Guessing, Tracking, and Learning: My Experience with Effort Estimation

01 May 2026

Guessing, Tracking, and Learning: My Experience with Effort Estimation

When I first saw that we had to estimate how long tasks would take for our Bow lletins project, I honestly did not take it too seriously. Most of the time I did not fully understand the task yet, so trying to guess how long it would take felt random. I thought I would just figure it out as I went. But after actually estimating and then tracking my real time, I started to understand why this process matters.

How I Made My Estimates

At the beginning, my estimates were mostly based on what I felt the task looked like. If something seemed simple, I would say maybe one or two hours. If it looked more complicated, I would guess four or five hours. Sometimes I compared it to something I had done before, like building a page or working with authentication.

Even then, I was often wrong. Some tasks that looked easy took much longer because of bugs or unexpected problems. Other tasks that looked hard ended up being quicker once I understood what to do. Estimation IMG

Did Estimating Actually Help

Even though my estimates were not accurate, they still helped me. Estimating forced me to slow down and think about what I was about to do instead of jumping straight into coding.

One example from my project was fixing the navbar minimization issue. I estimated that it would take about eight hours because it involved layout problems, dropdown behavior, and responsiveness. In reality, it took me about six hours of coding and one hour of non coding work.

Even though my estimate was not exact, it still helped me understand that the task would take time and attention. I did not treat it like something quick. It also helped me stay aware of how long I was spending on it.

Tracking My Actual Effort

Tracking my time ended up being more useful than estimating. I kept track of when I started and stopped working and separated coding time from non coding time.

For the navbar issue, my breakdown looked like this:

Estimated effort: 8 hours
Coding effort: 6 hours
Non coding effort: 1 hour

The non coding time included planning how the navbar should behave, researching how Bootstrap and Next.js handle layouts, and testing different fixes.

This showed me that a lot of the work is not just writing code. A big part of it is thinking, testing, and figuring things out.

What I Learned from Tracking

One of the biggest things I noticed is that I underestimate how much time I spend not coding. Planning and debugging take a lot of time, sometimes just as much as writing the code itself.

In the navbar issue, I spent about an hour just testing different ideas before getting the final result. That time was important because it helped me avoid breaking other parts of the layout.

Tracking also helped me improve my future estimates. After seeing how long tasks actually took, I started adjusting my expectations. For example, anything involving layout or user interface changes usually takes longer than I first think.

What I Would Do Differently

If I were to do this again, I would break tasks into smaller pieces before estimating. I would also track my time more consistently instead of trying to remember it later.

I would also pay more attention to non coding work because that is where a lot of time goes. Using a timer or tracking tool would probably make my data more accurate as well.

AI Use in Effort Estimation

I did use AI tools like ChatGPT while working on my project, mainly to help understand problems, debug issues, and generate ideas. I did not use AI to directly estimate how long tasks would take.

When I used AI, I spent time writing prompts, checking the results, and fixing the code so it would work in my project. I counted that as coding effort because it was part of the development process.

Most of the time, I had to edit or debug what AI gave me before it worked correctly.

Final Thoughts

At the beginning, effort estimation felt like guessing. By the end, it felt more like learning how I work. I may not have been accurate, but I became more aware of how I spend my time and how to plan better.

That made the whole process worth it.

AI Use Acknowledgment

I used ChatGPT to help clean up the wording and structure of this essay. The ideas and examples are based on my own experience working on the Bow lletins project.