/* Style Definitions */
mso-padding-alt:0in 5.4pt 0in 5.4pt;
font-family:”Times New Roman”;
mso-fareast-font-family:”Times New Roman”;
I love Yelp — it is an awesome site that has some of the
most comprehensive reviews of thousands of small businesses (I mainly use it
for finding good restaurants). I’m also
impressed by how the Yelp community has grown.
Building an explicit site like Yelp that relies on user
generated content is a really hard thing to do. Dozens of well-funded companies
tried and failed. Explicit sites face huge marketing challenges because they
need to get people to their site and start reviewing things before there is
much value for others to use the site – a hard chicken and egg problem to
Yelp has overcome these challenges by building in
hand-to-hand street marketing:
– Throwing parties for their
– Having marketing people get
stores to put up “People Love Us on Yelp” stickers
– Growing community by community –
so they are really strong in places like San Francisco
or New York but
they might be weak in another community
– Contacting businesses and getting
them to claim their profiles on Yelp
– Having great community managers
that encourage lots of reviews by focusing on game mechanics
How to challenge an
If you were going to build a Yelp competitor from the UGC
angel, you have two main options: be better than them at community marketing or
be better at inciting user reviews. But
it is going to be really hard to compete on either. Essentially, you’d be
fighting Yelp at the game they are best at so you’d have to be the best web
marketer to succeed.
You could also take a totally different direction and go for
a new medium (like mobile) and maybe we’ll see some companies do well here.
Another option would be to focus on building an implicit
version of Yelp. Unlike explicit sites that rely on user generated content,
implicit sites crawl the web and find already existing data and synthesize them
in one place. Instead of the huge marketing challenge that comes with building
an explicit site, implicit sites are really big technical challenges. That’s why great marketers should focus on
explicit challenges while great engineers should focus on implicit challenges.
If you think about how Yelp is organized, they have a series
of reviews on small business which generally have a common key, such as a phone
number and/or a postal address. So these
businesses are usually relatively easy to distinguish when crawling the web.
To build an implicit site about small businesses, the first
thing you might want to do is to crawl all the review sites, including sites
like Yelp, CitySearch, Menuism, New York
Magazine restaurants, Shamash
kosher reviews, VegGuide, and
more. You can then build collective profiles about the businesses.
You can also include government data about the business
including liquor licensing, lawsuits, Better Business Bureau, real estate
values, health inspections, and more.
Of course, the best site might be one that marries both the
explicit and the implicit.
Wait. Doesn’t Google already do this? Or if they don’t they could just turn a query into a business?
The other problem with Yelp is the ad revenue model: the business getting reviewed is paying for advertising on Yelp’s site. So of course there is always a problem with jerry-rigging (or should we say jury-rigging). Regardless of whether Yelp purposefully manipulated rankings, there is a strong incentive for low level Yelp sales scum to pursue unscrupulous strategies — like submitting their own reviews to punish a merchant (this explicitly adjusts the ratings without any “tech” manipulation). Why not just a pure, highly accurate rating system?
I like your diistinction: some businesses have lots of marketing challenges and some businesses have lots of engineering challenges. if you are startign a business, you should play into your strengths.
As a software engineer, I should start an implicit, not an explicit, business.