It seems that every year I find myself on another sales call with a current vendor trying to convince me that their price increase leads to more efficiency and savings which makes me think back to the beginnings of the cloud and SaaS transformation.

Much like “AI” is the current buzz word, “cloud” was the buzzword back then. CIO’s were being inundated with articles and sales calls telling them about the benefits such as:
– Lower prices due to economies of scale
– Improved security / disaster recovery
– Increased flexibility due to no capital investment
– No need to build out internal data centers

At the time I was working for Applied Minds which forced me to really evaluate these statements early on. Not only were we mandated to always be on the cutting edge but one of the founders (Danny Hillis) was friends with Jeff Bezos and AWS was just beginning to make people take notice. While Applied Minds did want the cutting edge they were also budget conscious so every solution needed to be weighed out.

This made for an interesting combination and helped create some procedures I still use today when evaluating solutions. To begin, I find a way to evaluate the true cost of a solution.

For example, one of my first AWS evaluations was for storage. We had a need for large volumes of storage and at the time storage was fairly expensive. I evaluated the major enterprise solutions and compared them to AWS S3 but at the time AWS didn’t have the storage gateway appliance they have now. So, I then looked at creating a custom solution that would do what storage gateway does now and presented my comparison. At the time S3 could beat the cost of traditional vendors but the solution to access its storage would need to be tested and built in house. It was around this time that someone at Applied Minds mentioned ZFS, a new file system built by Sun (now Oracle). After evaluating that solution it blew everything out of the water by using Solaris 10 and a Sun x4500 multi-disk array with Samba. It met both cutting edge and cost. To be honest it was a cool solution!

From there, I would evaluate everything as rigorously as possible including internal infrastructure. To evaluate internal infrastructure, I like to come up with a cost to completely build a company data center and include estimates for electricity, A/C, etc. Back then, my data center builds were costing less than AWS but now I find things break even or can cost less.

So, is cloud always the answer? No, it totally depends on what you want to do but first and foremost cost should be a driving factor which brings me back to these sales calls.

Now that cloud and SaaS solutions have been around for a long time it is increasingly important to evaluate whether we are using a service for a particular reason vs just the buzz word “cloud.” Over the years several vendors have had some fairly large security incidents and because of their “economies of scale” these incidents affected many organizations. Some incidents even resulted in weeks of downtime for customers because of a lack of proper DR procedures.

Most importantly, cost is a huge factor. In each of these sales calls it seems we are rapidly approaching (or have passed) the point where cost savings are a reason to stay with a cloud vendor.

Atlassian Confluence Cloud used to cost $3,600 annually for 100 users in 2016, now its $6,400. I also used to be able to buy Confluence Cloud and self host for $4,000 with an optional renewal that was a percentage of that cost.

With Gitlab you could have paid $4 per user per month in 2017, now the minimum is $29 per user per month! That is $960 per year in 2017 vs $6,960 now for 20 git users…..

Then there are larger infrastructure services such as Microsoft 365 which is a cornerstone of most companies IT infrastructure. They have released some amazing infrastructure over the years such as Entra-ID, Intune, etc but they have also been pushing customers towards more expensive licenses. This last year they increased their prices forcing customers to pay more for the same thing or for less.

It is getting to the point where this “promise of cloud” can become a nightmare. This is especially true with the issues in the USA at the moment, with other countries becoming concerned about their digital sovereignty.

So, what should we do?

First. ALWAYS re-evaluate your solutions and don’t be afraid of building something yourself! The beauty of SaaS was that we could offload some IT expense for maintenance and management BUT if these combined systems are becoming as expensive as paying for an IT person it may make more sense to have an IT person who can build / maintain your own solution.

Second. Build flexible solutions to avoid lockin! Yes, it is incredibly easy to use Amazon ECS to spin up Docker instances and yes Kubernetes looks complicated BUT it is also very easy to build out a Docker swarm environment on EC2 and it will cost less. From here, if you need to move to Azure, Google or any other environment you can just move your Docker instances. Even better, manage this all with Ansible / Salt and you just point your CM system to any provider you want if you need to move!

Lastly. Always, evaluate open source solutions and help support the community. Many of these solutions everyone is using today began as open source projects. There are plenty of new and cutting edge solutions out there and your support can create the next competitor. Tech has always evolved. Evolve with it and look to something new every once in a while. In the least, look at something new each time you get one of those sales calls…..