be careful with "fail fast".
Cancel
I agree with @agareev that they should at least understand what it means, but whether or not a business should adapt the model depends on the context of each business. I think that, with the pace of innovation moving faster and faster, some business are going to have to adopt the fail fast model in order to survive.
Cancel
Is it just fail fast?  I think many companies need to see failure especially at lower levels of IT, not as something to fear, but as something to be learned from.  There is too much expectation of perfection in some companies, and it just isn't well known.
Cancel
This is the only way to align with future as someone said that we are in 2nd half of chess board where pace if change is beyond our imagination. Unless we adopt a model of fail fast, fail safe, fix fast there is no way we can cope up with demands today. If business wants to move fast, there is no way IT can adopt traditional model......
Cancel
People are in essence already doing this though.  With TDD you see a lot of things fail before they pass.  I'm concerned this is setting people up to believe they've 'well tested' their apps, when subtler and hidden bugs are waiting to jump out and bite a consumer.
Cancel

I don't think that business failure is a comfortable idea.


However, "failure" in an experiment is not a failure. It's valuable data to learn from. It's about finding out what appeals to the customers and how. 


This must be backed up by the true agility of IT, capable to deploy prototypes fast and rolling back instantly if things go undesired way.

Cancel
I think this is different than TDD, although the objective behind each approach is to gain valuable information quickly. But, TDD has more to do with failing tests as you start to develop software, then fixing those failures. Fail fast and adapt is on a larger scale, where you put the software out there to see if it works or not, get some feedback, and adapt as necessary.
Cancel

-ADS BY GOOGLE

SearchAWS

TheServerSide.com

SearchFinancialApplications

SearchBusinessAnalytics

SearchCRM

SearchSalesforce

DevOpsAgenda

Close