Something a little esoteric, and a little philosophical, for a Sunday morning read.
While it’s difficult to pinpoint the exact time the balance shifts, it seems Wall Street has finally accepted the cloud as a superior solution to on-premise hosting. The impacts will be massive in terms of technological change and acceleration. That trend was highlighted last week with Google investing $1 billion in CME in return for the exchange moving their trading infrastructure to Google Cloud. Google clearly sees the change coming: they are willing to put a billion dollars behind it to try and capture some portion of that shift.
It’s hard to appreciate how much of a mindset shift this is: for an industry that historically has done everything technology focused in-house, shifting their hosting infrastructure to the cloud is a foundational mindset shift from doing everything in house, to only doing a few strategically important things in house.
This got me thinking… what would happen if every application, software solution, and vendor on the street was cloud hosted tomorrow? What would be the impact on the industry?
When I started Street Contxt in 2012, a lot of firms were still pushing to host everything themselves on prem, including both internal systems and third party vendors. There were many reasons for this, but they essentially boiled down to a view on security (risk), ease of integration, and historical preference. I’m sure there were other reasons, but those were the ones mentioned to us.
Since 2012, Street Contxt has been a hosted (cloud) only solution. We never developed an on-prem solution, a decision which I largely credit to our then CTO, and while it was the right long term decision, it was extremely painful in the beginning. We had several clients pushing us to develop an on-prem solution: they said it would “instantly” allow them to become clients, because the biggest concerns were around our cloud based infrastructure (hosted on AWS). We decided to hold out, and it appears to have been a very prescient decision.
There are a couple important things to note: the difference between architecting an on-premise application and a cloud hosted application are not trivial.
First, let’s look at the high-level view: it’s not as easy as “take the thing we’ve built, and make it work on the cloud instead of on-prem” (or vice versa). It’s the equivalent to rebuilding an entirely new application for the most part. What this means is that for vendors who have historically had an on-prem model, the shift to the cloud is not going to be an incremental amount of work for them: it could result in them having to fundamentally rebuild their applications from the ground up. That will be a massive delay in their evolution, and potentially fatal for their business. It will make that shift hard to stomach economically, and they will be caught between supporting their existing on-prem clients, and investing in the future (hosted) solution.
Second, from my view, hosted solutions have a massive evolutionary advantage over on-prem solutions. If you’re an on-prem solution, you have a much slower release cycle, where you deploy a ‘new version’ of your software. This is because deploying something on prem is much more complicated, as different clients have different versions and different setups of your software. That means some firms are only releasing once a year. Contrast that with cloud hosted solutions who are moving towards “continuous delivery” and can deploy every hour.
Additionally, cloud hosted solutions actually get insight into how users are leveraging the solution by looking at usage data. Those vendors can use that data to fix problems, improve the user experience, and ultimately iterate towards a better solution. For on-prem solutions, they are largely a black box, providing no feedback or insight into the user experience.
The only advantage on-prem solutions had historically was acceptance and security – it was usually faster to get approved. As it was hosted by the firm themselves, they were less concerned with the vendor’s security. It seems that final bulwark might be eroding, as the industry more generally gets comfortable with hosted solutions (given they have the right approach to security), and increasingly looks to either reduce or eliminate their internal hosting capabilities of third party vendors, except for limited edge cases.
So, you’re not a developer. You work in the industry, you don’t build applications: why does this matter?
If cloud solutions become the default for the street, replacing the legacy focus on hosted solutions, it’s going to be a foundational change for the industry. The ground will shift beneath your feet, which could change a few things:
Firms who get comfortable with cloud vendors earlier will have a competitive advantage over their peers who still insist things be on-prem – this will compound as the technological gap widens.
Vendors will start building cloud hosted first solutions. This might not feel like a big change immediately, but it will mean these applications can have a much faster update cadence. The new release cycle can go from years or months to days or hours.
With hosted solutions having insight into usage for the first time, expect vendors to focus more on the end user experience and adoption, instead of only focusing on the technology team facing features and benefits.
Cloud providers are going to start competing in capital markets – none of them (AWS, Google, Microsoft) ‘own’ the industry yet, so they will start partnering with firms and vendors early to lock down portions of the industry. This means there will be hosting deals to be had.
Expect way more integration. With more hosted solutions publishing an open list of APIs, technology teams will be able to focus more on integration and less on deployment
Expect a broader ecosystem of tools and solutions to develop. More niche problems will now be economical to solve. More vendors will be able to launch. This could lead to a Cambrian explosion of new solutions and tools with a faster time to market.
Expect a shift from in-house built tools to more industry wide tools. Between the superior economics of scale, niche focus, and continuity benefits, firms will accelerate the switch from legacy internally built solutions to market based solutions.
Overall, it means the pace of change in the tools and technology the industry is using should massively accelerate. Time to market will decrease. Time to deployment will decrease. Iteration cycles will massively shorten. End users will be prioritized and get a better experience.
Case in point: I can tell you that I haven’t been asked if we have an ‘on-prem’ solution in at least two years. It just doesn’t come up anymore. That era is ending, and the era of the cloud is coming, if not already here. It feels like it will be a major catalyst for change in the industry, up there with the FIX protocol. It will change how firms view and leverage technology in their strategy, how vendors interact with clients, and how fast new problems can be solved. It will be an exciting time for the industry.
As always, we’re here to help,