ROI Calculations On Custom Software Projects

‍

‍

Invene significantly grew in 2021. For those who aren’t familiar, we create custom software for fast-growing HealthTech companies. COVID-19, while bad for more traditional companies, has been very good for the Health IT sector. As such, our existing clients have continued to expand their relationship with Invene, and we are always landing new projects.

‍

As CEO of Invene, I’m always on the lookout for ways to make the company better. As such, one of my strategic goals for 2022 was building some custom software to help Invene’s internal operations scale more efficiently. Invene has never built an internal tool before, but due to our rapidly expanding team, I decided it was worth investigating. As such, I’ve detailed how I personally approach the software development process that I take with our clients. My hope is that it helps you make better decisions on what software you ultimately decide to build.

‍

In my opinion, the first and most important step in any software development is the return-on-investment (ROI) calculation. Most software projects that fail, fail because they should have never been built. Even if the program works flawlessly, it is destined to fail if it has no realistic ROI. At the end of the day, businesses are about making money. Custom software makes the company money in one of two ways: decreasing costs or increasing revenue. If the custom software cannot do this, it’s better that it’s not built.

‍

‍This was true in 2022. It is even more true today. GenAI tools have made it faster and cheaper to build software than at any point in history. The barrier to start a project has never been lower, which means the discipline to evaluate whether you should start has never been more important.

‍

The ROI you need is dependent on how big your company is. Smaller companies need bigger ROIs than larger companies. For a small business like mine, I’m generally looking for a 3x to 5x ROI within 12 months of finishing the product.

‍

Below is a real-life example of evaluating the ROI of our 2022 strategic initiatives. It only took a couple of hours, and I won’t spoil it, but it was well worth the time spent.

‍

Invene’s Strategic Software Development Project

‍

I wanted to build out internal tools to handle invoicing, payment, and reporting. Right now, my awesome Executive Administrator is manually doing this. She’s doing a great job, but the amount of work is increasing overtime as we add new employees, contractors, and clients. As such, I wanted to build a suite of software tools that handles some of this and she’ll act as a manual override. It’s not realistic to fully automate this process due to the constantly changing nature of Invene’s business. When building custom software, especially internal tools, it’s almost always better to have the software output a result and then have a human modify it. This avoids the problem that occurs with business software where programmers play whac-a-mole with a million different edge cases and no one is ever happy with the end-result.

‍

How Many Hours To Create?

‍

Engineers

All of Invene’s engineers are currently assigned to client projects. As such we have no spare resources to develop this software. That’s where the UTDesign Computer Science Senior Capstone Project comes into play! This is where three to five senior students at the University of Texas at Dallas (UTD) work on a software project for a semester. I’m a very proud alumni from UTD and my first two employees came directly from this capstone project. As such, I love sponsoring capstones as it allows Invene to try out and recruit junior-level talent. You wouldn’t want to use college students for production critical projects, but for an internal automation tool like this, it’s a good fit. P.S: Interested in paying for a UTDesign Senior Capstone Project? Feel free to reach out to me and I’ll give you an intro to the person who runs it! It’s a great program.

‍

‍

‍

As part of the UTD Capstone Project, one of my engineers must be Invene’s corporate contact to ensure students are hitting deadlines and to provide technical mentoring. This will give one of my engineers the opportunity to start building up their leadership skills in a low-risk environment. I predict it will take five hours a week for two semesters. Since we are a services firm, when calculating ROI, I use my engineer’s billable rate, not their actual cost to the company. The reason behind this thinking is that every hour the engineer spends on an internal project is an hour not billable to our clients. In my business, time is money – literally. 

‍

‍

UX/UI Designers

‍

While we don’t need UX/UI designers for this project, they help make the software look better. If I’m going to spend time developing an internal tool, I want it to look good. While there’s many programming libraries out there that allow developers to create decent looking applications, it is no replacement for a great UX/UI designer. Luckily, this is a simple application. I figure it will only take my UX/UI designers about 120 hours to throw together all the screens. 

‍

Domain Experts

To ensure the success of the project, it is a good idea to have a domain expert involved in the project. As such, my executive administrator would be the go-to person since she’s the one that will be handling the software manually. I’m assuming that takes five hours a week of her time for 26 weeks.

‍

Executive

Software isn’t all about the individual contributors to the project. It also requires a stakeholder’s time and buy-in to be a success. As such, when calculating ROI, I always make sure to include how much time the C-Suite, VP, or whoever, will spend ensuring the project is heading in the proper direction. For this project, I’m assuming it will take me around 30 hours to write requirements, get my team onboard, test, attend the final pitch meeting, etc. 

‍

‍

‍A note worth adding here is if I were running this calculation today, the engineering costs would likely be meaningfully lower. GenAI coding tools have changed what a small team can produce and how fast they can produce it. That changes the inputs, but it does not change the framework. The math still has to work.

‍

‍

‍

‍

What About Intangible ROI?

‍

Ah, but what about intangible ROI? The executive administrator and engineering lead will be gaining invaluable leadership skills. This will allow Invene to test out and recruit great CS students from UTD. It would also be a long-term asset of the company as we would own the IP, not our clients. It would free up the executive administrator time to work on more important tasks.

‍

Not to mention, what if I write a whole series about the journey of building this software start to finish? Surely a top-tier healthcare executive will read the blog posts and think “man, this James Griffin guy, sure knows how to build software” and we’ll close a multi-million deal because of my writing. Think about all the revenue this internal software would make us!

‍

All of these thoughts swirled around my head when I saw just how bad the ROI calculation was. They may seem ridiculous, but it’s true. I really want to do this project! It’s one of my 2022 strategic goals and I had already told UTDesign I had planned to do 2 capstone projects with them. I came up with all sorts of intangible ROI that made this project a no-brainer.

‍

But here’s the thing about intangible ROI – it’s almost always a load of bull. It’s a way that we, as executives, delude ourselves into thinking our project will be worth it, while in fact, it is better to axe it before we even start. Our time is so valuable. Unless it’s a massive return for the business, a project like this isn’t worth it.

‍

‍

‍The GenAI Excitement Loop Is the New Intangible ROI

I'll add one more trap that's become more common recently: the GenAI excitement loop. A team gets access to a new AI tool, builds a prototype in a weekend, and suddenly a project that should go through full ROI scrutiny gets fast-tracked because it "only took us two days to build." Speed of prototyping is not the same as proof of value. The math still has to hold.

‍

Roi Calculations Save Time

‍

This project had been one of my big 2022 strategic goals. However, the ROI calculation forced me to be honest with myself and take a realistic view of what Invene would gain. As such, I removed it from my list as it would be a massive waste of our time with very little gain. If a client asked Invene to build the software above, I’d instantly tell them no and advise them from building it.

‍

I absolutely despise ROI calculations with a passion. However, it’s the most important part of the project–internal or external. Otherwise, you get six to twelve months in before you realize you didn’t accomplish anything. Oftentimes, killing a project can be the best thing you can do for your business.

‍

When GenAI Makes the ROI Work, Own It

‍

Here's the flip side, and this is the part that genuinely excites me. When the ROI does hold up, GenAI changes everything. You no longer need a massive budget or a big vendor relationship to build something real. You can move faster, own the IP, and stop waiting on a platform roadmap that was never designed around your workflows. The organizations that will win in the next five years are the ones that run their numbers honestly and then move decisively when the math works. That is owning your destiny.

‍

Frequently Asked Questions

‍

How does Invene help organizations evaluate projects before committing?

‍

Before any build begins, Invene embeds with your team to understand your current workflows, data environment, and operational gaps. We map where the real cost drivers live, where automation actually makes sense, and where the numbers are unlikely to work. The goal is to give you a clear, defensible picture of what's worth building and what isn't before a single dollar of engineering budget gets committed. Contact us to learn more.

‍

How do I know if my software idea is worth building?

‍

Run the numbers before you do anything else. Map out your build costs, maintenance costs, and realistic savings or revenue gains. If you can't hit a 3x to 5x return within 12 months, the idea probably isn't ready yet. That doesn't mean it's a bad idea forever. It means the timing or scope needs work.

‍

Does GenAI change how I should calculate ROI?

‍

It changes the inputs, not the framework. Build costs are lower, prototypes come faster, and smaller teams can do more. But maintenance, hosting, and the cost of internal time all still apply. And a fast prototype is not a business case. Run the same calculation you always would, just with updated numbers.

‍

What's the biggest mistake executives make when evaluating a software project?

‍

Trusting intangible ROI. The moment the hard numbers look bad, it's tempting to start listing leadership development, brand visibility, and long-term strategic value. Most of the time that's a rationalization, not a reason. If the tangible return isn't there, be honest about it and move on.

‍

Should we build internally or rely on a vendor?

‍

That depends entirely on your ROI calculation. What's worth asking is whether you're relying on a vendor because they're genuinely the best solution, or because building felt out of reach. GenAI has closed that gap significantly. If the math supports building, owning your own IP and roadmap is worth serious consideration.

‍