Spill the IT Ep09: Resource optimisation, don't get caught out

Intro (00:05):

Welcome to the Fasthosts ProActive podcast: Spill the IT. Each episode we'll sit down with some of the amazing ProActive team and chat through their experiences of the ups and downs of IT infrastructure management in small businesses. There's always plenty to chat about.

Graham (00:27):

Well, here we are again at Fasthosts ProActive HQ, here to record another podcast session. With me today, I have Gary again who was on our last session as well, who's the Solutions Consultant here. I'm Graham, I'm your host for today. But before we start, Gary, for anybody that's not tuned in before, not heard your dulcet tones, why don't you tell people a little bit about yourself and what you do here at Fasthosts ProActive?

Gary (00:54):

Hi Graham. Thank you. Yeah, so I'm a Solutions Consultant. That means I'm responsible for designing solutions that meet all the business needs and allow us to add the maximum value into the organisations that work with us.

Graham (01:09):

Fantastic. You're the one that gets down in the weeds.

Gary (01:13):

Most often, yes.

Graham (01:15):

Fantastic. Well, today we're going to be talking about resource optimisation and that's not about people, that's going to be talking about the IT infrastructure, optimisation of your infrastructure. So before we get started on that, I thought it would be really good, Gary, to hear some of the maybe horror stories, some of the things that you've seen where people have been particularly bad at that sort of resource optimisation. They've just not thought about it, just not got their ducks in a row. What have you seen in your vast and long career?

Gary (01:52):

I think there's been a multitude over the years in my time in IT, and I think the one common theme around all of them is not keeping an eye on things. If you're not watching, things will sneak up on you. I've seen that happen time and time again. Organisations are then forced into undertaking activity that takes resources, people resources, financial resources, away from strategic business goals. So when you have to do these things, you want them on your terms and in your timing.

Graham (02:30):

I mean it's interesting, isn't it? You say something there, and we've not done a podcast on this subject yet, but I guess visibility of your whole operation, understanding what's going on. Where people just haven't got the dashboard in front of them, they just don't know what the impact's been. Are you seeing people not paying enough attention with those types of approaches?

Gary (02:52):

Yeah, there's multiple angles into this, right? So a really big buzzword within IT at the moment, because we do like a good buzzword, is observability. I'm hearing that term used a lot. Effectively, it's what insight can you gain into your IT operations with using the data that you've got available? A lot of organisations have the data. They perhaps don't know how to visualise it. They perhaps don't know how to interpret it. Bearing in mind, what we're talking about in a lot of cases is a 24/7, 365 operations. They don't have the time, or if I'm being honest, inclination, to capitalise on that data and really, really use that observability to prevent them from having problems.

Graham (03:40):

How human is that? I mean, we talk about having these tools, but we still need to look at things that are going on in the business, don't we, and make those what I call human decisions? Are people relying too much on the sort of technology to give them the answers all the time? Are they just giving it a cursory glance or are they not actioning on what they're seeing? What do you see about that?

Gary (04:00):

I think it's, people have the data, as I've said, I think they either underestimate the amount of effort that has to go into doing something with that data because the collection part, and to an extent, the presentation part of that is fairly easy because you can automate that. That's the bit the computer can do for you. But as you said, the big missing part of that is the human understanding. AI is probably not far off, but AI can't do that for us just yet in every situation. So if parameter X is 100, what does that mean to the business? What's the impact of that particular metric breaching going to be in real terms?

Graham (04:54):

We don't have Michelle here with us today who heads up the sales and marketing here at Fasthosts ProActive, but I'm sure it's one of the key conversations she has when she's talking to new customers and actually making sure that people have got that really knuckled down really tight.

Gary (05:08):

It's one of our biggest value areas. We have a team who are looking at this data, looking at the dashboards 24/7, 365, and these are highly skilled professional people who for X number of years are doing this exact same task, but for our own internal infrastructure. So we're talking beyond the enterprise grade scope and this is what they do.

Graham (05:42):

You're like the VAR of the IT world, aren't you?

Gary (05:45):

Completely, I mean.

Graham (05:47):

Great analogy I think there. We might have to do a blog on that.

Gary (05:49):

Keep that one. Look, we're eating our own cooking here. So all of these solutions, all of these offerings that we're giving to our proactive clients or making available to our proactive clients, are based on things we've been doing at scale for many, many years. So as well as the service itself, they're getting the wealth of experience. And, as you've said, that human element that sits behind them that you can't code, it only comes with training, documentation and experience.

Graham (06:24):

Yeah, I guess you've just got their back. You've got people's back. You're making sure they stay good at what they're doing and just say, "Look, you could be falling over very shortly. We need to do something here."

Gary (06:34):

We don't want to speak to our clients.

Graham (06:38):

Yeah, of course you don't.

Gary (06:39):

In operations, when you are speaking to a client, I would say in eight out of ten cases, it's probably not good news potentially. So the less we have to speak to people, the absolute better for them because it means everything's working, things are doing what they need to be doing when they need to be doing them, and as you say, people can sleep again, right?

Graham (07:03):

Yeah. Last time we had you on the podcast you were being interviewed by Charlotte. I know the last one that you spoke about was around myth-busting in relation to migration. So when you're looking at migration and migration projects, how important is it for you to consider resource optimisation within that migration process? How much of your time in the planning, and you said that's what you get involved with, how much of your time is spent in understanding that before migration takes place?

Gary (07:38):

It's varied, but I think the big challenge that I see time and time again... People tend to fall into, in my experience, one of two camps. There are the organisations who have no visibility of what their platform's doing, at what times. They just don't have that data available to them. In most cases, these are self-managed platforms and the organisations managing them haven't had the budget, the skills, the inclination, whatever that may be, to build out the platforms that will give them that observability, and so they have nothing. They don't even have a baseline to go from. So that makes migration challenging or just day-to-day operations challenging.

(08:25):

The other camp are the people that understand very well their peaks. And again, we see this a lot in E-commerce. So depending on your market, there is a likelihood that you're going to have peaks and troughs. So if you sell school uniforms, you're starting to think about getting busy in the next kind of six weeks or so. A lot of organisations understand those peaks. What they don't have the ability to do is understand their operating models within the troughs. For a lot of platforms, they're being forced to resource for those peaks all year round when it's unnecessary. And the joy of the Fasthosts ProActive platform is that flexibility of scale. So a principle that I use when I'm designing solutions is making sure that we start at a baseline and then we can scale as we need to, and we can do it quick because we've got people watching 24/7, 365. If we find that a peak is greater than we thought, we can change that within minutes so that it doesn't become impacting. Again, that agility is one of the things that we bring to the table with all of our solutions.

Graham (09:42):

Just as you were talking there, Gary, I'm thinking about everybody talks about scaling up, meeting that peak, but it's just as important to understand and managing, dialing back. Getting back to a more cost-effective way of working and managing when that peak is dropping off and getting customers to that point where they know that they're not overly spending their money on the capacity that they need to buy to meet those demands.

Gary (10:13):

Yeah, and I think a lot of that behaviour is driven by traditional physical and even some virtual platforms that perhaps aren't as agile as the one that we've architected here. So when you buy a physical server, there is a limit to how many drives, how much memory, how many CPUs you can put into that box. You can't easily change that after it's been set up. The only way to put something new in it is to physically move it, physically install a new piece of resource. Similarly, with some cloud platforms, adding resource also has a potential interruption in service. So I think that's what's driven some of that overspend in It, in that people have to resource for their peak times all the time. With our platform, you don't have to do that. You can flex that up and down, as and when you need to, without having to worry about that interruption to service.

Graham (11:17):

So last time you were speaking... In previous podcasts, we've spoken about the hybrid model, haven't we, where people have got some on-prem equipment and responsibility, and obviously then they're using cloud as well. What's the complexities? Does that get muddled? Does that sometimes add a tier of complexity that really, sometimes it doesn't work as well as it should?

Gary (11:40):

Of course, but I think the biggest complexity with that particular model is budgetary. So if you've got a big on-premise estate that will have a cost attached to it. So when you are looking at migration of those systems off premise, there is going to be a portion of dual operability, which again, can become particularly costly. An area where we focused with customers in this situation is looking at individual workloads. So rather than saying, let's take this massive server IT thing that does job X, job Y, what we suggest clients do is look at job Y, let's build a solution resource specifically for that one workflow. That keeps that spend down. And again, I keep saying it, but because of the platform flexibility and agility, we can then add additional workflows in as the project moves forward. And again, minimise that excess spend by resourcing for the job at hand rather than where we think we might be at our busiest period with everything running.

Graham (12:58):

I guess finally, just one other consideration, it's important for you and your team, your customer success managers, to really keep close to your clients, I guess, to understand what's going on. Are there quarterly reviews? Is it every six months? Are you getting together every other month? I mean, it must help having that fluid conversation with your customers, just understanding what's going on in their world.

Gary (13:28):

It's essential, okay. So what we are doing is we are becoming part of the IT function of our customer's business. So in order to make decisions that are based on business needs, as opposed to just abstract technology, we have to be embedded in that customer. We have to understand beyond the data that we have in front of us. Certainly we have an infrastructure analysis service and a solution analysis service that customers can take. What that enables us to do with them is run through overall performance every single month. Let's look at how things have been running across the last four weeks. Have there been any breaches? Have we noticed anything? Have our engineers identified that at three o'clock in the morning, server X gets really, really busy and we're not sure why, so we want to dig more?

(14:31):

We maintain calendars of customers key events. So by understanding what's happening in the business, again, it allows us to put more relevance and context around these numbers on the board. When we can see, when can expect busy periods, we know the calls, we can then make recommendations. We will retain that data as part of monthly reporting packs and say to organisations, "Right, last year during your busy period, we had an issue here. Let's get ahead of that. Let's put some more resource in this particular area because then we don't have to deal with it on the fly. We're ready in advance."

Graham (15:12):

It's fascinating. Every time I come into Fasthosts ProActive, I just get a little bit more of a deeper understanding. I guess for you, dealing with all those different customers who are all doing all different things, it must make it quite interesting, understanding when they've got new product launches, new services, new events and things like that. So it must be quite nice to have that diversity as such.

Gary (15:34):

It's great you get to be a part of many, many, many businesses. If you're like me and, I won't say I've got a short attention span, but I like to be varied.

Graham (15:48):

Yeah, of course.

Gary (15:48):

It gives me the opportunity to work with a huge number of organisations that just wouldn't be possible unless I was in this kind of role.

Graham (15:57):

Well, Gary, you've been great again, so thank you very much for coming on. Yeah, absolutely fascinating subject. Really enjoyed that conversation. So next time when we're going to be recording, next time we're going to be talking about the challenges. I mean, very relevant to this. Really, we're going to be talking about the challenges of the online seller, new, old, or even evolving, and what's going to be going on, what's going to be the future state of online retail, and let's look at that maybe. We're going to talk about what that's going to look like over the next five years. So have a think about that one, because I'm sure you'll have a comment on our next podcast. But until next time, we'll see you all again soon.

Gary (16:33):

Thank you very much.

Graham (16:34):

Thanks again.

Outro (16:35):

Thank you for listening. We hope you enjoyed this episode. You can subscribe on Spotify or Apple Podcast or visit proactive.fasthosts.co.uk for more info. See you next time.