Interview: Tracour, combating tech rumors 1 fail at a time

The Internet is filled with content and the accuracy of the content has always been subjective to the author that creates it. Specifically, in the technology sector, we see tons of rumors and content that is not always factually correct. This issue has been around as long as the blogs have been driving traffic for advertisement revenue as we all know, a well sourced rumor is golden for traffic but when content is poorly crafted, who holds the creator accountable? 

Recently, a website launched that is devoted to putting technology news rumors "on notice". The site is called Tracour and it was founded Neowin's own Senior News Editor, Brad Sams and developed by Braeden Petruk. Since Neowin posts more than its fair share of news stories based on rumors from other sites, it would seem like this kind of database would come in handy. Sams recently gave us some more information about Tracour.

First, what can you tell us about Tracour?

Tracour.com is a content accuracy engine. We track rumors from all of the major tech websites and then rank them on their accuracy. In short, we can tell you how accurate Engadget is, or the likelihood of a rumor from The Wall Street Journal is to be correct. With our database of over 800 data points (and growing rapidly) we can identify key sources of information and reduce the noise around rumors.

How did the idea to come up with a web database for evaluating tech rumors come from?

We have all seen them, the rumors out of left field (Apple making a TV anyone?) that seem to generate a lot of noise but never come true. What we are doing with Tracour, is holding websites accountable for the information that they publish, if you start a rumor, we will track that content for accuracy. After the rumor expires, we will rank the end result and apply a statistical average to the domain so that consumers can see how reliable any given domain or author is when they write a report. The idea is simple, hold those accountable for the information that they propagate.

How does the database work in terms of putting in information about tech news rumors and gauging how accurate they are?

Right now we are crowd sourcing content. When a new rumor hits the wire, it is submitted by a user on our website (tracour.com) or on Twitter. If you are a user of Twitter and see a rumor we should track, simply tweet “Track @tracour Paste Link Here” and it will be added to our moderator queue for review. Once a moderator has approved the rumor, it will be give a trigger date. On the date, a moderator will be alerted to rank the rumor to see if it was fulfilled. The moderator can score the rumor on a scale of 0-100 (0 being incorrect and 100 being perfect) which will then be applied to the domain and authors scores that are public facing. We do have plans to automate the flow of the site and have the algorithm to do so, but seeing as we have very limited resources at the moment, we are using crowd sourcing to keep costs contained until we can expand our operations.

Does a low score for a writer on the Tracour site basically mean this writer's stories are not as reliable as others?

Correct, if an author or domain has a low score, it means that they are missing the mark for accuracy of their rumors. It also means that writers with a high score should be trusted as their track record has proven their accuracy.

We have created a leader board for the top domains and authors in the tech industry to help highlight those who are consistently doing great work.



Some websites seems to just post rumors, such as Digitimes, that are later found to be either partially or totally inaccurate, yet they are still popular. Why do you think this situation exists?

The interesting thing about Digitimes is that they have an awful track record, but they have also nailed some pretty big items. For example, Digitimes correctly predicted that Microsoft would build a tablet nearly a year before the product was announced.

Basically, Digitimes is not a reliable source of information but sometimes, they do get it right, which is why we can’t totally ignore them. 

What's the ultimate goal for Tracour in terms of monitoring the accuracy of tech news rumors?

Our goal is to becoming the ranking engine of the Internet. We are sort of like Klout but we cover domains and authors and use hard facts to draw our rankings. Imagine this situation where Enadget says the iPhone 6 will launch in June and The Verge says the iPhone 6 will launch in August. We can use our data and cut and slice it hundreds of ways to tell you who is likely more accurate and which source should be trusted. 

Do you see Tracour moving into other media stories such as politics or entertainment at some point?

Certainly, our custom built Honesty Engine can be used with any data that has a URL. Currently we are a small team building the service and are only focusing on tech rumors at the moment. But, our service can easily be applied to stock predictions, analyst’s forecasting, sports rumors or even tracking weatherman forecasts; if it has a URL, we can track it.

Finally, is there anything else you want to say about Tracour?

Tracour is trying to reduce the noise around rumors and increase the signal. Our database is growing quickly and we are already seeing third party websites use our data to enhance their posts. Our pipeline includes tons of features that will help consumers see insight into past rumors as well as upcoming products that will be launching in the future.

We have monetization plans in place and have already been approached by two outside firms who want to purchase feeds off of our data but we believe in offering a free service to the consumer. The goal is not to burdern the site with advertisements and offer the majority of our features for free.

You can do a lot of fun things with data, far more than I have talked about during this interview, and our ambitions are only outweighed by our to-do list. Keep checking up on us as we expand our services and features. 

View: Tracour.com

Report a problem with article
Previous Story

New Microsoft holiday stores in NYC and Chicago opening Oct. 26

Next Story

Microsoft training videos leak, tell you how to sell Windows 8

5 Comments

Commenting is disabled on this article.

I like this idea, but I worry that it might miss a larger mark: spin.

Stories that get information wrong need to be called out, but spin is not inherently 'incorrect' and won't be caught by engines like this. This might punish liars, but it ignores anyone who simply misrepresents the truth to push a point. I know spin isn't a big part of rumor reporting, but sites like these could give a good reputation to otherwise heavily biased sites.

For example, Fox News has a pretty darn good track record of calling elections before official results. I'll let you decide whether that makes them trustworthy.