Blockfeed App Surfaces Hyper Local News

The death knell has been sounded for local journalism at regular intervals during the digital era. Craigslist ate classifieds and gobbled up the local newspaper’s business model in the process. Then social media served up a free stream of gossip from your social circle and interest groups, pulling readers’ attention away from the printed page. Safe to say, it’s been something of a perfect storm battering traditional local news media.

Yet there has never been so many sources of local news available, given the web has removed barriers to publishing, social media has got everyone accustomed to self-publishing (and self-publicizing), and smartphones are ubiquitous e-readers in everyone’s pockets. Local news is out there and people are interested in reading stuff, so distributing relevant content to the right people is one new challenge.

And that’s where Blockfeed comes in. This New York city-based startup is aggregating local news sources, from small blogs to established newspapers, geolocating relevant news stories to a hyper local location — such as a particular street or block — and then serving those stories to readers based on where they happen to be at the time they open the mobile app. Thanks to smartphone location-positioning tech, knowing a reader’s location is trivial.


Blockfeed users are redirected to the actual news source if they tap to read the story, with the app just acting as a local news source aggregation and distribution platform. In theory — and if it gains significant traction — the platform could help to level the playing field for local bloggers competing for attention with established media outlets by surfacing their content alongside more mainstream media new stories.

“The front page, which is our primary way of displaying information to the user, uses distance as the major relevancy factor. But it also uses social activity as part of how we prioritize things,” explains co-founder and CEO Phil Perkins.

“So, for example, if an article is getting shared a lot on social media it’s relevancy goes up. But one thing that’s cool that we do is we actually normalize that kind of velocity against the publication itself — so we enable even small blogs to theoretically compete against your major publications, like your New York Daily News or your New York Times.”

Currently the app is live in New York City only, after soft launching on iOS at the start of last month. Thus far it’s gained around 900 active users without any marketing. It’s launching on Android today, and stepping up the noise.

Perkins tells TechCrunch the ambition is to replicate the model in other major urban cities in the U.S. “This is our first major push to market the app publicly. Our retention numbers have been awesome, and our median sessions per day has been about 2.2 for the past month,” he says.

“We want dense urban areas, and we want a high-quality local journalism scene,” he adds, discussing where it may expand the app to next. “We are going to be cherry picking cities. It’s very likely we will continue a bit on the East Coast and then ultimately jump to San Francisco and LA. So on the East Coast potential candidates might be Boston, Washington DC, Philadelphia.

“When a given area is pretty much well served by one single publication and there aren’t many people engaged in publishing blogs and they’re not really engaged in their community from a journalistic perspective there’s really no purpose for Blockfeed.”


Blockfeed’s approach is a rethinking of how you slice and dice local news media. It replaces the role of human editors deciding which content goes on the front page with location technology plus a pretty simple algorithm that assesses how new a story is and looks at how many social media shares it’s gained to determine what content to dish up to the reader.

The startup is not purely tech driven. It is currently employing one human to cast an eye over the news stories it’s pulling in — Perkins refers to this process as “ingesting” — to make a final call on whether stories should be included (i.e. whether they are relevant as local news), and, crucially, to help verify the location of the story (such as by cross-checking with Foursquare’s location API or Google Places).

To be clear, the human is not acting as an editor in the sense of fact-checking these local news sources. Blockfeed is aiming to be a platform, not a publication in its own right, so it’s human employee is acting more in a curatorial than editorial capacity.

“There are several stages of transformation that occur to map [local news] content to content that’s suitable for the Blockfeed platform. There’s an attempt at categorizing it, so that’s using some machine learning, there’s an attempt to extract location, there’s some cases where these feeds actually contain location data — that’s extremely rare though, it’s like one percent of the time — so we’re using some named entity recognition, and we’re pulling places of significance from it,” says Perkins.

“A city like New York city which has a flow rate of, let’s say, 700 articles a day, one single editor is able to handle that entire feed. They’re pretty much extremely rapidly confirming that most of the fields have been populated properly, potentially geolocating it and approving the article… confirming that the data is correct and the location is correct.”

“At the end of the day we really depend on really high quality data. We ultimately want good location reads on every article that we ingest. So often there is a human being involved. But that human being is augmented with this technology that I’ve built,” he adds.

I can achieve complete coverage of New York City at this point with one single human curator.


Does this hybrid tech plus human eye approach scale? Yes argues Perkins. “When you look at even a medium-sized publication… they actually maintain quite a large staff of editors. I can achieve complete coverage of New York City at this point with one single human curator. So I believe I can start scaling out to multiple cities with just a handful of editors, and ultimately be in every major American city with what you might consider a normal size staff for a publication.”

The other question here is quality. Given Blockfeed is not verifying the accuracy of the news it’s pointing people to it can’t be compared with a newspaper — where the editorial staff is there to fact check and stand up stories.

“Ultimately we’re attributing all of the content to the sources, and we just direct you to the third party site to actually view the content,” admits Perkins. So it’s very much ‘reader beware’ when it comes to judging the veracity of the local news they are discovering via the platform.

Moreover, while locally motivated bloggers may be keen (or incentivized) to cover certain types of local news, say like a new trendy bar opening on their block, other less easy to report on/monetize community goings-on may well fall through the cracks of a citizen journalism, crowdsourced news model.

So while Blockfeed is an interesting idea for distributing local news to an engaged local community thirsty to hear about new stuff on their block — and as a way to support local bloggers by bringing them additional, engaged readers — it’s a stretch at this stage to say whether it might be able to fill some of the gaps left by the digital decimation of a robust, well-resourced local press engaged in the often tedious but democratically vital role of scrutinizing local politics.

Still, Blockfeed is a business first and foremost, rather than a social crusade to save local journalism. And the team sees big potential for its platform to generate rich local data that could be monetized via local contextual advertising or by selling as B2B intel to third parties wanting to move in on a particular local community.

“As a business model ultimately all of our users are coming to Blockfeed for their local news. We are aware of their location or whatever location they’re actually interested in viewing, and we’re aware of all this local content so if you think about that it’s actually a great platform for local, contextual advertising,” says Perkins.

“The data itself is actually really valuable. And there are potentially many different commercial clients. These aren’t just dry data points. These aren’t just simple crime statistics… These are actual news articles… and they’re chock full of information. They were written by human beings, they may contain links and all kinds of other information — so it’s a really rich data-set to be geolocated. We will be the top source for that kind of urban information.

“So let’s say I’m a real estate agent, for example. I can actually show you the real cultural history for the past two months of this block with this data. I’m not just giving you empty crime statistics. Or let’s say I’m McDonald’s and I’m looking for the location of my next chain restaurant or whatever I can actually see, for this given section of the city. What are the history of openings and closings of businesses so I can assess my risk? So I think ultimately that data-set that we’re building will potentially be very valuable.”