ForwardThinking |
|
Rumors of SaaS’s death may be exaggerated |
As we return from a holiday break and jump headlong into a new year, I’ve been thinking a lot about the speed of change inside large organizations, and how it doesn’t usually match technology hype cycles. |
I’ve been around long enough to remember at conferences in the dawn of the new millennium when enterprise software companies were pushing instant messaging as the next big thing in the enterprise. While you could argue that we eventually got there with tools like Slack and Teams, there were a lot of steps along the way. |
It really required the development of mobile and cloud computing for these tools to make sense, and took 15 years for them to begin reaching the enterprise mainstream. |
The same goes for SaaS. Before the cloud emerged as a software delivery platform, we got our software in boxes or via servers sitting on prem in our companies. Yet ultimately SaaS changed how we buy software using the subscription approach, and it changed how we deliver software. Instead of large expensive upgrades we have incremental changes delivered on a regular basis. But it certainly didn’t happen overnight. It began with a few intrepid beginners like Salesforce, and success begot success. Today, almost every software company uses the SaaS delivery model. |
Yet tech leaders, it seems, want us to believe that we can go from zero to 100% when it comes to new technologies without the inevitable bumps, bruises and lots of change management scars along the way. As a prime example, we recently heard from none other than Microsoft CEO Satya Nadella, speaking on the BG2 podcast, claiming that AI agents would replace all software as we know it. While this technology is surely promising, it is really only a notion we started hearing about last year. |
That, friends, is a key example of over-hyping. AI is definitely changing software, although it seems unlikely that we are likely to go back to boxes and on prem. The idea Nadella is advancing is that we can do without these large enterprise software packages because agents can operate with a light touch from humans to undertake a bunch of the stuff we used to have to do manually. |
The thing is, he’s not the first to say it. boldstart founder and general partner Ed Sim said it last fall. Speaking on my buddy Brent Leary’s ‘A Few Good Minutes’ podcast, Sim suggested that AI agents could essentially kill the value proposition for companies like Salesforce that provide a set of front-end services on top of a database (which frankly just about everyone in the software business does). |
“If an agent can treat anything at Salesforce just like a database and there's no user interface anymore that someone has to use and you can end-around the whole thing, where does that leave Salesforce over time? Does that give someone an opportunity to be the insertion point between Salesforce and the end user? And the answer to me is absolutely yes,” Sim told Leary. |
The thing is, while all this is possible at some point, I don’t think it means all software as we've understood it is suddenly going to go away over night, or even any time soon. There are still many questions, as I raised in this column before the break. Agents have the potential to create as many problems as they solve around management, orchestration, security, governance, identity and a host of other things enterprises have to worry about. |
Nadella himself brought up thorny issue of permission and access as agents move around an organization and across systems. He admitted it is still something that needs to be worked out -- and make no mistake, that's not a small problem. |
For now let’s tap the brakes on the idea that 'SaaS is dead,' and keep in mind that enterprises move slowly when it comes to substantive change, and folks weary from hearing repeatedly about the next big thing, don’t necessarily believe the hype machine without some substantial use cases as proof. |
New technologies typically gets layered on top of what came before. Rarely does anything get ripped and replaced. There’s too much risk to the business to do that. If you don’t believe me about the rate of change, look at the fact that 70% of the Fortune 500 still use mainframes. |
I’m not saying that the nature of software as we've thought about it won’t be changing. It absolutely is, but will software as we know it suddenly change over night? I highly doubt it, and as with all new approaches, there will be bruises aplenty as we make this transition. -Ron |
|