I was recently invited to review a website, Woya.com – a consumer electronics shopping site. At first I didn’t really think that it was relevant and was going to politely decline, but I diligently went and took a look at the site expecting to see familiar price-comparison features. I was surprised to find something that is more like an online shopping analytics package, there is some very clever tech going on here which I felt was worth writing about.

All the things that you should be doing when you shop online

Let’s face it, if you are a smart online shopper, you trawl different sites, compare prices look at availability etc. Most of us, however, are too lazy and just go straight to Amazon, ebay or whatever shopping site we prefer. Woya does all of that trawling for you and presents a wealth of information to help you make informed purchases. For every product it looks for retail, auction AND classified listings.

Woya gives you easy access to all of the information

For any product, you can see the current Retail, Auction and Classified listings available. You can see the product specifications, and related products and accessories. Very handily, there is a page to see all of the user reviews aggregated together with filters available for positive, negative, most helpfuland most recent.

The compete page seems like an interesting concept, comparing the current product with similar ones based on price and user rating.

There’s more than just information

Where Woya starts to get really clever is in the analytics based features. In addition to aggregating all of the published information, it also provides an overview of trending prices, availability and simple advice based on the analysis.

Save yourself some effort

If you are looking to buy a specific product at no more than a certain price, the DealDigger function is a great idea. It lets you set up the parameters for your target price, new or used condition and whatever combination you want of retail, auction or classified listings. DealDigger will then send you an email when it finds the product with your matching criteria. You will be free to get on with your life safe in the knowledge that Woya is doing all of the hard work for you.

Social Shopping

One aspect of the site which is easy to overlook is the integrated social networking. It’s likely that these features will become more useful when the site has a large community of active users.  As far as I could tell, the idea is that you can list items that you own, recommend or want. For any product, you’ll be able to see if other site users own, want or recommend that item. Hopefully you would be able to contact another site user who owns the item to ask them specific questions about it but that is just speculation on my behalf.

Interview with the founder

I asked Qiang Sun, the founder and Chief Code Monkey for Woya, some questions about how the site came together.

What is the biggest technical challenge in building / running the site?

A couple of things come to my mind. 

One of the unique features of the site is to incorporate data from unstructured data sources (e.g. classified listings) in the price comparison. The listings need to be properly categorized under products in our catalog. This requires that the system be smart enough to “understand” what product each listing describes. It is an extremely difficult problem to solve not only because there are many different ways to describe the same product but also that related products may share similar descriptions. Furthermore a seller may intentionally use less straightforward description in order to make his listing stand out.  We have developed sophisticated algorithms using natural language processing and statistical analysis to successfully tackle these issues. What we are able to achieve is pretty amazing IMHO.  As a result, our users can access a wider range of data sources during price comparison and set up alerts to get notified when the desired products show up on sites such as Craigslist.  No other website is able to offer similar services as far as we know. The result from such an AI system can never be accurate 100% of the time and we are constantly making improvements.

Running the site on a small number of second-hand Dell servers is also challenging. The system is built on dynamic data. Millions of deals for hundreds of thousands of product need to be refreshed at least once a day from a dozen different data sources. Scores of crawlers and analyzers are running around the o’clock to get the job. Since I am the main developer of the site and the sole IT person, it can be difficult at times. But this is more of a resource problem that is common among all early-stage startups.

There is so much information on view and so many features, what sort of process did you go through in order to come up with the site design?

Well, the short answer is that I wished we had spent a bit more time streamlining the design of the site   The site itself is only about one third of the work as we have complex backend/middle tier which was under development before the site was designed.  There is sort of a master plan – a blueprint based on how I envision the ultimate shopping site should be.  But there are so many things we want to build but so little time/resource, we decided to focus on price analysis for version 1.0.  The major sections of the site and the navigations between them were nailed down over the course of only a few days.  A freelance designer helped with the visual design of the site while we continued working on the backend and middle tier.  The UI design went over two quick iterations based on feedback we solicited from friends and family.  The initial HTML slicing was done by a 3rd party after which we started to put all the pieces together.  Due to the team’s lack of web development experience, the prototype of the site relied heavily on native ASP.net controls. Over the past few months these controls were replaced with [custom] controls and more Ajax features.

Are there any specific changes or additions that you are working on for the site?

A key focus for the next couple of month is to re-organize/re-skin certain sections of the site.  The site is meant to help user analyze shopping related data with ease but during the course of development we did not do a good enough job staying with this principle.  As a result, the UX was compromised.  Information is not always presented in an intuitive manner and the complexity also leads to confusion for some users.  A set of changes will be introduced in the coming months to address this problem. 

Has user feedback shaped any decisions in the development of the site, and were there any surprises?

Yes, absolutely. Since the beta went online a few months ago, we have received lots of feedback especially in the area of the UX.  A large percentage of the feedback is related to the issue mentioned in answer to you previous question.  It is not a total surprise because we knew that things were rushed through during the design phase so it is inevitable that this kind of problem surfaces.

The Site is US only at the moment, do you have any plans to expand into other countries?

Not anytime soon. We plan to expand the product catalog to go beyond consumer electronics first.  And in parallel we plan to introduce data analysis in areas other than price.  International expansion will only be considered if we can build a successful business in the states first.

It’s often the case that UI and UX can be somewhat neglected on projects that are heavily technology based and it’s good to see that this issue has been acknowledged and is being dealt with. The site does a good job of presenting the data available but as a developer, I’m comfortable with complex interfaces and I can see why less technical users may find some of the pages slightly overwhelming.

I wish Qiang every success with Woya and hope that it won’t be too long before there’s a UK version available.