News

Megite : Interview with Matthew Chen of Megite.com

Today I had the chance to speak with Matthew Chen, the co-founder of Megite, which is a real time news aggregation service. You’re probably familiar with their model but if you’re not, it’s kind of a mix between Google News and Digg in some ways, but focuses on breaking news throughout the blogosphere and then the reaction to that news or opinions stressed by other bloggers.

 Megite : Interview with Matthew Chen of Megite.com

In a time when blogs (even news organization run blogs) have become the instantaneous messenger of the daily news, services like Megite are becoming more popular.

I’ve known of Megite for a while due to the large amount of referrals Search Engine Journal receives from them from time to time, and thought that since social news, linking and blogging have become such an important piece of the SEO, SMO (Social Media Optimization) and online marketing universe, that the chance to speak with Matthew about his service would be quite valuable to our readers.

We discussed his site, his video service, blog marketing and touched upon web publishing and monetization. Enjoy :)

Loren Baker : Thank you for taking the time to speak with me today Matthew, for starters, What is Megite?

Matthew Chen : Megite is a social news aggregrator with smart discovering and ranking algorithms. Mainly it is based on the blogs and news mining.

Loren : How does Megite choose and rank the stories and discussions on its homepage?

Matthew: 1) First Megite has a crawler that can fetch blog posts and news from web;

2) Second Megite needs a list of seeds to start;

3) Then it comes to our algorithms: Megite analyzes every new posts and mines the relationship between posts and news; It also uses natural language processing techniques to analyze the content; and mine the web for the positive and negative feeds (votes)

4) Then Megite ranks the blogs and stories by several factors: popularity, importance, freshness,
opinions from web, etc

We are going to adding click through data and social recommendation in the near future.

Loren: What do you mean by mining the web for the negative and positive feeds? Do you mean comments, links to that post, possible votes on Digg or Reddit, or opinions of a blog post written by other bloggers?

Matthew: Right something like that. From the text description, we need some techniques to decide whether it is positive or negative. For the number and star rating, things are easier.

Loren: Possibly an incentive for bloggers to add rating options to their blogs?

Matthew: It could be. But never over do :)

Loren: How can web publishers include their blogs within Megite?

Matthew: It may be selected by us as the starting seeds; Or a blogger can submit blogs to us; Or we mine the logs to find new popular blogs; What is more, the Megite algorithm can auto discover new blogs and add them to the system.

Loren: You recently launched Megite Video, could you please explain that a bit?

Matthew: We adopt the same infrastructure and algorithms to Video aggregration. It aggregrates the popular videos measured by number of blog posts, votes and discussions, etc from more than 20 video sharing service sites.

Then, we present them in the same way as other Megite channels.

The difference from other video recommendation services is that we introduce the video clustering and put related videos together. In this way, a user can see videos from different topics.

Loren: I receive a substantial amount of traffic from Megite, where do you receive most of your traffic from?

Matthew: In terms of geo location, some big part is from North America. For the refers, we see lots of blank or empty refers, we guess they may be from bookmarks or just enter into browser navigation bar. Also we see significant traffic from rss readers or blog posts

Loren: What about from services like StumbleUpon, get much referrals from them?

Matthew: Yes. The traffic from them is sudden birth. We get huge [traffic] from them one day (or several hours), then drain very quick.Their traffic pattern is very interesting.

Loren: I along with some other bloggers I’ve talked to tend to sometimes confuse Megite & Techmeme services as being the same. Is there a difference between Megite & Techmeme?

Matthew: One difference is that we introduce the auto discover of new blogs, so not just A-listers. So the coverage should be broader. But in some sense, it may affect quality a little.

Another difference is that we introduce the personalized megite by giving the read list (either a list or in opml file). For example, recently TopRank asks us to create one for them at http://www.megite.com/toprankblog we even create one for techcrunch from the opml file in share.opml.org at http://www.megite.com/techcrunch and one for search engine watch from the blog lists on their site.

People like this because they can get a quick glance what is happening now. We see Megite as an intelligent RSS reader, not just a simple meme.

Third, we cover more verticals because it is very easy for us to add new verticals, normally one or two hours work.

Loren: Any spin-offs in the works? Or will you be keeping all of the verticals under the Megite.com domain?

Matthew: The good thing to keep in one domain is that a user can find most of interesting stuff in one place, for example, a persona may like to read technology and entertainment. The problem is that the main page is only branded to one vertical. We are thinking to pick top n from each vertical into the homepage.

But one thing we are considering is that for different language, we can create different domains
Currently we have some inquiries about other languages, they are trying to partner with us.

Loren: Have you been following the direction of Findory and their personalization technology? Think Megite may be able to identify a person’s niche or choice of news and serve blog stories based upon that profile?

Matthew: Profiling is very hard, specially for news. Users may have interesting on some topics, but they don’t want to always see the same topic again and again. They always look for new stuff. So we may not go the way like Findory.

But one thing we are doing is that we can recommend popular or related blogs to a user if it is not in their reading list.

For Megite personalization, we are more inclined to help them to manage their read lists and provide add value service for their reading behaviours.

It is risky to guess what they need.

Loren: Here’s a question for you, how does Megite make money? Techmeme has their sponsorships, Digg does AdSense and Federated Media, how about Megite, I don’t see any ads on the site.

Matthew: Several revenue channels:

1) Provide hosting services who want to create verticals (or license our software, but it is risky to give out code)
2) Sponsorships
3) Ads
4) For the personalized megite, we are thinking about about.com model, sharing revenue with creators

Loren: What type of sponsorships or ads are you currently offering, or is this something in the works?

Matthew: We are working on that. Currenty we have several site sponsors in talks.

The good thing for Megite is that we don’t have big spending, so we are not eager for funding or money. Make our users happy is the highest priority.

Loren: That’s a nice priority to have. What did you do before Megite? And is this a full time venture for you?

Matthew: I am a software engineer. I worked in startups before, started startup before. I do other consultant work to make a living. My partner works full time.

Loren: Excellent, well is there anything else you believe our readers will find useful, in terms of search or blog syndication?

Matthew: Several things may help them:

1) Create personalized (vertical) Megite to provide more coentent;
2) Create River of News service to provide real time news;
3) Improve the quality their blogs and get more attention from memetrackers

Loren: Well then, what 5 tips can you give for improving the quality of blogs?

Matthew: Have RSS output with full content, users will appreciate that by have less clicks.

Loren: Yes, less traffic to the blog but perhaps more readers in the long run.

Matthew: Understand users needs and write with expertise; Keep track of new content, try to be the first one to break new stuff (like techcrunch); If not, then have own opinions and insights, agree or not agree with big names; Keep working every day.

Loren: Right on! Thanks Matthew, it’s been a pleasure.

Matthew: Thanks Loren

If you have any questions for Matthew, follow up comments to the interview or would like to comment on Megite and social news aggregation, please feel free to leave a comment below.

Screen Shot 2014 04 15 at 7.21.12 AM Megite : Interview with Matthew Chen of Megite.com
Loren Baker is the Founder of SEJ, an Advisor at Alpha Brand Media and runs Foundation Digital, a digital marketing strategy & development agency.
Screen Shot 2014 04 15 at 7.21.12 AM Megite : Interview with Matthew Chen of Megite.com

You Might Also Like

Comments are closed.

4 thoughts on “Megite : Interview with Matthew Chen of Megite.com

  1. Great interview.

    Whilst I am writing this comment I am attempting to probe Matthew into giving the secret of the natural language processing away and how to influence it ;)

    Your readers might also be interested in the Google Toolbar Button “search mashup” I knocked together.

    It searches using one of the CSEs created based on Lee’s OPML file (not my CSE), and includes the Megite feed based on the same data.