Move to the cloud, they said. It will be cheaper, they said. It will be faster, they said. It will be easier, they said. Well, it’s possible for some of those things to be true, but the “easier” part, not so much.
Amazon Web Services, or AWS, is pretty famously obfuscated in how it works, and you can easily end up spending far more than you realized. I inherited an AWS installation in early 2020 that the company was spending $1,800 a month for the most basic website you could imagine, just a very simple WordPress site. As I went through and started shutting things down, they’d start back up again; I ended up in a weird dependency circle that was incredibly difficult to unwind. I was able, after much effort, to reduce the cost to $70 a month.
First, let’s talk about the hidden options. These options on data don’t present themselves until you mouse hover over a certain area; sometimes the area is big, sometimes it isn’t.
In the image below from the User list in Workspace, the area highlighted in yellow doesn’t appear until my mouse hovers over that line. The standard user flow is to look at the data and try to see where the function is that they want, not to keep moving the mouse around the screen, hoping for options to show up.
This is inconsistently applied throughout the system, which makes it extremely cumbersome to manage an installation. Don’t even get me started on interfaces that require you to switch back and forth between the mouse and the keyboard. Talk about a productivity killer, then finding what you need is a real problem, which brings me to the meat of this all, the Search!
Google is famous for search, right? Ok then, why can’t I search for all the instances of a string in GCP or Workspace? Meaning, I type someone’s name, like Beth, and the only thing I get back is their user profile record. What if they have Google Voice? What about an email forwarder setup? I have to dig through a half dozen ever-changing menu layers to that particular subsection to see if the user is configured in it.
GCP is the same thing. Suppose I’ve configured a particular IP address to access various objects in the system because we have a tightly configured system to avoid unwanted intruders. In that case, I can’t search for that IP. So instead, I have to go into every subsection and configured object to find if it is there and then make changes. This is a lot of fun with folks that don’t have a static IP address.
So, they’ve obfuscated the user interface, they’ve taken to moving functions around inside that interface, while at the same time not keeping the Help system up to date, and then made Search not be particularly useful.
How is this progress? How does a company whose entire claim to fame is Search not implement Search in a useful fashion in their own widely used products? It boggles the imagination and makes the job harder than it needs to be if you are managing these systems. AWS is even worse, though, but that’s a story for another day.
In the olden days, time was spent refining the UI/UX experience, working with users, and calibrating for learning curve and speed for both new and seasoned users. I noticed something; however, when Apple moved to a flatter design and everyone followed suit. Windows Server 2012 was nearly impossible to work with; nothing looked clickable; it was like a sheet of paper. Google went to their Material Design, with positive and negative consequences.
We ended up with a “follow the leader” mentality and a desire to keep updating UI/UX so designers could justify their job. Unfortunately, this made it harder for users to relearn how to use their programs and ultimately made much less user-friendly interfaces, which study after study shows. So, note to Big Tech, clean up the mess you’ve made by working with actual users and not the echo chamber inside the company, and then stop touching it.