In this, the third in our series of predictions blogs, we turn to the increased focus on platforms – that is, coherent and managed layers of technology, rather than point solutions or best-of-breed deployments. Let’s see what our expert analysts, Iben Rodriguez, Andrew Green, Ron Williams, Paul Stringfellow, and Ben Stanford, have to say about platforms in all their guises. TL;DR – With a still-moving tech landscape, enterprises can make decisions based on cost-effectiveness, building skills, and keeping it simple at the core.
Iben: Two opposing things are happening. Solution providers that came out with a point solution to solve a specific problem are now trying to become platform players. Big vendors are absorbing point solutions and building out their platforms. At the same time, the opposite is happening, where CIOs want to provide microservices or be more dynamic.
Ron: Sure, platforms are becoming more sophisticated and diverse in what they can do, but that doesn’t mean every platform is an expert at everything. When you look at platform X, it may do very well in three or four categories. Platform Y does well in two or three others. Yet the marketing is that each platform now takes care of everything. That is a risk to enterprises because they may not get the best they need for any particular purpose.
Jon: Perhaps organizations have had enough of marketing. I’m seeing enough around cloud repatriation back to on-premise, and a move away from Kubernetes, back to monoliths, to suggest a rebalancing. Rather than trying to build applications as massive, distributed, cloud-native things, how about we just build them on a stack like we used to? Enterprises are taking stuff off cloud and building it in the old way, and sometimes finding that it’s cheaper or more efficient to do that.
Ron: Keep in mind that it’s a two-sided coin. Time to market drove the microservice business. Monolith works if you have a well-defined application that won’t change massively, and I don’t have time to market problems with it. But for things that have to change weekly or monthly, the microservices model will continue to be very strong, even though it’s complex.
Jon: Yes, we’re going around the buoy again. Outsourcing begat right sourcing, and then offshoring became smart shoring. Do we need a new term, like “right-clouding”? I’m sure there’s a better term, but it’s about keeping responsibility. It’s another iteration of the “Oh crap, we couldn’t just give away the farm and expect the farm to run itself” cycle.
Andrew: It depends on the type of organization. Startup developers are more comfortable and have experience working with three cloud environments, but it is harder to learn about and work with twenty on-premises hardware/software vendors. Though interestingly, I see more products purpose-built for cloud use cases than for hybrid environments, despite widespread agreement that hybrid is the future.
Jon: Perhaps that will change with AI demand. Data management platform providers will benefit from AI, whether AI delivers or not. It will drive demand for data management products – including pipelines, catalogs, and everything else. Everyone will want to do more with data and use more data in the process. They will come up against governance, compliance, sovereignty, and quality challenges and will need to address these as they go.
Paul: The storage industry could step back into disciplines that have been largely forgotten, such as cleansing data sets or more easily moving data sets around – for example, putting a data set into a training model much more easily than trying to lift the whole thing, Also revisiting technology like cloning – instead of copying lots of data sets, creating versions that can be removed once they are used.
Ben: There’s a people question with all this. We’re negotiating the rocks in AI, trying to work out what’s going on. If you’re an IT leader, your understanding of the choices you’re making, the complexity you’re buying, and the staffing requirements for a skilled IT team to help you make good decisions – these areas are getting tougher, not easier.
Iben: For enterprises, this means keeping things simple first, and building on top. In Gordon Ramsay’s cooking show, he goes into a restaurant, and the first thing he does is look at the menu. The poorest performing restaurants have huge menus that are super complicated with too many offerings. So, he says, “You’ve got to get your menu down to one page. Make it really simple. Focus on doing these things well.” Then, customers can order something that’s not on the menu. “If you feel comfortable doing it, you can do it.”
For further reading on enterprise strategy for platforms and how these drive vendor relationships, check out our interview with Koby Avital!
The post 2024 predictions redux: data, storage, infrastructure, and services platforms revisited appeared first on Gigaom.
Source : 2024 predictions redux: data, storage, infrastructure, and services platforms revisited
