We are all told that the cloud will be a cure for our problems. Cheaper, faster and better. Are you considering Cloud as well? But how actually should I implement it in my company – You may ask. I know – it is hard. For that reason, I wrote this article. I will tell you what 7 things that you MUST know when considering Cloud. Want to choose wisely? Read carefully!
Everybody is talking about the cloud. Some people are going even further and skipped cloud – now it is AI, the blockchain, and virtual reality.
So many new technologies, features and you don’t even notice them around you. You don’t know how the others can catch up with everything and how they apply it in business.
Do you think you are alone?
No! You are not. Most of the organizations are still exploring and evaluating new technology. They struggle with the same things as you are.
The holiday season in Europe is in full swing, and I’m just about to embark on mine three weeks of travel.
Before I will do this, I decided to take my CTO hat on and share with you 7 hard things I’ve learned about the current state of cloud and organizations over last year – working with tens of our customers around the world but also speaking with peers and experts from other organizations.
Have you already adopted the cloud in your company? If yes – don’t leave this article. Read it or even scan and share your opinion. Were you following the same things when considering cloud? That knowledge may be very important for the others.
To make it useful for you I will also share some practical steps you should take within your organization to get ready. Changes are coming, and you will be prepared.
#1: You are not there yet? Don’t start big, start small!
You hear that everyone is doing cloud, but you are not there yet. So you need to start but there are tons of projects you need to take care of, and you don’t know if it is even worth considering.
Now someone complains about the performance of your e-commerce site. Or the cost of running the document management system. Perfect! You’ve heard that cloud will solve it: cheaper, faster, better. Containers are a solution for everything!
Let’s hop on it.
You have just defined a perfect recipe for disaster and putting cloud project on-hold in your organization for next year.
If you are not there, don’t hop on this train with your most significant cargo on the first trip.
You are missing at least several vital ingredients for success: You don’t know:
- The services, vendors and their platform capabilities.
- How to build and operate it and how it is different from existing solutions and the way you delivered.
- If you have appropriate skills within the company with the right experience.
- How to evaluate the exact cost of building and running it. With consumption-based services, expenses can quickly get out of control.
Always start smaller
Should you stop and do nothing? No – you need to start smaller. Don’t try to learn on the biggest project or critical business system for your company.
Start smaller! How? There are plenty of opportunities:
- Your marketing department needs to run a short campaign, and host landing pages + have a way to communicate with users over different channels including mobile messaging. There is Azure App Services, Logic Apps, security service like Azure AD B2C and services like Twilio – let’s build PoC of it.
- One of your servers is ending its life. You need to replace it. Wait! Maybe we can move it to the Azure IaaS and in the process:
- Establish our first subscription.
- Learn how to build it.
- Determine how it is priced and how to operate it.
- You wonder how to provide backup and disaster recovery capabilities for your organization. Azure Backup and Site Recovery is something which might help you. Evaluate it.
By the way: you are wondering what to do with old Windows 2008 systems still running there and going out of support? You can extend your support window by moving it to the IaaS and then think how to upgrade it.
In the excellent book “Ahead in the cloud” I read about the superb rule of 2 weeks. If you need to migrate something on-prem to different hardware or data center, look if you can move it to the cloud or re-write/ re-configure it to some service within two weeks.
If the answer is yes: Do it!
#1 Key takeaways here:
- Start small, don’t aim your central business system for first cloud project.
- Pick something you can deliver for production in 2 weeks. Then deliver.
- Look for opportunities in new projects/maintenance cycle – this is where you can find low hanging fruits.
#2: There are no talents in this area. You need to grow them
It sounds good. First project. But we don’t know the cloud. Let’s hire someone who knows it and will teach us.
Some truth here: there is not that much talent with real skills in the cloud out there.
Lots of people try it, lots of people play with it, but still, people with real skills are not that common. Another aspect: cloud attracts mostly developers and people with mixed skills – they might not be the first one to jump on your current environment.
But you know what – you have already those people. They are working for you right now. They know your organization and know your operations. They are your employees, analysts.
You want to have the best cloud engineers for your company: grow them. You will also have them for years as they will have an exciting job.
How? Start your Center of Excellence for a cloud. Don’t make it big.
Here is the recipe for a quick start:
- Get three people from different departments or skillset together like the developer, infrastructure administrator, and networking guy.
- Put them in the same room and assign them a task. Real task – like push the first server to the cloud IaaS or build a PoC landing page for your marketing based on PaaS.
- Give them fixed (but realistic) time frame and set the goal of putting it to production.
Hey! You’ve just put together your first crew of your cloud dream team. Give them resources to learn:
- Start the first subscription and take advantage of free tires of cloud offering. Every vendor has them. Did you know about Azure Free?
- Buy Pluralsight and A Cloud Guru subscription for them. Let them learn.
- Let them go to some conference and meetups – there are plenty of people to learn from, and probably there are some groups and meetups in your area.
You are on the road to building your cloud dream team and long satisfied employees at the same time.
#2: Key takeaways here:
- Your best specialists are already hired. They were just not given the opportunity to learn.
- Build cross-functional cloud “dream team” and give them a real project to deliver.
- Set time frame, give them resources to learn – there is plenty of online resources.
- Make your PoC with production deployment.
- Spread lessons learn to others within the company. They will join you!
#3: Cheaper comes in different flavors – Not always in direct monetary value
What you will hear from vendors and consultants is: cloud is less expensive! Let’s hop on it, save money and rave in joy.
You are after your first deployment, you moved your primary server to the cloud and here is your first bill. Where are my savings?
It is not true that the cloud will always be cheaper in monetary value. The server in a cloud might cost more than your existing hardware. PaaS services in your solution are usage based. If you use them a lot, there might be a big bill at the end.
Cheaper in a cloud is not coming from direct savings. It might come in many flavors:
- Less operational costs to run the service because of automation.
- Quicker time to deploy the service even on IaaS because you might not have to way for those servers to arrive.
- More service deployments per month and a speedier time to market for your services and apps.
- Less time spent on implementation with your valuable people focused on the new things instead of solving deployment issues.
Based on the already mentioned book Ahead in the typical cloud organization might see benefits of cost reduced by 30% but it is expressed in total value with operation and deployments.
You need to define your definition of cheaper, and you need to measure it. The last thing is important – if you will not have a way to measure it, how you will know you are getting there?
Do it another way
Don’t try to figure out some complex system for it. Make it empirical. Find some easy way to capture the current state and track change and its direction. Things to measure might vary:
- Time your team spends on deployments.
- The number of deployments for app per month.
- Time of outage of services because of maintenance and deployment tasks.
- The time it takes to onboard new service.
One more thing. This one is important! It is hard to expect things to be better if you will do them in the same way. If you want to get benefit from the cloud deployments, you need to learn to do it in a cloud way.
Have you heard word DevOps? Yes? Me too. Too many times in different contexts. Whatever it means for who is saying it, it is an operating model where you automate your operations and make it part of a process. Not separate tasks from development.
The first task for your dream team (see #2): Even if it is single server they have to make its deployment automated that it is being deployed, configured and being operational without manual interaction in a process. Make it a challenge!
#3 Key takeaways:
- In a cloud doesn’t mean you will pay less on a direct bill. Think and define your benefits and definition of cheaper or savings.
- You will not know if you are getting there if you don’t measure it. Set your KPIs, find a way to measure it for your definition of savings.
- From day 1 and first project on-board on a new method of building, deploying and operating solutions. You will not get there if you do it in the old way.
- Even if you are starting small and doing a proof-of-concept, think about how to build your cloud environment and how to measure and assign costs on projects/departments. you will need it later.
#4: Not everything you hear from vendors is as shiny as it is. Evaluate, build and iterate
Are you following all the vendor’s announcements? Do you know all the services and improvements in them?
Every cloud vendor is building more and more services as demand grows. There is more and more options, services, possibilities.
You look at the videos from latest Build or AWS:ReInvent, and you think – wow. It is cool. Let’s do this.
Hold your horses. Remember. Cloud is a living thing. Vendors build and deploy it on the fly.
If there is a new service and it has a promise of solving your issue don’t hop on it and implement. It might not be what you need; it might be harder to achieve than it looks like, it might cost more at your scale than you think it will cost.
Remember – even if service is serverless and it is part of Platform as a Service (PaaS) there are still servers and interfaces behind it. It might have its limits. It might need some configuration and development.
Don’t accept all that is given as a solution. Remember – at the end you will develop, maintain and operate it, even if it is serverless and PaaS.
Think about support, costs, operations, and deployments.
Should you stop considering cloud?
Nope, but you need to build your process for evaluation and tinkering with it. Have you read it until here – you have a way of evaluating it:
- Start small with some real case project (#1)
- Assign your cloud excellence team on it to assess on the service (#2)
- Find your way of operating it and measure for your use case (#3)
Make yourself informed and make it empirical. Evaluate, build, iterate. It should be your mantra when it comes to cloud services and technologies.
Your people will love this approach.
#4 Key takeaways:
- Don’t take promises from vendors and service descriptions as granted. You need to evaluate it in the context of your organization.
- Use your Cloud Center of Excellence team to assess and onboard new services you think might be useful.
- Learn limitation of new services and ways to build, deploy and operate it. There might be quirks around it.
- Evaluate, build, iterate.
#5: There are no products anymore: solutions
Do you remember a time when it was products off the shelve we were buying to address problems? Then it was a never-ending cycle of upgrades and learning of its limitations.
Now the hard message to be delivered: When it comes to cloud platforms and offerings there are no products anymore – there are solutions.
Of course, you can buy a SaaS product. You still can buy some product to address some simple needs.
When it comes to addressing some problem with cloud platform you will quickly learn that:
- It consists of multiple services and interfaces which applied together builds a solution.
- The solution requires some development and integration typically. It might be possible to leverage the platform for it but still – expect it to happen.
- To maintain it, you need to learn how to build and deploy it in an efficient way.
That is why your teams should always be cross-functional and why you should expect some development always happen in your cloud projects.
It is not about deploying ready products. It is about building from the active components.
#5 Key takeaways:
- Your team should consist of mixed skills with some developers and operations people at hand, even if you operated later solution delivered by some vendor on top of cloud services.
- Expect development in every cloud project.
- Prepare yourself to handle things like source control, continuous integrations, and deployments within the organization. If you have not done it yet – time to learn it.
#6: Business is eating IT, and it is good for it
Have you heard that Software Is Eating the World? It is true. In our space what is happening is Business is Eating IT.
Sounds like a buzzword? Yeah, I thought so. Here is an explanation.
Business is faster right now than ever before. Some years ago when your company would want to deploy a new service, they will come to IT and will ask you about doing it. It will trigger designs, procurement, hardware and so on.
You know what will happen right now? If there is a need, your business will find a partner in some consultancy or development company, will use its magic credit card to do a procurement part (yeah, I know it is not that easy) and will make their own it.
Don’t you think it is happening? It is. We have at least a few such examples right now within our projects at Predica, where business decided to go directly to vendors on cloud technology because IT was too slow on solving the problem.
There are lots of vendors already doing it for you.
My point here is that if you are in it, you are in the best position to be a business partner for your business. You don’t have to be scared of it.
Seems like people didn’t lose their job because of a cloud. It is the opposite – their number grows if a company embarks on this journey.
What you need is to learn how to experiment, operate quickly and deliver quicker on the business needs. Have you read my previous points?
#6 Key takeaways:
- Talk to your business about their needs. You are the best-suited partner for them.
- Don’t try to stop them with inertia like “we don’t do this-this way.” They will find a way to go around you.
- Use your cloud center of excellence team and new operational skills to deliver faster on their needs.
#7: Set your compass: you need to know where are you heading to make right decisions
The time will come and you will need to decide where to move. Should you:
- Build something?
- Do it on-premises or in the cloud?
- Make it be an infrastructure or PaaS deployment?
I’ve been recently in a couple of situations with the customer where after learning all the options and possible solutions they were stuck without being able to make decisions.
Why? It was not about lack of information or uncertainty about costs. All the data were at hand.
It was inertia caused by lack of compass – lack of asking yourself the right questions.
- do you want to achieve at the end?
- is your approach to services?
- is your core business and what is not?
- should you do as the organization, what to grow internally, what to outsource?
Those are not easy answers, and it requires some thinking around it.
Hard fact – no business consultant will tell you what to do. It is you who know your business and you who has access to all the people with knowledge about your company goals and directions.
There is one key takeaway here – you need to know your direction to make decisions and not make it only because others do it this way.
How to know it? It is not easy, and it is a process. Besides that, if you want some take look at some good content during the holidays, I encourage you to watch this video as starting point: Keynote: Crossing the River by Feeling the Stones – Simon Wardley, Researcher, Leading Edge Forum.
Seven things when considering cloud – here we are
I hope it makes a good read for you during your holidays or slower summer time.
If you have any questions, I would gladly know them and try to answer them. Post them under this post in the comments section.
Do you have your learnings or critical takeaways? Remember – there is plenty of people still looking how to do this. Share it here, and they will benefit from it.