Seagate’s Grace Liu On Simplifying Your Cloud Deployment

Hi James, happy Friday to you too. Very much look forward to our chat about simplification.

Very, very, very active week. We are in the middle of a global digital transformation implementing SAP, so as you can imagine extremely hectic. So very much look forward to the long weekend.

Yeah, I think it's the same for any large-scale transformation program, be it business transformation, digital transformation or just simplification of your cloud computing. Reason being is what do you have today? The thing you want to simplify is a fabric. So simplification means you are destroying the fabric you are having today and you may tear off a piece, you may puncture a hole. So this is the challenge or complexity you are facing. So what is required is the same thought process, the framework, the approach you are facing any big scale large-scale transformation program. You have to think about the result you want to achieve at the end and then come up with a plan. And during the process, I would say the process is as equally important as the result or sometimes far more important than the result because without a good process you will not be able to achieve the end result.

Yes, absolutely. I think it's pretty much all about components. I think this is something really important especially for the infrastructure folks because you really have to think about there is an objective you want to achieve, a business objective you want to achieve at the end of simplification, not just a technology objective. You have to put yourself in the shoes: you have a customer, you have to know who your customer is, what kind of value you're going to deliver to your customer by simplification. So for example, the infrastructure folks they have to keep in mind your customer is the application guys. The application guys they do not really care about the compute, storage and networking, these components. They really care about can you offer them a really good sound infrastructure that never goes down, high performance, and can run their applications seamlessly. So you have really to keep that in mind when you design your technology objective. What are you trying to simplify? And then take that mindset into your own technical transition plan. What do you want to simplify? I think that's very much a mindset thing. We always, no matter which specific functional domain or role we are in, technology professionals we have always to keep in mind who is our customer, what we do, how do we deliver the business objective to our customer.

It's interesting. Okay, so they really get focused on the components, not the larger solution.

Yes, absolutely true. I probably can expand it in two ways. And the first is, I would say the perfect solution never exists. We IT professionals, our job is always to find the fine balance between a number of factors. It could be technology, it could be cost, it could be time to customer because timing is also of extreme importance. And those three components, those three factors constantly change. So I'll give an example: many companies at the very beginning they really like cloud computing because it gives the technology and it gives you the speed to deploy, but very quickly they realized cost starts to become prohibitive, very very expensive, especially when your data grows to multiple petabytes. Then you realize there is a hidden cost, you have to pay enormous amount of money to do egress cost. Then you have to chase and find the next fine balance between those three things. So the perfect solution never exists. The second point I want to bring up is something new because the world we are living in, we have a data explosion. Data explosion not just about the volume of data, it's really the variety of workloads, different type of data jobs you have to run. And in that case you have many many options. So I give you a very specific example: many many years ago there was no big data platform at that time, the only option you have is to dump your data, run the workflows on the same relational database platform. Moving forward now, we want to move those workloads to a big data platform, to the cloud or other on-prem solutions. But then it's very interesting: you always find a 10, or even though a minority may not be 10 but a minority, they actually really love what they used to have because what they used to have, the relational database, gives them the performance, even though it doesn't handle that kind of big volume. So again, when you do simplification, you have to look at all your constituents, your customers. Some may really love the new solution, give them. Some may really find those are the unique corner cases, they really like what they used to have. So you really have to think about all your customers. What is the right solution, the word is not very perfect, what is the right solution for every use case.

So I think it goes back to the fabric, right. So now you suddenly realize my approach may very well be I give you a new fabric and you will have the customer say I really don't like the new fabric, I like what I have today. So again goes back to before you start your journey, you understand all the use cases you have deployed in your current solution, your fabric. And then design the approach: who want to move out to the new solution versus who want to remain, who still find good value out of an existing solution. Then comes into the cost benefit analysis, the traditional program management, the negotiation. All these old school not so sexy things, but still these are extremely important in this simplification journey. So that's why it's pretty much the same, those are the same elements you will see from business transformation, digital transformation programs.

I feel we tend to see this in the application world. For example, in the application world how things start: you have a business requirement, you have an issue, you have a problem you need to stop, so you go out and do your evaluation and you find a software you really like. So for example, you look for a solution to manage your sales motions, you go out and find a tool, you really like it, it's initial success. And then it happens to be that you may have a similar, you have an adjacent requirement in overall sales marketing work, you have a marketing requirement. Then what do you do? What could usually happen in this scenario is: go back to those software providers, they always want to look for opportunity to expand their business, grow their revenue. So the space they tend to grow into is the adjacent spaces too. So you typically see a company who started off offering sales solutions, they want to extend into marketing space. So when those two things come together, when you have initial success with the company in this specific domain and you look for something else and that company happens to want to upsell cross-sell their new product into that new space, so when those two things come together you would say let's jump in, let's maybe forego some of the due diligence. But I think for IT professionals, the moment you really need to pause and be careful is when you have initial success. Those are the moment you need to still pause and think about: is it the right approach?

So the digital transformation of Seagate is the umbrella name for a number of initiatives. We started off by simplification of our infrastructure. So we moved, we consolidated data centers, we moved some workloads to the cloud, we keep some workloads on-prem. Now we are in the process, because we are also finding the cloud could be cost prohibitive. Now we are repatriating some of the workloads back on-prem. We're looking for the cloud to run some of the workloads that does not charge egress cost. So that's coming to the like, I don't want to spend too much time on the like cloud approach. So that phase has been completed. What we are doing now is we are implementing the next generation ERP system to run Seagate business, not only for the current business but a futuristic platform to support and enable the business growth for the next 10, 20 years.

Thank you very much James. I really enjoyed the conversation.