February 24, 2016

My Niche

As a lowly grad student, toiling away in my single core computer I was often intimidated knowing that when I submitted a paper to conferences, it would be put up against those from the likes of Microsoft Research, Google and Yahoo! Labs. How could I ever think to compete against all those collective years of experience and seemingly endless computer resources? My adviser’s advice was not to go up against their strength, but to find my own. Big corporations that answer to share holders and consumers are not as free to tackle every idea. Her advice was to find something I could differentiate myself with, and focus on that. Find my niche.

It’s good advice that’s as applicable to starting one’s own business as it is to research.

I have been intimidated from doing the financial apps for many of the same reasons. Need a mortgage? Type in “Mortgage Calculator” (or even just “Mortgage Calc”) into Google and you get one on the search results page. How could I ever compete against Quicken and Mint?

Find my niche.

There’s a reason most mortgage and loan calculators are designed to only handle fixed rates, besides just simpler math. I imagine most people who are calculating how much a mortgage will cost them over the long haul are at least a little risk adverse, and those who are at least a little risk adverse are probably more likely to prefer fixed rates. But as any good financial person knows, stability can be more costly than risk and there are times that an adjustable rate might make more financial sense. I can either go after a small percentage of a large market, or a large percentage of a small market. I choose the latter.

So here are the ways I’m trying to differentiate myself from everyone else:

Smart Input Parsing

I’ve always found it mildly irritating when online forms choke on input that isn’t exactly as expected. Ever have to re-fill the whole form because you forgot to put dashes in a telephone number? Or because you only filled out two digits of your birth year instead of all four? It should be obvious to the computer what you meant. If a human can guess what you likely meant, I want my apps to as well.

One of the recent ways I incorporated this philosophy is with the interest rate on my loan form. Whether you type in 3%, 3, or 0.03, the form assumes you meant a 3% interest rate. That’s because a 300% or 3 hundredths of a percent are both rather unlikely interest rats. Possible, (I’ve seen some pretty predatory payday loan rates) but unlikely. It’s also unlikely that someone typing ‘3’ meant ‘300%’ given the way we humans tend to express percents.

More than Just a Formula

I love math. When I first started thinking about creating a webapp business I wanted to focus my efforts on bringing math to problems where math was traditionally not applied, such as the Labor Prediction Calculator. I’m especially proud of the math that went into the Image Color Palette, a mixture of machine learning, color theory and vector space math.

If all my app is is a implementation of formula, especially a standard well-known formula, there’s nothing to separate me from anyone else. Users have no reason to choose my app over the competition.

I especially like it when I can add a human touch to it. Sadly, not everyone shares my love of numbers. Numbers can even be a bit off putting. I like the way my Miscarriage Odds Reassurer tries to put abstract probabilities into perspective.

Data, Data and more Data

Data is definitely my thing, and one of my strengths that’s rare to come by. The more data I’m able to incorporate into my apps, and distill into meaningful results, the more I will differentiate myself from all the others out there.

Lately the data I’ve been munging has been my web traffic patterns. I hope to create some more data driven apps in the near future.

Now I just need to turn these disjoint niches into a single philosophy I can incorporate into my business. A task for tomorrow.

Posted in Work Life | Tags:

Leave a Reply

Your email address will not be published. Required fields are marked *