The past decade or so has been defined by organizations realizing the benefits of the public cloud and adopting hyperscale architectures. And the benefits of the cloud are plentiful - flexibility, availability, performance, distributed access, regular upgrades, etc. And when cloud adoption became mainstream, I'm sure we can all clearly remember cost benefits being cited as one of the main drivers.
Yes, there were arguments about how enterprises would shift their cost base from CapEx to OpEx, but generally the view was that things would work out cheaper overall. However, in many instances this hasn't proven to be true. One of the problems being that companies often lifted and shifted their systems and architectures to the cloud, and continued operating as they would have done on-premise.
But that doesn't mean that this is inevitable. A number of factors at the moment, all happening in confluence, mean that it is now essential for enterprises to take a look at their IT operations - with the view to automate and optimize. Otherwise, you're simply burning cash in the cloud.
This is something that is on the mind of Avantra CEO, John Appleby, who has spoken at length previously about how companies are sitting with hundreds of SAP consultants and operations people, simply keeping the lights on. And whilst Avantra's interests mostly lie in the sphere of the SAP world, Appleby's views ring true with what we are seeing in the broader market.
And this isn't a straightforward story - even though the solution is relatively achievable. Buyers need to understand what's at risk here, which isn't just about cost optimization. Yes, that's part of it, but this is also about protecting and retaining talented operations employees, who aren't going to hang around to be woken up at 2am to do necessary maintenance. And it's this latter part that should be of particular concern to those not yet automating operations, as that's potentially a greater risk than reducing cost.
But let's take a step back. Appleby provides some context on what's driving this need and says:
I think first of all when people did the business case for the cloud, that was predicated around cost optimization being part of the business case. They said we're going to run these non-essential systems part of the time, and then we're going to do auto-scaling on demand, so we can deal with the peaks and troughs, and that forms part of the initial business case for moving to the public cloud in the first place.
And we all know that business cases don't necessarily get realized at the beginning. But if you just compare the simple cost of running a VM on your local VMware versus running on a public cloud, it is more expensive on the public cloud. It's always more money. But there has to be fringe benefits from doing it.
And what we've seen with this move to the cloud is that people sort of skipped over squeezing the benefits out of that. So they've got this platform with all of that flexibility, but they didn't actually take account of what's happening and they've got this creep of budgets going up.
Appleby adds that this is being compounded by buyers recognizing that the cloud offers ‘infinite storage', but that too comes at a cost, and the end result is that the total cost of ownership is just being pushed up and up.
But in some ways this is slightly inevitable. It's a lot to ask for organizations to get this right from the get go (unless you're the likes of Amazon, which have been optimizing at scale), and this is all part of a journey to better performance in the cloud. Appleby adds:
There's been a shift over the last 10 years to get off those assets and move to a hyperscale world. And that journey happens in phases. I think for many companies that journey is well underway, for some it's complete, for some it's starting. But either way, you've come to kind of the end of phase one.
And now you say, well what am I going to do next? And so you've got this modern platform - but you've got the same old stuff on it with the same old model. And so I think people are now saying: right, what do I do next? I think that logical step is to say, we need to look at the unified support model, we need to look at how we do cost optimization and auto scaling. It's the logical next phase to the cloud journey.
Understanding the problem
As noted above, Avantra's interests in this area largely lie in automating SAP operations. It has also made interesting announcements in partnership with ServiceNow, which allows for the automatic forwarding of Avantra notifications to a ServiceNow instance to create incidents filled with detailed SAP performance and landscape data. Event records can then be directly to ServiceNow, synchronized between the ServiceNow CMDB and Avantra, to ensure monitoring of all SAP objects.
However, this isn't something that Avantra, SAP and ServiceNow are thinking about in isolation. You only have to look at the work that Google Cloud and ServiceNow are doing around optimization to see that this is something that is going to come to the fore for buyers in coming months.
And whilst this is technical detail, ultimately this is about Avantra playing a role in helping organizations simply discover what they have and to identify the problems using automation. Appleby compares the situation at present to when companies didn't know how many staff they had, whilst trying to carry out talent management programmes. He says:
One of our customers uses Avantra and ServiceNow to provide a single view of inventory. That sounds banal, but almost nobody has actually got that. Often these companies don't even know what they've got.
If you look at HCM, those problems were solved 10 or 20 years ago with PeopleSoft, Workday, etc. 20 years ago a lot of companies didn't know how many employees they had. How do you talk about talent management when you don't have a record? You have to have a record. And once you have a record, there are a few things you can do.
First of all, you can have a look at what's anomalous - so what is operating outside of the envelope? And often those things are correlated. So you might have user response, you might have capacity restrictions, you might have backups that are running, or you might have batch jobs that are running. And typically, by the way, these layer on top of each other. And then at that point is when you have a user experience problem or an outage.
The difficulty with SAP specifically, is that in order to provide even a vague level of analysis of a problem - incident analysis, real-time operations monitoring - you need to go really deep into the application. So we increase the ServiceNow taxonomy to provide all the details of SAP and we feed that into the ServiceNow Service Graph, into Asset Management.
So, once you understand that, and you understand the anomalies, you can then start to provide workflows to people to acknowledge the problems that are likely to happen ahead of time. And then the last piece is that you can use that to inform automated maintenance.
Why this is important
But as COVID-19 has accelerated ‘digital transformation' in the enterprise, as is evident from the stories we have heard over the past few months, this too has accelerated challenges that were on their way prior to the pandemic. Whilst Appleby's core point is about optimization for the cloud, centered around cost, this too is going to be critical for companies that value their operations staff and what they do. Want to attract and hang on to the best talent? You better be thinking about this, now.
We fight tooth and nail for our engineering resources that we hire, and everyone is in the same boat. These IT operations people, their salaries are increasingly becoming more expensive, they are becoming more scarce, and they don't want to be doing the same repetitive stuff. So you've got market momentum, market forces, acting in both directions. They want to be working on automation - so when we hire for testing for example, we hire test automation engineers, and not people doing unit testing.
We also have customers that have redeployed operations people into projects, for example. Another piece, which is often more important with enterprises, is quality of life. They don't want their team to have to get up at two o'clock in the morning to do maintenance windows when it can be automated during office hours.
And as we've learnt over the past 18 months or so, a strong company is nothing without its people.
I think this is a really important topic that I've not spoken to many people about. And I partly think that's because it's not the most sexy of topics and it's something that companies struggle to get to grips with. But I think the people element is what ties this whole thing together and will make it a top priority. Organizations fighting over engineering and operations talent will soon have to think about why they aren't attracting the best people - and automation and optimization should be at the front of minds when having that conversation.
But also, money does talk. And with organizations investing heavily at the moment in their digital and technical infrastructures and programmes, money saved is money that can be redeployed to areas that can actually increase revenues. As Appleby says:
The thing I find most interesting is that a lot of customers are still content to imagine that this isn't a problem. And that's insane. They've got teams of 100 SAP operations people and they're content to literally set fire to that money every year. And I don't understand why people are content with that.
I think it starts with the fact that CIOs are fearful of touching the SAP island. They're not sure what they want to do with it in the long term, and so they tend to sort of leave it be. But we're talking about such a large amount of money that's spent.