Help me buy!

Jul 13, 2012 17:50

Deciding what to buy is still broken! I recently wanted to buy a new phone, and boy, was it a tough thing to research. It required a combination of multiple visits to the store, lots and lots (and lots) of sites visited and revisited on the web. In the end, I ended up selecting a phone almost by serendipity.

Here's what I wish I could've done, and what were the problems encountered along the way:
  1. I should be able to say what the important specs are, and get back valid recommendations. Think filter. So, I can specify a minimum screen resolution and size, max price, good battery life, good camera, etc., and be presented with a set of choices. Seems like a fairly simple thing to do for all those sites with specs data to enable.
  2. Products recommended must pass a review filter. Obviously, the better the aggregate reviews, the better the product, but its almost never that simple. For instance, I almost bought the LG Optimus L7, but the reviews showed that it had a nearly obsolete processor. I ran into spec not holding out in actual reviews many, many times! An issue with flipkart reviews - many reviews for the phone seem to be made by people commenting on the specs, without even looking at the phone IRL!!
  3. Equally, a multitude of reviews from different sources were essential to get an unbiased and complete view of the product. There are a ton of "review" sites which are basically SEO bait - they take the specs as announced, pronounce that this "looks good" or the equivalent, stuff keywords - but which are completely useless. The ones which I consistently found useful, i.e., hands on reviews, were thinkdigit, phonearena, and gsmarena. phonearena was especially helpful, providing alternate suggestions. Even these, though, tended to be incomplete and inconclusive for positive decisions, and had to be enhanced by reviews by users who'd bought the product.
  4. I'd have loved to see how the phone looks (and better, feels) compared to my old phone, a milestone. High res side-by-side videos or photos could have helped, but the store was the only way to do this.
  5. There didn't seem too much price difference between online and offline for many models - at least those I was interested in. Hence, it would've helped to provide location specific availability and price info. MySmartPrice had the best price data out there, but it was only for online sellers.
When I started my company (WisdomTap, old website now defunct) I intended to solve this very problem. I intended to do it by scarfing up reviews from across the web - including enthusiast forums - and apply NLP to get detailed, feature specific reviews. While it would've helped somewhat, its clear that it wouldn't have actually helped fully.

I think that there is a crying need for tools to make it easier to decide what to buy. Maybe somebody will find this useful and actually solve this problem!
Previous post Next post
Up