David Smoley, CIO of global pharmaceuticals giant AstraZeneca, is a veteran of enterprise cloud adoption, having been one of Workday's earliest customers at his previous company Flextronics. Since joining AstraZeneca he's presided over the rollout of a half-dozen cloud applications. In conversation at DocuSign's London conference this week we asked what advice he would pass on to other enterprise CIOs. We distilled these six key points from what he told us.
1. Define the business problem
The starting point is never the technology. Focus on defining the business problem first, Smoley advises.
We always try to start with where there's pain, either in high cost, or slow speed, or complexity. A lot of it's driven by simplification.
Once you have that problem to find, look for solutions options. I do think I'm a big believer, as a technologist, in technology as that sort of battering ram to get through and lead in change, but then making sure that there's a solid business case.
There's some sort of payback hypothesis. There's a strong business partnership. You've got the team that's working on it, the sponsorship is represented by the business side that's going to benefit as well as the technology side.
2. Use the proven tools
Thanks to others like Smoley who have already taken the plunge, there are plenty of proof points out in the world to show what can be done. He says there's now a proven toolkit of cloud solutions.
What's cool about this time, for CIOs in my view, is there's essentially a toolkit. When I look at Workday, ServiceNow, Box, Office365, and you could argue Google though less so in Europe, and then DocuSign for sure, and I'm probably missing a few, but there's a handful, five or six, key tools that are proven. Concur, I would throw Concur in there as well.
If you're not using them, there's a problem there. By bringing them in, you know it'll be a huge success ...
When you have that tool, Workday, DocuSign, whatever, it makes breaking through those old processes so much easier because you've got a template and you can put it in and drive it as opposed to meeting and process mapping and re-engineering things which is kind of hard.
My view of cloud is that it's great because it saves money. You can do more with less. It's faster, and in general — mostly because it's contemporary, based on agile, iterative development — it's so much more intuitive.
3. Set ambitious goals
Many enterprises, and IT people in particular, are often over-cautious in what they set out to achieve, he says. His approach is to set "stretch goals" such as the classic lean goal of achieving twice as much for half the cost.
You set goals and targets that are beyond reach, recognizing that in many cases, if you achieve 80% of your goal, you may have actually gotten 20% more than if you had set a safe goal and settled for something that everybody felt good that they could accomplish and achieve.
In my industry and my company, that's culturally where we've been. We're risk averse. You take months to sort out exactly what you think you can achieve safely. Everyone agrees, consensus, then you go forward.
From the beginning, thanks to this toolkit of cloud tools, where in other industries, I knew you could deploy Workday in less than a year. I was able to come in with that data point and say, 'We're going to do Workday. We're going to do it in a period of two years, and we're going to do it globally.'
People just say, 'You can't do that. You're crazy.' Because I was in that new position with a mandate from the CEO, I was able to say, 'I am crazy. We're going to do some crazy stuff here, but here are places that it's been done before. It's not impossible.'
At AstraZeneca, IT is getting close to achieving that goal of doing twice as much at half the cost, with operating expense falling from $1.3 billion to around $800 million after two years, while improving metrics such as reliability, fix times and project delivery.
It's mostly around creating a mentality. In our case, it was imperative to create, to put in people's heads, that it's about being better and less expensive. It's an and, not an or.
By actually delivering thanks to cloud, and thanks to insourcing, where it's been another big strategic lever for us, we've delivered it in a way that has won people around now to seeing that it can be done.
4. Don't hold back
When Smoley spoke to diginomica last year about adoption of DocuSign at AstraZeneca, the company was heading for 40,000 envelopes electronically signed during the year. This year that figure has already been passed and the final number for 2016 is likely to be close to 100,000 — with a significant impact on process times in all those instances. This is typical of the viral adoption many cloud applications are able to achieve.
Speaking to us this week, Smoley cited the example of Concur, which has replaced an in-house developed expense management tool.
That's one where [we had] internal resistance ... There was always an argument for not going there. [After] a year or two, finally we said, 'No, we're going to go for it.'
Now that we've implemented it, it's off the charts. People love it. It's so much easier and so much better. That's the value. You get that simple, easy to use, intuitive.
5. Change the relationship
Demonstrating what can be achieved not only builds trust but also shows the contribution IT can make to the business, says Smoley.
Initially that relationship changed as we were able to deliver on our promise. Point one, building trust and credibility around doing what we said we were going to do consistently, which had been missing or inconsistent. Then point two, by reducing cost in meaningful ways, we were allowing resources to be shifted into discovery of drugs and other areas where it was more impactful to the company. Those two things have enabled us to build a trusting relationship.
We're now much more engaged and involved in the strategic digitization of the business as opposed to being chasers and order-takers and fixers.
Business teams are also picking up on the agile approach that IT has been using in some of its projects and applying that in some of their own activities, he adds.
Obviously agile can be applied into all sorts of different areas including drug discovery and development in particular.
The business has actually taken a real keen interest in what we're doing in agile. We've found that as another example of a point where some of the things we've done in IT are proving inspirational and helpful to other parts of the business. That creates that partnership around how to get things done.
6. Partner with innovators
It's important to work with vendors that are pushing the envelope, says Smoley. That was what led him to partner with Workday at Flextronics and has informed other partnerships at AstraZeneca.
I'm a huge believer in cloud but it's not in cloud per se, and over time as cloud's become less unique and a bit more the marketing title for everything, there's good instantiations of cloud and bad ones as well which makes it a bit more confusing.
To me, it's really around finding the right best solution for the problem you're trying to solve. In the Workday case [at Flextronics], we were trying to consolidate and put HR processes in a large, global business. The logical choice, or the typical choice, was SAP, but, when you looked at it, it was a nightmare to install, it was incredibly expensive, most companies hadn't really ever really got it fully installed.
We just saw a great partnership where we had the right spec, if you will, to help them build their product and the right landscape or canvas on which to do it.
It was a perfect marriage, but it was picked because of the leadership and the thought process and culture of the company, and the track record that they had and the commitment that they made to us.
The same applies to cloud infrastructure as well as applications, he says.
I think that the infrastructure enablement is key now. There's two pieces to that. One is the things you need to do in order to support and enable the application piece. Obviously when you're running a DocuSign or a Workday, these sorts of things, your network has to be configured differently. You've got to be able to leverage wherever you are around the globe versus traditionally funneling things through a limited number of access points for security and control. As we move away from the build the high walls and the moats and protect by locking things down, you just can't operate that way. We've become more cloud enabled in the tools we use to secure and protect the environment.
On the other side, obviously we're looking at storage and compute from a cloud perspective. I wouldn't say it's huge in terms of the metrics around either how much is spent or how much compute or storage, but the projects tend to be particularly exciting and interesting. They tend to be more in the science base. We do a lot of work with Amazon, see them as an absolute innovator in the space, a key player. That's primarily in the science with large data sets and highly variable compute workloads.
Invaluable insights from an enterprise CIO who was an early adopter of the cloud for core applications.