Monthly Archives: February 2009

Rants and Raves

Why I don’t trust Yelp

10 Comments
Image representing Yelp as depicted in CrunchBase
Image via CrunchBase

Earlier this week, Kathleen Richards of the East Bay Express published an article titled: Yelp and the Business of Extortion 2.0. The main point of the rather scathing article is that Yelp offers to hide negative customer reviews of businesses on its website, but for a price.

Jeremy Stoppelman, the co-founder and CEO of Yelp, posted a rebuttal on the Yelp Official Blog: Kathleen Richards – East Bay Express and several followup posts as well (see the Yelp Blog for more).

Now normally, such blog banter wouldn’t hit my radar and even if it did, I would usually ignore it. However, the Yelp story hit a nerve and prompted me to post about why I am a little dubious of Yelp.

In December 2006, I had a bad experience at a local business (Tea Time at 542 Ramona St, Palo Alto, CA 94301). I posted a review on Yelp. Here is the full text of the review:

Photo of Manu K.Manu K.

Palo Alto, CA

1 star rating

12/19/2006

Yikes! I’m so surprised to see the other reviews for this place on Yelp. Here’s why:

Three of us were meeting for a business meeting at Coupa Cafe at 10:30 AM. Coupa happened to be very crowded and so we decided, hey, why not go next door to Tea Time since it was empty. And I mean empty, there were no customers there at all.

We go into Tea Time and find a place to sit. Turns out all the power outlets happened to be along the wall where there was only seating for two. There being three of us, we pulled up a extra table and chair, ordered a pot of tea and sat down do chat and do a demo.

A couple of minutes later, the guy running the store (apparently not the owner) comes by and tells us they “we’re not allowed to change their seating configuration because they have all these other customers that they need to accommodate”. May I remind you that we were the ONLY customers in the store at the time 🙂

We told him we would put things back in their position when we leave, but he insisted that he had to put the chairs back in their position right away. So we told him to give us our money back first and merrily went on our way to one of the several other great places Palo Alto has to offer.

What amazes me is how someone can be so silly as to piss off their only customers in the store and that too to make sure they can keep their chairs in the right position!

So we obviously have the worst  possible opinion of this place and I bet that their lack of attention to customer needs will render them out of business soon. Afterall, this is the Valley — lots of people go to cafe’s to plug in, get on the net and talk shop, the beverages/deserts are often the perks which come as side effect. So here’s wishing all the best to the guy/gal who owns this place — they’ve still got a lot of learning to do about how to run a business.

I posted the review and then also emailed a copy to my cohorts who were with me. A while later, when I visited Yelp, I figured I would check to see if my review was still there. I was shocked to find that when I looked through all the reviews for Tea Time, my review, with its 1 star rating was no where to be found.

A little further digging today, right before writing this post, showed that the negative review I wrote only shows up when I am logged in to my Yelp account. When I log in to my Yelp account, Tea Time has 47 reviews. However, when I visit the site in another browser, without logging in to my account, it only shows 46 reviews.

Now I’m not an avid “yelper”. This was the one time I felt that my experience with a local business was bad enough to warrant taking the time to write a review on Yelp about it. The fact that the review doesn’t show up for anyone except for me, seems to suggest something that needs further review. Jeremy has written about their black-box algorithm, but I can see how and why an algorithm should be used to compute the final ranking for a business. However, removing a full review, which is the raw data, seems inappropriate to me.

I would be perfectly okay with allowing the business to rebut the review given by a particular user and and if a business takes the time to do so, perhaps it should reduce the weight assigned to the review accordingly. However, I think it is inappropriate to not show or remove a review.

I am perfectly willing to give Yelp the benefit of the doubt here. They have a great site and I use it often to check reviews (but don’t write many). But my faith and trust in Yelp’s site would be considerably higher if I the one review I took the time to post, would actually show up on their site. If they have a rational explanation, I’m all ears and would welcome a response in the comments.

Post to Twitter

Rants and Raves, Thought -- caught in the act!

Information fragmentation in the world of Web 2.0

10 Comments

Web 2.0. Yes, that wonderful, wonderful world of Web 2.0. Where information and technology will solve all of the worlds problems, make it possible for us to communicate instantly with anyone, where all human knowledge will be instantly searchable in nanoseconds, and yes, it will also eradicate poverty, achieve universal tolerance, global literacy and intergalactic peace.

All hyperbole aside, Web 2.0 has been a wonderful thing. Today, we can search human knowledge using Google, we can communicate instantly with friends and family using Facebook, we can publish our thoughts and share our opinions (like I’m doing here) with the world using WordPress and we can publish/receive news (like a plane landing in the Hudson river) while it happens on Twitter, all among a myriad of other things. For the record, I love Web 2.0 (as a user) for its ease of use and for the so many immensely useful services it provides (though I must admit I’m not a fan of the business models that often accompany Web 2.0 companies, especially advertising related models).

Image from ConnectionCafe

But, “Houston, we have a problem.” The problem is what I would call “information fragmentation” in the world of Web 2.0. Yes, each Web 2.0/social media site in isolation may be very easy to use, but to try and get all of them to cooperate and more importantly keep track of all our information is becoming a complete nightmare. Here is just a short list of just some of my information assets:

Gosh, I can’t even keep track of them in a list, and I certainly don’t want to link to all my accounts on various services for privacy concerns. Just recently, I discovered that I had a MySpace page! I didn’t even know I had a MySpace page, and in fact, I used to take pride in saying that I didn’t have one (they built their network by being spammers in my opinion). Until I discovered that I did — probably one I set up years and years ago, and never thought about it twice after that. (I have since promptly deleted the MySpace page, so that I can stick to my assertion that I do not have a MySpace page!)

Our bits are spread out all over the web. In fact, it’s probably more accurate to say that bits of us are spread out all over the web. Subtle difference. The deluge in the number of services out there has now resulted in new services (like FriendFeed and Ping.fm) which try to help you take control of all your media and interactions. However, aggregators only work in theory. In theory, the pitch for an aggregator is that “We will be the one stop shop for X.” The problem is that there are a dozen one stop shops. And so you ultimately end up building an aggregator to aggregate the aggregators! The travel industry is a great example of this with the airline sites, then sites like Expedia, Travelocity and Orbitz. Then sites like TripAdvisor, Kayak, Vayama. I’m just waiting for the next level up (Update: Didn’t have to wait too long, as soon as I finished this post and went back to check on my feeds, I found this on TechCrunch: Travelzoo’s Fly.com Launches Yet Another Travel Search Aggregator). It’s the Madoff scheme (Ponzi is becoming too old for people to know what it means any more!) of aggregators.

Likewise in social media, everyone is trying to aggregate everyone else. Till recently, I had my Twitter tweets being cross-posted to my Facebook status. I recently severed that connection. Now I have FriendFeed aggregating my tweets, my Google Reader shares, my blog posts all together. Oh, and I also have a FriendFeed tab in my Facebook profile. Everything is going in circles, I guess that must be the true indication that I have embraced Web 2.0 and social media, and more so, that I feel caught up in its embrace. An embrace of so many tentacles that it’s like being caught in, oh, oh, wait for it… a Web!

Lets make this more concrete with some examples:

  1. Blog comments: When you post to a blog, you typically have comments on the blog. But then you also have trackbacks and pingbacks. And now you also have tweetbacks. The conversation has been splintered. I can get comments on the post, I can get tweets back in response to the post, I can get an email, I can get a phone call, a direct message on twiter, a like on FriendFeed or now even on Facebook, a comment on Facebook. It’s just all over the place. Fred Wilson has argued in his posts that (paraphrasing) comments should be treated as a first class citizen — as a true part of the conversation that is ocurring. But, comments are now being splintered all over — on Twitter, on FriendFeed, on Facebook, on blogs, in emails and in direct messages. Capturing that conversation has become and continues to become more of a challenge. There have been some good steps like Disqus integrating FriendFeed comments, but that is only the beginning.
  2. Status messages: Status messages are everywhere. And the buzz on the web is that status is the hot thing, popularized by Twitter and Facebook’s “What are you doing?” question. (Fred Wilson says that “Status is the ultimate social gesture”) But there is also GoogleTalk status. When I update my GoogleTalk status it is seen by people in my contact list who are also on GoogleTalk. But there is no (simple) way of connecting that to my Facebook status (emphasis on simple, though Xoopit recently announced GMail and Facebook integration through a plugin), to my Plaxo status, to my LinkedIn status. Ping.fm provides some hope of being able to do this. And I’m sure if I took the time to figure out the map of where all I want my status messages propagated (and avoid any circular references) it may very well do the trick. But, all of these services are supposed to be mass-market services. It shouldn’t require this much effort and thinking to make it all work right. While on one hand I am pleased by the diversity of options, on the other hand I lament that there are simply too many options.
  3. Pictures: I made a conscious choice to not post pictures on Facebook. Even though Facebook is intended to be for friends, there is a huge amount of intermingling of family, friends, teachers, co-workers, professionals and business contacts on Facebook. Yes, they provide a way to keep all these lists separate, but it just takes to much work. So I have my pictures on Picasa, in private albums that I share selectively. But then I also have a Flickr account and I also have a TwitPic account for posting images to Twitter.

I think the point is clear by now and so I won’t keep beating a dead horse. Web 2.0 and Social Media are wonderful and great, but at the same time they provide us with a plethora of options accompanied by a lack of easy interoperability (the kind that my mom could figure out). Information is being produced and created in unprecedented ways and at an unprecedented rate. It is being shared in unprecedented ways at an unprecedented rate. But, it is being fragmented in unprecedented ways. This is problem and an opportunity (as always). However, this is one case where I’m skeptical if just yet another technology (Google!?) will be able to help vacuum together all our digital bits from all over the Web.

Post to Twitter

Thought -- caught in the act!

The decreasing importance of longhand

6 Comments

Last week, I was thinking about how sometimes change happens and we don’t really notice it. It’s the old story of how to boil a frog. One such change that I think is happening around us is in the very foundation of basic education and in the three R’s: Reading, Writing and Arithmetic.

The change in arithmetic started happening when we started giving kids calculators. When I went to school, calculators were banned. You had to learn how to do math in your head, at least basic math. But I won’t dwell on the poor state of basic math skills since that’s not the core point of my post.

The bigger realization has been in how reading and writing are changing. When was the last time you picked up a pen/pencil and wrote a page full? How about a paragraph? Or even a full sentence? The pen, though metaphorically mightier than the sword, is becoming useless for anything more than signing our name on the credit card receipt, on a check or on a legal document. Yes, we may hunt for one when we need to scribble down a reminder or a quick thought, but whenever it comes to writing anything meaningful or of substance, our tool of choice has become the keyboard.

What does this mean for education and for the children who are learning how to read and write today? Do they really need to learn how to hold a pen or how to write longhand? The importance of longhand has diminished to such an extent that I worry about its very existence in the near future. It is more important to know how to type today than it is to hold a pen. In fact, I bet that very soon when someone needs to scribble something, they won’t be reaching for pen and paper, but reaching for their phone or whatever device it is that is almost surgically attached to their hand. (BTW, in case you haven’t tried it yet, I would strongly recommend trying out ShapeWriter on the iPhone. Here is the iTunes store link for ShapeWriter Free and ShapeWriter Pro.)

Writing longhand is a skill that I compare to boiling a frog — with every passing day, with every blog post and with every tweet, we’re boiling the frog that is writing with an instrument that still vaguely resembles a quill.

Post to Twitter

Thought -- caught in the act!

Running at periscope depth

For the past couple of weeks, for most of 2009 in fact, I’ve felt that I am running at periscope depth, meaning I feel that I am able to learn a little bit about a topic, but haven’t had the time to go deep. This is quite the opposite of the training you receive by being in a PhD program, where you are required to go deep and become the expert in a particular topic.

Taken during an Anti-submarine warfare (ASW) e...
Image via Wikipedia

I use the submarine analogy because it best describes what I mean. If you are on the surface and literally skimming topics, then you know very little and can easily be overthrown or thrown out of control by a wave or the wind. By contrast, if you are under the water, you have a little bit more control and are less susceptible to the wind and the waves, but you still lack the depth of knowledge. While I was at Stanford, I felt that I was able to stay at periscope depth, but still be able to make deep dives from time to time. It was a requirement of the role and the task at hand.

In my new role of helping startups, I find I spend a lot of time at periscope depth, but I haven’t been able to make as many deep dives as I would like to into various topics — sometimes topics that would be of benefit to the startups I’m working with and sometimes those for just my own edification.

A big part of the problem has been the glut of information created by social media (mostly useless information, but every so often there are a few gems). Keeping up with my feeds, trying to get a healthy dose of the twitter stream, and most of all managing the beast that is email takes up a large chunk of my week. The weekends have really become catch up days for things left over from the week. Fortunately, the information traffic on weekends decreases just enough to make it possible to write a blog post for instance!

Has the increase in social media activities such as Twitter, Facebook, blogging etc made all of us incapable of making deep dives? Most of the time the web is an “echo chamber,” as it has appropriately been described, for the self-absorbed. Innovation happens not only be being aware of the acitivity around you, but by being able to get a deep understanding of a particular topic. When are entrepreneurs working on building stuff, when most of the time they’re keeping busy just trying to keep up? These are just some of the questions I am pondering over, while I try to correct my own observation of being at periscope depth.

I hope to make some proactive changes to help correct this and carve out some time to still do the deep dives that I believe are essential for being fully informed and not just partially informed.

Post to Twitter

Rants and Raves, Thought -- caught in the act!, Twitter

Twitter, can you be my Zephyr?

1 Comment

While I was at Carnegie Mellon, we used the Andrew system as our primary computing resource. Andrew was so ahead of its time in so many ways that till today I still long for some of the features that it offered back then. The coolest part of it was that it all worked together. I still remember using twm, or motif along with Tardis as my console, and mucking around with my X11 preferences to setup my environment just right. But the thing I miss the most about Andrew is Zephyr.

Zephyr was the first real “instant messaging and presence” system in my opinion. We used Zephyr long before there was ICQ, or AIM, or Yahoo Messenger (or Yahoo Pager as it was first known), or GoogleTalk, or Skype. But Zephyr was the coolest thing since sliced bread. Not only did it allow you to maintain a list of friends whose presence you could monitor (mk5a has logged in to orion.weh.cmu.edu for instance) but it also allowed all kinds of cool functionality that IM never really caught up with.

Let me try to describe the coolness of Zephyr from memory. If there are any MIT/CMU alums out there who remember it, I would welcome their input to correct me wherever my memory may be failing me.

Lets start with the UI. Zephyr offered both a command line interface as well as a rich UI built around it. The command line interface was brilliant for being able to send messages even when you were telnet’d in over a dialup (shudder!). The graphical UI was cool — It not only showed a list of all your “friends,” but also showed which machine they were on. Since the machine names at CMU were logical and geo-coded by the name of the cluster they were in, you had a pretty good idea of where the person was on campus. For example, weh.cmu.edu meant Wean Hall — the home of all things geeky.

When a user logged it, their name got added to the list of online users with a little + sign next to them. The + sign would fade away. When they logged out, it showed a little – sign next to their name and then removed the name from the list. You could of course also have popup notifications about when a user logged in.

The standard UI for Zephyr used message pop-ups which appeared on the top left corner of the screen. The messages would pup up in individual windows (like the “toasts” do in Windows). The user could determine if they wanted the messages to automatically close after a predetermined time interval or stay stacked until you clicked on each message. The entire message window was a giant button — so clicking or hovering over it was not a problem.

Now, I mentioned “friends” above. but like Twitter and FriendFeed today, Zephyr also allowed me to watch people who I may not know, i.e. the “follow” vs. “friend” functionality. This allowed the discovery of lots of smart (and sometime even cute ones!) people — or as they say for Twitter/Facebook now: “Facebook is for people you know, Twitter is for people you want to know better.”

Zephyr had all kinds of utilities including one called zlocate, which alowed you to find out where someone may be logged in. zmap allowed you to produce a ASCII map of a particular cluster – you could tell exactly in what corner of the room someone was sitting in by using zmap. The wall utility allowed you to blast a message out to everyone within a particular cluster (or with a wider distribution if you had the powers). It also had a ultra fun zbomb utility which allowed you to flood a user with so many zephrys that it would take them a while to close out all the windows! 🙂

But I haven’t even gotten to the coolest feature yet. The Zephyr instances. These were vitual channels that one could subscribe to to become part of an opt-in community on any topic. The best examples are the food instance and the help instance. If you subscribed to the food instance, anytime there was free food anywhere, it would be broadcast on the food instance. (At Stanford we have the Gates food maling list, but that’s not as instantaneous as the Zephyr food instance! And as a student, when it comes to free food, you want to get that message quickly, before all the food is gone!)

The Zephyr help instance is where the magic happened. If you had a question, any question all you had to do was “zwrite -i help What’s the syntax for a regular expression that finds email addresses?” or better yet “zwrite -i help What’s the phone number for Dominos” and within seconds someone out there on Andrew-land would respond. We had “wisdom of the crowds” before it was called wisdom of the crowds. Zephyr was your oracle. It knew it all. You could ask (almost) anything and get an answer. In fact if you don’t know it already, the answer to “What is the meaning of life?” is 42. I learnt that on Zephyr.

The wisdom and instant gratification of Zephyr instances surpassed anything I had ever seen. It truly worked like magic.

Today, over 16 years later (geesh, I’m getting old) we finally have Twitter. Twitter is getting closer to what we had with Zephyr. It’s still not there. but it has the makings of it. By connecting my Twitter account with my Facebook status, I’ve finally been able to recreate the “oracle” I want. For instance, I was having trouble with my wireless router. A quick post to Twitter about which new router should I buy, resulted in a series of responses on Facebook with recommendations for which router to buy. Within 10 or 20 minutes I knew what my choice would be (in case you’re wondering it’s the Linksys WRT54GL — note the L, not the newest models, but the L). Similarly when I was trying to get some data on the paid click rates for search engines, I as able to ask Twitter (though I only got responses on Facebook!)

David Pogue (@pogue), the well known tech columnist for the NY Times conducted a similar experiment recently. Twitter really has the potential to become the oracle that taps into the wisdom of the crowds. However, Twitter is not perfect. It needs LOTS of improvement. I have lots of ideas about it too (UI improvements, threading, search, media, channels, discovery, ranking, visualizations, etc.) but that’s a topic for a whole other post.

But so far, in the absence of being able to return to Andrew and Zephyr, I’ll end with…

“Twitter, can you be my Zephyr?” (I hope so!)

P.S. In case you’re wondering, I’m @manukumar on Twitter.

Post to Twitter