All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Which is the best business model for using cloud providers?
LosPollosHermanos
Member
I am setting up a cloud hosting business that uses several popular cloud providers. So DigitalOcean, Google, AWS etc. Without getting into too many specifics, for discussion purposes assume general hosting servers. So assume apps such as WordPress etc. I am not doing that specifically as there are already sites that do that. But I am using a similar idea as them.
I am stuck on 2 possible business models.
Business Model 1
One is where I charge for everything including the hosting and they just select which cloud provider to use. In that scenario they do not have access to the cloud provider client area and all the things they provide such as console etc. I will probably build my own console for them to use. They will not have linux root access but they will have SSH/SFTP access to a master account on linux with limited non sudo capability. A similar existing service is cloudways.com
Business Model 2
The other scenario is they own the cloud hosting account with AWS or whoever. So they have access to all the AWS provided things such as console, they have full root access to the server, and they are billed directly by AWS. I provide the software as a service that automates all the installation and configuration of their server, backups, notifications etc. It's all centralized so they don't have to do that separately with each individual cloud provider. They just provide my website with the API access keys to their cloud provider so my website can control the server, install my SSH keys and agents etc. This is similar to runcloud.io services.
Business Model 1 considerations
The cloudways.com scenario would be for people who want me to do everything for them and don't want to get into any of the gory details of linux. It also make it a little harder for customers to leave because of the fact I am doing everything for them. I probably won't even let them have access to their backups other than for use on my service. I don't think it would be that important to this type of customer anyways. One downside is that it's more complex to do it this way, there will be a higher level of support required, and I can have more issues being stuck with hosting charges due to fraud if I am not careful.
Business Model 2 considerations
The runcloud.io scenario simplifies things. I don't have to provide a tty console. I do need to add a mechanism where I can save and use customers API keys but that should not be too difficult. I don't have to worry about fraud as much because the customer is the one paying the hosting charges. I just charge a monthly fee for the service and possibly some tiers of service. The downside is that it will attract more of the do it yourself customer and would be easy for them to just use my service for initial setup and then cancel or go somewhere else. So churn will be higher and revenue is potentially lower.
Again, I am not doing anything that competes with runcloud.io or cloudways.com. I am just using them as examples of the two different types of business models I am considering. I have most of the frontend and backend already built. I am just at a point where I need to decide on the direction.
Comments
It would be helpful if we had some idea of what you were selling.
Is it hosting for a specific app that you own/resell/are an expert on? Let's say you're providing white-glove wordpress hosting (an example you offered). In that case, I'm coming to you for Wordpress and I don't want to get into which provider because I don't care. Could be on AWS, could be on a laptop in your garage. If I have have different tiers where I can choose price and reliability, give me options, but I just want one bill from you.
Model 1 is a support nightmare - you need to interact with tons of providers.
Model 2 is a support nightmare - user screws stuff up, user doesn't follow setup, user installs other things, user runs Ubuntu 20.02 and you require 20.04 and your setup doesn't check and it takes you three weeks to figure out why it isn't working correctly and blah blah blah.
I would just pick a provider that meets your needs and customer needs, and give customer options like location, availability tiers, etc. Then you provide the solution with access you dictate. If they're clever they can figure out that you're hosting on DO or AWS or whatever, but so what.
Webhosting. Not Wordpress and not that particular vertical market but it doesn't matter. If you know how to run a WordPress hosting business then this is no different other than the expertise in whatever the application is you are hosting.
It is open source software, like Wordpress. I don't own it. If you just want one bill from me then that is Business Model 1.
If it was easy everyone would be doing it and it wouldn't be worth it. I already do something similar on my own infrastructure so this is just kind of v2 for me. Support is not that big a deal.
I can certainly launch with just one cloud provider if I wanted to. The software is designed to support multiple providers from the get-go and it's not that much harder than if I was designing it for just one. I have parts of it running with multiple providers already.
The customer also chooses location, storage size, memory etc just like how runcloud and cloudways do it.
What are you trying to achieve? Just to offer max number of possible locations for your service? Or your customers would be using few multiple locations at once?
If you are still not sure what it is you can look at the websites of those 2 providers I mentioned.
One of the sticking points seems to be why support multiple cloud providers. The answer is because a common interface to use multiple providers is one of the value-added features. That includes things like common backups, common notifications and monitoring etc. etc. for all the supported providers. Either one provider at a time or several at the same time. For development by the hour or for production hosting long term. There is no effort to hide any of that and pretend it is my infrastructure.
I could just use one provider (which I may start with initially) but the value-added is not as compelling in that case. By designing the software to support multiple providers from the start it's not that much more effort to code it.
Seems like you're leaning towards exactly what cloudways is doing. Multiple providers, transparent resources and one click apps from the get go. They have big pockets, 50-60 employees, 4m+ revenue, huge customer base and are most likely still in the red.
Unless your offering something real niche, id just start overselling vps nodes, least you might make some money
I don't think people consider it's hard to implement, it just might not be worth an effort (reselling). Depends on what added value you will offer.
As S. Jobs suggested - start from sales and create a technology, not the otherwise - create technology and try to sell it. That's my favorite quote from past weeks
I would go for model 1, i think the potential customer base is larger and in the future you can cross sell more services to this type of customer (if you want). Also with model 2 there can be more discussions about responsibilities.
Thanks for the feedback. I have gone back and forth on it a few times and I'm still not sure what do to.
I have already been doing this for a long time. Just not in this way. LIke many other people I am losing customers to Google and AWS anyways and frankly, these multi-billion dollar companies can do a lot of things better than I could ever hope to be able to do managing my own infrastructure.
You keep using that phrase and I hear this a lot, but it depends where you draw the line.
Yes, you won't have to manage servers, switches, and hardware. But you still have to manage the OS, security, patching, backups, etc.
They are not competition for what I am doing. There are several companies doing what they do which is going after the WordPress crowd among other things. I guess that is where most of the business is but it's not of any interest to me.
It's not a monolithic website so I could scale up development by farming out parts of it without giving up my overall design. Cloudways may be doing something similar for all I know.
There are reasons for not doing it other than "hard", I suppose. I think of it as writing multiple modules for billing system, let's say 200€ each and there would be some additional work, so add few hundred on top of that. Actually there would be even more work after all supporting it, ok. Perhaps I oversimplify as I don't get the point here.
I have a lot of that written already. Billing was not the hardest part but I hate doing it so it seemed hard.
You know, I'm just sitting here, refreshing my screen and waiting for a final version of your post to come out
So many alfas and betas 
If you were asking which business model is more sustainable or profitable - I don't know.
If you were asking is it hard or costly to do - I tried to calculate.
If you were asking if it's worth doing - you answered it yourself, as you already do that.