Software as a Service has been a huge boon to businesses. Suddenly, for a manageable monthly or annual fee, businesses have had access to high quality software that is maintained for them, updated with new features regularly and accessible without the hassle of installing special software on computers or other devices. But there is a downside. Customisation is limited and you're limited to functions the developer thinks are useful. But application platforms promise to change that.
Those of us with a few grey hairs will remember the days when on-prem software ruled the world. The beauty of those platforms was that we could install them locally and have complete control. For large enterprises, that meant modifying the applications so they fitted our business processes exactly (or, at least, far more closely than the original developers could manage). But there was a cost.
As well as the effort to develop the changes, upgrading when the core software was improved by the original developer became almost impossible. For example, in the late 1990s, the company I worked at wanted to upgrade from JD Edwards 5 to version 7. That was a project that took almost two years as we unwound all the mods made to the Version 5 software and applied only the ones deemed essential to the version 7 update.
With the advent of Software as a Service (SaaS), we can now access software through a web browser. But we've lost the ability to mould the software to meet our specific needs. However, that's changing
At this year's Twilio Signal conference, held in San Francisco, the company's CEO Jeff Lawson talked about the engaging world of application platforms.
"How can we deliver the best of both words?" Lawson asked. "We all know that once you set things into pixels that you're stuck with those decisions".
While considering this problem, Lawson realised that Twilio had solved this problem inadvertently when they created a tool for creating flexible user interfaces that sat over the APIs the company is best known for. It meant that the company had developed a programmable back-end system and a programmable front-end.
This gave them an application platform that, he said is as "easy to deploy as SaaS but as flexible as APIs".
Interestingly, although Twilio says they are the first company to bring this model to market, we have already seen similar offerings. WordPress is not dissimilar in that the basic service delivers a useful application but the use of plug-ins and extensions allow you to substantially customise the user experience. But Twilio goes further by allowing the backend to be customised as well.
Twilio is putting its money where its mouth is. The company released a new application earlier this year, Flex, which is designed for contact centres. They targeted this niche as traditional call centre software is highly customised and, in many cases, several years old as upgrades have become untenable because of the software modifications made.
In order to demonstrate the ability to modify the application, Lawson live-coded a new function into the standard app that added new SMS functionality to the contact centre program. And I'm comfortable it was live coded as a tech rushed out from the back to let lawson now he missed a semi-colon that would have made his code fail.
In addition to being able to code your own changes that will persist as the core software is updated, Flex supports plug-ins for authentication services, sentiment analysis and other functions.
The mantra is that you can deploy like SaaS, integrate like on-prem and iterate like the web.
One of the criticisms I think has been justifiably levelled at the SaaS industry is the lack of customisation options. But the emergence of application platforms that allow businesses to customise SaaS without requiring a massive development budget gives them a halfway point between vanilla SaaS applications and the need to create bespoke software that is either deployed on-prem or through Platforms a Service (PaaS) providers.
Anthony Caruana attended Twilio Signal 2018 as a guest of Twilio