What Yelp Needs To Do In Order To Now Scale Successfully

For the past year I have been amazingly bullish on Yelp (YELP), both the company and the stock. I’ve held this view for several reasons, but mostly due to the untapped potential of what Yelp has the opportunity to become, and what I felt would be an awakening of the market to that potential. Remember, the market does not integrate new information as it goes, it’s often a situation where the market doesn’t care until it cares. This is the inefficiency that can be taken advantage of.

Yelp’s run from $20 to $60 a share is evidence that this has taken place, while YELP still doesn’t turn a Non GAAP profit on a quarterly basis. Their revenue growth rate continues to hover in the high 60% range, neither accelerating or decelerating.

In my opinion, a large portion of the run has come from changing opinions regarding Yelp’s ability to continue to grow at that rate versus previous sentiment that they were already decelerating to a large extent and their ability to monetize was not that large compared to their social media peers at Facebook (FB) and LinkedIn (LNKD). As you can see in the chart below, they continue to crush both Wall Street and Estimize consensus estimates for revenue quarter after quarter:

The other monkey off their back seems to be the changing sentiment around how much of a threat Google (GOOG) is to them in restaurant search. Google bought Zagat (horrible purchase), but hasn’t really done much with it. The real threat was from the Yelp like recommendations right in the search page, specifically on mobile, but it seems that they’ve really failed to gain much traction with it, or at least Yelp isn’t seeing any harm from it.

I believe deeply in the theory of unbundling, meaning as an industry matures each piece of it gets broken out into its own service, its own business. What is taking place over at Craigslist is the perfect example. I think the industry has matured enough that if the average person wants to search for a restaurant they go to Yelp, they won’t just settle for Google, just as if you are searching for a place to crash in another city, you go to Airbnb, not Craigslist short term sublets.

But Yelp has a problem. As it acquires more and more content from its users--very valuable content--it is running into relevancy issues. The signal to noise ratio is getting bad, and it’s hard to trust what I see there anymore given the rampant self promotion that takes place. Yelp is trying to fight this through legal battles, completely the wrong way to do it.

The right way to scale is to use data science in order to provide relevancy and trust. This can be accomplished through using a system of weightings and measurements that are not difficult to build. I implore Yelp to move in this direction before people abandon the platform due to too much noise. Below are a few ways that you can increase the signal.

1. The more reviews you’ve posted the higher influence you should have overall in the rating of the restaurant
2. Your reviews within a specific category (Brazilian steak houses) should count more if you have reviewed others in that category
3. Reviews within your specific neighborhood where you have left a lot of reviews should count more
4. Underweight people who are persistently negative or persistently positive, there is a huge selection bias for what people review, you want to overweight people who review things across the spectrum of great to horrible, not just when it’s at both ends
5. There may be a discrepancy between reviews from different age groups for the same restaurant, this should be highly relevant to a recommendations system.

1. If I have reviewed a specific Brazilian steak house positively, find other restaurants that steak enthusiasts who have reviewed the same restaurant as I did and give me those recommendations.
2. Is there a place in my neighborhood that I haven’t reviewed yet that everyone is raving about (trending algo?), give it to me.