Backfence: Lessons Learned

There has been a lot of speculation about what went wrong at Backfence. To date, the company's investors and I have tried to stay out of the second-guessing in the blogosphere and the trade press, largely because there are private business matters involved that we've chosen not to discuss. Indeed, as with many early-stage companies, some of Backfence's problems were internal and self-inflicted, and actually had little or nothing to do with the many reasons wildly speculated about in industry blogs and in the trade press in recent days.

However, as Backfence's co-founder, I thought it would be helpful to discuss some of what we learned from Backfence--and why I'm still very optimistic that a similar model can and will succeed. As a pioneer in the user-generated, hyperlocal field, Backfence hopefully will pave the way for many other efforts to create locally focused online communities that ultimately will become profitable businesses.

Over the past couple of years, we’ve seen the appearance of a large number of different variations and models for creating and operating user-generated, citizens’ media or hyperlocal sites. (See Steve Outing’s excellent 2005 essay on the 11 Layers of Citizen Journalism—there surely are more by now—and the excellent recent overview of citizens’ media by Jan Schaffer and the University of Maryland J-Lab.)

Like Backfence, all of these nascent efforts are fascinating laboratories—and also like Backfence, none has yet proven to be a successful, sustainable long-term business model. So it’s difficult at this juncture to say what’s “right” and what’s “wrong.” But based on the Backfence experience, here are are a few things I believe are essential for success of a user-generated hyperlocal site:

* Engage the community. This may be the single most critical element. It's not about technology, it's not about journalism, it's not about whizbang Web 2.0 features. It's about bringing community members together to share what they know about what's going on around town. A top-down, "if you build it, they will come" strategy absolutely does not work, and that's not what Backfence did—otherwise, we'd have launched in a lot more than 13 communities in two years. Instead, Backfence employed a group of talented journalists and community representatives who sought out and interacted constantly with members of each of our communities to encourage them to participate. Over time, in our more mature communities, this really bore fruit. Could we have done more? Absolutely. It's essential to success. You have to get a critical mass of community participation and eyeballs coming to the site. Also essential: engaging existing local bloggers, Flickr members and other existing local voices. All of these must come together to make a successful hyperlocal site. You have to get the community involved. There’s no substitute for that.

* It's not journalism—it's a conversation. (Actually, it's whatever the community wants it to be.) The magic of hyperlocal sites, be they Backfence, other startups, Yahoo Groups or local blogs, is that they provide a forum for community members to share and discuss what's going on around town. The back-and-forth of a good online conversation can be as rich, deep and interesting—or more so—than traditional journalism. In fact, the role of journalists in this process is overrated—except maybe by journalists! The less involved site managers are, aside from lightly moderating the conversation, the better. Proponents of some other citizens’ media models argue that journalists are essential to hyperlocal sites as thought leaders and examples of professional reporting. But that adds considerable expense, and in our conversations with community members we learned that the intrusion of an all-knowing journalist would tend to stifle, not enhance, community conversation, by setting top-down agendas and crowding out community members. Let the audience decide what's important and choose its own leaders.

* Hyperlocal content is really mundane. We heard this criticism all the time. You bet it is—if you're an outsider looking in. To members of the community who actually live with these local issues, it's vitally important. It's precisely that mundane content, and the conversations around it, that brings life to hyperlocal sites. I find that when I look at supposedly thriving hyperlocal sites, they seem boring and pedestrian to me. Exception: WestportNow.com. Why? Because I grew up there and know the place and its players. It's that simple: It’s relevant to me. From outside, hyperlocal content all looks mundane. But it's information and conversation that's important to a specific local audience and flat-out unavailable anywhere else—the far end of the Long Tail. That's not mundane—it's competitive advantage.

* Trust the audience. We were asked all the time, mostly by nervous journalists, how we avoided having Backfence become a nasty free-for-all. There were many answers: We installed profanity filters, required registered membership to post or comment, asked members to use their real names and put "report misconduct" buttons on every post and comment. But most of all, we trusted the audience to do the right thing—and invariably it did. All of that is why we can boast that, aside from removing the usual offshore classifieds spam (English bulldogs or capuchin monkeys, anyone?), we very, very rarely had to police Backfence by deleting content. It happened just a handful of times over two-plus years. The audience took responsibility for what went on at each local Backfence site and debated local issues in a civil manner because it was about THEIR community. And as in their physical community, they were proud of it and took care of it.

* Focus on strong, well-defined communities. We chose the communities in which Backfence launched based on demographics, population density, local governance, commercial viability and competition, among other factors. But as much as any of these, we chose them because they had a strong, well-focused sense of place and community pride—I live here, I don't live over there. Moreover, they were well-defined geographically. Beyond a certain size, communities lose their focus--there are too many different governmental bodies, local organizations, schools and people to get a clear grip on what it means to be a community. It's possible to argue, in fact, that a hyperlocal site ideally should operate at the neighborhood level—that even a town is too big. Trying to create a hyperlocal site that covers a large area increases the potential population and spreads the focal points of interest too broadly. You only care about the high school your kid goes to; the one across town might as well be 3,000 miles away. It's all about focus: local, local, local.

* Leverage social networking. The rise of MySpace, Flickr, YouTube and the commercial version of Facebook—virtually all of which have happened since Backfence launched more than two years ago—demonstrates the power of social media. Local communities are social beehives anyway. Why not take advantage of existing local connections and the virality and marketing reach of social tools such as member profiles, "friending" tools, widgets and the ability of members to exchange messages with each other? This was an element we unfortunately were unable to get off the drawing board at Backfence, because of business issues and other priorities.

* There is most certainly a robust hyperlocal advertising business. Indeed, local advertisers are eager for new online advertising vehicles. I've seen it suggested repeatedly that Backfence failed because it couldn't sell advertising to local merchants. Not true. In fact, we sold ads to more than 400 advertisers, more than any other similarly sized hyperlocal effort that I'm aware of. It was clear that we had staked out an affordable and lucrative corner of the local ad market. Ads in local newspapers—even community weeklies—are too expensive for many small local businesses. Alternatives like the Yellow Pages, Val-Pak-style coupon flyers and local radio are similarly pricey. And most small businesses don't know from AdSense. That presents a ripe target for a talented, hard-working ad sales team concentrating on offering low-cost ads to local businesses who want to reach members of their communities through hyperlocal sites. It's a rich, untapped marketplace.

* Keep costs down. To become a successful business, hyperlocal citizens’ media efforts have to be lightweight and lean. Many hyperlocal sites today are one-off labors of love, with little or no hope of becoming profitable. Nothing wrong with that, unless you want to make a living at it. To make a successful business, a hyperlocal operation has to be run as much more than a hobby. More than likely, the costs (and revenue) need to be spread among many sites to create a successful business model. The Backfence formula averaged about one staffer per community site, and in retrospect, that probably was too rich. Oh, and you don’t necessarily have to pay community members to contribute to and participate on the sites. That’s not their primary motivation. They want to be seen among their neighbors as sources of local knowledge and opinion, and that’s more valuable to them than being paid. Indeed, we never had a single member ask for compensation—and when we offered coffee cards in exchange for posts, Backfence users actually rebelled. They thought we were making the site too commercial, even though ads didn’t seem to bother them at all—because local ads are local information.

* Partner with a media company or some other distribution source. Because of the critical need to market to and engage the community, it's better to piggyback on a print or broadcast partner’s existing community relationships and marketing power. It's very, very difficult to start from scratch in a community and get to critical mass without help. For a variety of reasons that made sense at the time, Backfence chose not to go the media partner route. But as newspapers and broadcasters have become more savvy in the past few months about their need for hyperlocal efforts, it makes more sense for hyperlocal entrepreneurs to hook up with media partners—or for media companies to start their own, dedicated hyperlocal efforts.

* Hyperlocal works. You need patience and hard work to embed yourself in a community and become a vital cog in the life of that community. But when a community comes together, it’s striking. We saw it happening in Backfence's more mature communities after a couple of years—in fact, even after cutbacks earlier this year left us with no day-to-day community outreach, site traffic and posting activity continued unabated, and even rose. Were they at the levels we wanted? Of course not (they never are). But that proves that once the community gets involved, a successful community site can almost run itself. I said "almost"!

* Hyperlocal is really hard. Don't kid yourself. You don't just open the doors and hit critical mass. We knew that from the jump. It takes a lot of work to build a community. (Look carefully at most hyperlocal sites and see just how much posting is really being done, especially by members of the community as opposed to be the sites’ operators.) Anybody who's run a hyperlocal site will tell you that it takes a couple of years just to get to a point where you've truly got a vibrant online community. It takes even longer to turn that into a viable business. Unfortunately, for a variety of reasons, Backfence was unable to sustain itself long enough to reach that point. In an entrepreneurial startup, there's a blend between capital raised and a point of self- sustainability. Typically, that takes many years. Even though it raised—and spent—$3 million, Backfence was unable to make that leap. Ultimately, the investors and I determined that the company's fundamental internal problems made it impossible for us to reach that transition point, which is what led to our decision to shut down. It happens; most start-ups fail, unfortunately.

And yes, Backfence was one of those startups. I'd like to thank the talented, hard-working employees who spent so much time trying to make Backfence a success—reaching out to the community, interacting with members, selling ads, keeping the site running and adding new features. They were the backbone of the company, and unfortunately lost their jobs earlier this year because of issues beyond their control. I'd also like to thank our venture-capital investors at SAS Investors and Omidyar Network, who took a big risk on an unproven concept and then took a large financial loss when we were unable to successfully execute on that concept. The investors have been consistently helpful and supportive and even now are continuing to explore ways to possibly revive the company in some form, even if there's no remaining financial benefit to them.

Most importantly, we all believe that the core Backfence concept—user-generated hyperlocal citizens’ media—is sound. Someday soon, somebody will make it work and turn it into a successful business. If there's anything I've learned from Backfence, it's that the power and potential of local communities still is waiting to be tapped. In the end, Backfence was a well-intended experiment that unfortunately was not successful in the form it took and at the time it was launched. I hope many other hyperlocal citizens’ media efforts can learn from what we did and rise in our wake.

We'll see you around the neighborhood.

Comments

[Jay Rosen]

Thanks very much, Mark. Very informative. I give you guys a lot of credit for going out there and doing it, taking the plunge, making the mistakes, adjusting the model and of course dealing with doubters who can't be bothered to spend 30 minutes finding out what you are actually doing.

I found especially informative the part about partnering with big media sites.

One thing you did not go into here is: what kind of content did you find citizens successfully contributing that other citizens wanted, and how did the actual content differ from projected content, meaning the ideas you had at the start?

[Mark Potts]

Thanks, Jay. A couple quick comments on content:
* We really didn't have any predetermined expectation of what the content would be, because we knew the communities would determine that for themselves, based on their specific local needs. In keeping with that, it varied from community to community—McLean was heavy on youth sports; Reston had a lot of discussions about local politics. These reflected the interests in those towns.
* Some of our best content was long, back-and-forth discussions about local issues, where the meat was in the discussions (much as often happens on your blog). We also had a number of stories from the community that were picked up a few days later by local media.
* We had a lot of posting of announcements and press releases by local organizations, which generally didn't foment much discussion, but often received a lot of page views as the community checked to see what was going on around town.
* Least expected: The success of our user-generated event calendar, in all communities. It was deep, it was comprehensive, and it was entirely user-contributed.

[yellowbot]

Thanks for the information, Mark.

The content information in your comment was also very interesting.

Best of luck!

[getvendors.com]

Hello Mark,
Thanks for posting your observations. Very useful reference for anyone engaging in local space. One thing though, don't you think folks building hyperlocal sites should accept the fact that their content will be mundane and think of building "community business" around that. What I mean is against the news of Paris Hilton or Facebook zillion dollar potential IPO, story of a new shop in neighborhood is always going to be mundane stuff..yes, even if you are going to buy bread from that shop. It just does not have that urgency ..that pull. But Paris Hilton/Facebook news earns money on CPM basis as against the local shop will be willing to pay for $3000 in ads if you are getting him/her exposure in front of local crowd. I know nobody wants to say this but for hyperlocal to succeed first innovation and differentiation has to be in terms of business model. If you see CPM based ads or google ads on a hyperlocal site then you can be sure that it won't succeed. You can not have business model of volume play for hyperlocal sites. Whoever does that successfully can engage whole heatedly in creating niche local presence that does not have glamour or scale of larger communities but still has very vibrant and useful presence on the web (with its abundant resources). With right business model, these communities would have got metamorphed into success stories.
-Ashish

[Phil Wolff]

So what plans have you made for your communities? Will they be closed? Copies saved at Archive.org? Put in a not-for-profit trust owned by the community members? Sold to a bigger outfit promising to keep the motors running?

I've been on small sites that just walk away and that hurts. I wonder what happens with a bigger site with more loyal, committed community.

[Mark Potts]

Ashish: There's no question that the hyperlocal ad sales model is very different than the national model. On a local level, as you say, AdSense and CPM are both problematic. The sales proposition to local merchants has to be based on helping them understand that they can target and reach their best local prospects.

Phil: The disposition of the Backfence sites and contents is uncertain. The company's investors are entertaining a number of possibilities that may or may not include continuation of the sites and content in their curret form. It probably will be up to an acquirer to decide how the archives are handled.

[John White]

Mark:

I launched a hyper-local media service in Winnipeg last year and had to shut down do to cash flow issues. We were staffed by professional journalists, but also had a community space for users to post.

Registered users could comment on any story.

We shot video, audio, photos and had blogs and regular print-style stories.

We were able to cover breaking news ahead of the big players in the marketplace, and had positive reaction from the majority of users.

Cracking the cash flow nut in finding the right balance of staff is the key...

[earlpearl]

Thanks very much for writing about your experiences and lessons learned.

As a small business operator in the DC region I inquired about advertising at backfence and ultimately decided against it based on cost benefits of costs versus traffic. My business has high rankings at search engines for appropriate phrases and generates the majority of sales off of internet advertising (after decades of YP and newspaper classified and other media) In fact earlier today I spoke with a web developer and someone who follows search engine optimization and as he perused clients stats he found most traffic off the web from most clients was coming directly off of searches.

Advertising media on the web is a tough nut let alone creating vital, large, appropriate communities. I have been a participant in interest group forums that have experienced varying levels of interaction and success.

I suppose it is a learning process.

I'd simply like to thank you for your efforts and comments. Much to learn.

Dave

[steven johnson]

hey mark, what a thoughtful and gracious post. I really appreciate the parting thoughts, and look forward to seeing your next move....

Steven (from outside.in)

[Karen Bridgman]

Mark,
Way down here in New Zealand we have been following your site closely, because we are in the process of launching New Zealands first ever hyper local citizen journalism network. Of all the sites in the US, Backfence was the one we thought most likely to succeed and I am very sorry it hasn't worked out at this stage. The revenue model at www.iworldpeople.co.nz includes web pages, local classifieds and site ads, all user generated, on top of publishing of news, events, opinions and photos by businesses, organisations and citizens. We will certainly take into account the lessons learnt at Backfence as we develop our sites and wish you all the best.

[Rocky Agrawal]

Thanks for the great summary, Mark. Sorry, things didn't work out with Backfence.

I'm a big believer in hyperlocal, but am beginning to wonder if the uber-hyperlocal portal is the way to go.

Will people instead stitch together services on their own to meet their needs- e.g. get community events from Upcoming, restaurant reviews from Yelp and neighborhood info from their HOA Website?

[Angela Button]

Thank you for your thoughts, Mark. How to make hyperlocal sites successful is going to be one amazing puzzle to play with over the next few years. Will be in touch again soon!

[Kevin Roundtree]

Mark,

Thank you for your article. Kaesu, a company I develop for, has created a model that hits on just about every item you discuss in your article. We leverage relationships with local newspapers (mostly weekly) and newspaper associations to help reach local communities. We have combined free and paid (place in print) classifieds, a display ad archiving and online distribution system (similar to AdSense distribution), a local business directory and tools for businesses to add content. We also brand sites for the newspapers so they can continue to own their markets ( i.e. mtoliveweekly.coolerads.com - A paper in Budd Lake, NJ).

[trademark lawyers]

But is local advertising enough these days?

[Focusdigitaluk]

Just wanted to say thanks for probably the most interesting hyperlocal article i've read in recent memory despite the fact its over two years old! I'm looking into doing something similar in the UK so your insights are much appreciated.

Martin Reeves
www.alderleyedge.com