mistakes learnemyElisha Tan is the founder of Learnemy, an online marketplace that finds you the right instructors and classes in Singapore. Follow her on Twitter at @learnemy.

I believe that I’ve overspent S$10,000 on developing the Learnemy web application.

Before I hired someone to code Learnemy, I tested my idea with only CSS, Wufoo form and a phone and this is something that I’m proud of: I got traction without actually having a product.

Pardon me for sounding arrogant (it’s not my intention) but I really would have never thought that I could make the mistake of spending on things that didn’t matter. Every bad decision that I regret making just seemed so right at that point in time.

I was spending money on things that didn’t bring in value and they mainly fall into two buckets: bad prioritisation of features and early optimization.

Note: Before I continue, I need to clarify that I’m not blaming anyone for the mistakes I’ve made. Everyone has their own opinion and I take ownership for the opinions I agree with. And ultimately, I had the authority to make the final decision.

Bad prioritisation of features

I used to envision a feature that would direct visitors from overseas to a landing page, instead of the home page, so I can collect their emails and notify them only when I expand to their countries.

Guess how much it cost me (in terms of development fees): S$800.

Guess how much it made me? S$0.

This feature could be replaced by a simple “Only available in Singapore” on the home page, and specific calls to action for people who are overseas. Or if they have enrolled in a class they didn’t realize is happening in Singapore, it’s not difficult to do a refund.

This is just a prime example of the kind of mistakes I’ve made.

Update: Elisha added, “The $10k was calculated based on the lump sum I paid to hire contractors to work on tasks that didn’t matter. The $800 was one task that hit me the most about how I was spending my money, that’s why I remember it so specifically.”

What I’ve learned: How to prioritize features

Features should be prioritised based on how much money that feature can directly or indirectly make. Features that will make most money should be coded first.

I really understood this when I went for a scrum course. It gave me a framework that I used to think about what features should be coded, and the order that it should be coded.

Early optimising

I’ve read so many mistakes about early optimization that I did not expect myself to fall for this. But I did.

My app has gone through the hands of three people, including myself (a newbie), so the codes were messy. So a contractor gave me this analogy:

“The state of your codes is like a house built on unstable ground. You can add more levels to the house but since the ground is not stable, this is not a long-term solution. To add more features on this app, you will create a lot of bugs and you will end up spending more time fixing bugs than moving on. I would suggest you rebuild it from scratch.”

This suggestion is good. I took it. But at that point of time, the startup wasn’t making much money. So I shouldn’t focus on long term, but you know, staying afloat.

At that point, there was not enough user data to know what users want, so the plan was to redo pretty much the same thing. I only realized that this doesn’t make sense when the V2 codes were halfway done. But money was spent and the codes weren’t used.

What I’ve learned: Optimize on branding and user experience first

Branding and user experience matter more than the quality of the code because of the nature of my startup. The whole matching process can be done via phone and email like the many tuition agencies in Singapore. The brand and reputation is worth a lot more than the codes.

These are the mistakes I’ve made, and what I would have done if I can go back in time. Maybe what I thought I should have done instead are not the right answers either; I won’t know. But entrepreneurship is all about finding out what works best.

What mistakes did you have, and what have you learned from them? Share them in the comments below.

This post was first published on Learnemy’s blog as 2 mistakes I made that cost me $10,000.