What gets measured gets done. You’ve probably heard this before (even though experts argue where the saying came from). With the constant flurry of data coming our way, it’s hard not to want to measure everything. Any little bit of data can
help, right?
Kind of. There’s a caveat: data saturation is a real thing. It happens when you measure too many things and
you get drowned in data. You have it, it looks pretty and insightful, but it can cause decision paralysis.
When you create your own software, you naturally want to measure as many things as possible. Anything you measure can be tracked and improved upon—and it’s tempting to track everything “just in case” you want the data in the future.
Before you go down the analytics rabbit hole (which is incredibly hard to escape), focus on tracking what matters.
What Are Your Goals?
At Far Reach, this is where we start with all of our clients when they want to develop new software, add to existing software, or measure a new release. What are your goals?
Without a clear answer to this question, why bother measuring anything at all? If you know what you want to achieve, the metrics you need to measure will be much easier to narrow down.
So, before setting up four different analytics apps or pulling 10-page reports from your software and attempting to make sense of it, take a moment to think about what you want to achieve, when you want to achieve it, and what metrics will tell you you’re
there.
Are you looking to boost productivity? Increase usage of a new feature in your software? Retire some of the features very few people are using? Improve some of the existing features? Reduce subscriber churn?
All these are valid choices. And for all of them, data can help.
Tracking within your software is typically pretty simple—there’s likely a lot of information in your database and there are many plug-and-play solutions like Application Insights that can help. When your goals are clearly set, you can work out which metrics to measure and use these tools to make it happen.
What Should I Be Tracking?
First off, there is no one-size-fits-all when it comes to software analytics. Just because your competitor tracks certain metrics doesn’t mean you should too.
In addition to your goals, there are other factors that will determine what you should track. For instance: is your software for internal or external use? Internal software may have different performance criteria than one that is available to your customers
or external partners.
When software is used internally only, you may want to focus on metrics connected to productivity and processes. On the other hand, when your custom software is used externally, things like satisfaction, ROI, and usability take center stage.
You can (and likely will) create your own combination of metrics to measure and report on. And as with everything in agile/scrum, you can always add, remove, or reprioritize them, depending on what your data reveals or on how your goals change.
While there is no ideal combination of metrics that applies to every custom software system, there are some
common things many companies track.
1. Most and Least Used Features
Knowing what features users mostly rely on can help you decide which of them to improve first. Similarly, when you know what features aren’t that popular, you can make a data-driven decision about them: should you improve them so they gain more
users or retire them altogether?
2. Who’s Using Your Software
You can track this metric in external- and internal-use software. In the first case, it will tell you a lot about your target audience and how you can improve your solution. You can learn which companies and users within those companies are using your
system and the features within it, which can help with decisions about future priorities.
In the second case, it might reveal software flaws—or human flaws, like resistance to change—that you can address.
For instance, if you created a solution to be used company-wide and discover that only a handful of people from a few departments are using it, then you may have a usability or change management problem. If team members find the software difficult to
work with or clunky, they may opt to do things the old-fashioned way. If you discover Gary in accounting still doing reports the old way, you can use data to figure out how to make the software better and make Gary’s job easier if he uses it.
3. Errors
The first step in getting a top-notch custom application is to always know what doesn’t work. We recommend proactively monitoring and reviewing errors in custom software systems. In many cases, you can quickly deploy a fix and avoid major issues.
In other cases, the error may not be mission-critical and you can spend more time coming up with a long-term fix.
4. Performance Issues
Small performance issues, when fixed, can make a world of difference. A slow software system has a big impact on workflow and team members. Common issues include poor-performing queries or pages that have unnecessary database calls. By fixing problems like these, you can improve the application’s performance and, with it, the productivity
and the performance of its users.
Analytics Are Just One Part of Custom Applications
There's a lot to think about beyond analytics when you're building custom software. Learn 11 things to consider in your next development project.
How to Use Your Analytics Data
Once you’re tracking the right things in your software—you know what works, what doesn’t, and what could use some improvement—now what?
Our recommendation is to use the agile/scrum approach and work iteratively to prioritize and implement fixes and
enhancements in your system. Remember that custom software development is not a one-and-done project. Continuous development is what ensures the success of your custom software. Even if you launch with a stellar product, your company goals, the
market, and the industry you are in will change. It’s only natural that your software should evolve along with them.
As long as you keep a keen eye on your goals and make a purposeful effort, tracking the right metrics and then acting
upon them can become a routine part of your workflow. Want to talk about this further and get insights from our experts?
Reach out!