AI generated image —

The ROI question for not-for-profit technology investments

Krishna Kumar K
4 min readFeb 23, 2024


If I had bought NVIDIA shares in Feb-2023, my ROI would have been

232%!!! I missed the chance. That is not the point of this discussion.

Like many of us, I am from a middle class family. We spend every penny very carefully. We want good ROI for everything. Even when we go out on vacation.

Wait, ROI is return on “investment” right ? How does it make sense for our expenses.

Let’s say we spend money on flight, hotels and experiences during a vacation. What is the “ROI” ?

Let’s simplify the definition as (Current Value of investment)/(Cost of Investment). The denominator in this case will be $ amount spent on the vacation, but the numerator is the intangibles. The happiness you experienced, the stress you reduced, the amount of time spent with family and so on.

You cannot put a dollar value to it. Well, in a very convoluted way you might argue that if you don’t go for a vacation, you might get burned out, your health may suffer and you might end up spending more on your hospital and medical bills.

But this analysis is not needed in our decision making when we are choosing between different options.

What we are trying to see here is between different options, how can I maximise my numerator (the value, the intangibles) for a given denominator (budget). Or even reduce the denominator.

Let’s apply the same principles to a not-for-profit technology investment.

Let’s imagine that a city administration wants to promote healthy eating habits for its citizens. They come up with an intervention - a smartphone app that directs people to the nearest eatery with healthy food choices and personalised recommendations on what to eat based on their budget. If developing this app and promoting it costs 500 K dollars, how can they justify the ROI to their finance team.

It is very unlikely that this app will generate revenue because that is not the goal for this app. The spirit of ROI is relevant here, not the exact definition.

There are some key questions to answer:

Why is promoting healthy eating habits important now ?
There should be clear arguments on why this is relevant at this point in time based on prevalence of diseases, genetic factors and so on. Of all the investments in healthcare, this should come up as one of the important ones to make that can have a significant positive impact on health outcomes. Health outcomes is an end in itself. Future healthcare cost savings because of better health outcomes is a short sighted argument in my view.

Once this is established, the next steps is to discover the most effective way to promote healthy eating habits.

Among all the interventions one can think of what is the most effective one:

  • Encouraging movies and TV shows to promote healthy eating by giving incentives
  • Spending money on ads to raise awareness
  • Reducing cost of healthy food options by offering incentives to vendors
  • Developing and promoting a smartphone app that directs people to the nearest eatery with healthy food choices and personalised recommendations

To make a decision, what we need is estimates on the cost of executing each. We also need a measure of the impact and that is where experiments and pilots help. We then evaluate based on the following

  • Scalability — Maybe we are okay with the cost, but we want the intervention to scale across the population
  • Cost effectiveness — Maybe “promoting healthy eating habits” is important but is not the top priority, so we want to cap our investment here and perhaps choose the most cost effective one even if it is not the most scalable. The ‘ROI’ comparison will help here
  • Sustainability — Is the intervention sustainable in the long term ?

So the ROI question is valid but it is not the pure definition of ROI that should be considered here. I remember my finance professor saying — “we can compare apples and oranges as long as we know their dollar value”. I think we need not know their dollar value, all we need to know is — if we barter apples, how many oranges do we get :)



Krishna Kumar K

Product Guy. (Worked at Indeed, Microsoft ...). I write about product management, startups, analytics and machine learning. Occasionally I digress...