human search Archive

ChaCha president and co-founder Brad Bostic appeared on Fox Business today to discuss the human search engine. Check out the video below for an example query on why Brad believes ChaCha is better than computer-driven search. Sounded similar to when Mahalo CEO Calacanis always compares Paris Hotels search on Mahalo to Google.

Here’s what Brad discussed:

• ChaCha’s power is in the mobile service
• He compared why people help on ChaCha to why people help on Wikipedia – it’s because people are passionate about the topic(s) 
• ChaCha is an on-demand assistant for anyone with a mobile phone
• ChaCha has raised $16 million in private equity funding along with a $2 million grant from the State of Indiana

Last week we wrote about the new payment plan being instituted by Mahalo. Today Mahalo has revealed the pages that are currently paying the highest, and the price for every page they are looking to create. Here is how the Community Manager explains the pricing change, "We look at the demand for the search term, the size of the serp, and the amount of research (time) one might need to spend to do it justice. So, the lower priced SeRPs *take less time to do.* This is a critical point — it means that stubs (very short SeRPs with a handful of links) are now open for the Greenhouse to create."

Stubs are getting in the $1 per pricing just for quick Google rankings for Mahalo. Looks like the average is about $5-7/page. There are 1400 pages listed in the system to be created.

Here are the highest priced pages – those at $12-15/page (many of the "location hotels" are paying $12/page):

• Seville Restaurants (along with nearly every other big city in the world)
• Cheap Hotels Marrakesh
• Cheap Hotels Kobe
• Cheap Hotels XYZ City (I count about 40 other cities)
• Cheap Hotels Auckland
• Dynasty Warriors 6
• IFC Entertainment
• Sony Pictures Classics
• Warner Independent Pictures
• Picturehouse
• COX Inhibitors
• Allegra
• PGA Tour
• Mixed Martial Arts
• The Art of Power

For the $1/page, I note that EVERY car make and model is listed. Each option package also has a separate page (i.e. Camry XE has a different page than Camry SE), most major universities, every country in the world, along with every wrestler you could imagine. Other notables below $10 include:

• Miss America
• Miss USA
• GigaOM
• Nikolai Volkoff
• King Kong Bundy
• Der Spiegel
• Joseon Dynasty
• George IV of England (along with every other George)
• Henry V of England (along with every other Henry)
• Ramada International (along with every other hotel chain in the world)

Should Mahalo be considered a blog/wiki in the eyes of Google – meaning quick rankings like a blog gets?

Search engine Mahalo is changing their payment program to their guides beginning this Monday. Here’s the update from Mahalo, "Beginning Monday, January 14th, we will be assigning smaller dollar amounts to extremely short SeRPs or very fringe topic SeRPs that aren’t of high priority, and likewise, we will be assigning higher prices to SeRPs in especially difficult verticals or which are in high demand."

Makes sense to do the higher priority pages first, pay more for them and capitalize on the SEO referral traffic while the term is hot. When a story comes out, if Mahalo can get to it first, they might immediately rank 1st in Google based on the timing-changes Google has made. No information was provided on what the actual dollar amounts will change to – we will learn about those next week. Mahalo is also reporting today that they have passed the 25,000 search engine result pages mark. Congrats to the team including Sean Percival!

Let’s take yet another look at the SEO play topic. Anyone remember Jason Calacanis saying "SEO is Shit" at SES Chicago just over a year ago? Anyone remember Jason stating here on CN that "If Google wants to index us well that’s fine with us, but our model is not based on being well indexed in Google or Yahoo." He then further stated his views on SEO here. I still stand firm that a percentage of traffic on Mahalo is actually employee/guide traffic.

Marshall Kirkpatrick is loving the Mahalo noting, "I find Mahalo a good place to start learning about many topics". Today we learn from Hitwise that Mahalo traffic is growing nicely. Hitwise has put together some charts and stats about Mahalo that show a great growth curve. At the end of the post, they note, "Mahalo receives most of its traffic from Search Engines (76% last week)". How could this NOT be a SEO play? If Jason was serious about not being interested in search engine referral traffic, then he could easily block it. Let’s not even touch the internal promotion of articles to the social news sites.

Hats off to the Mahalo team for their early success. I just hope that Jason will admit that Mahalo is a SEO play. Sure in 10 years Mahalo might not need Google/Yahoo to drive traffic, but it’s clear today that they do and are maximizing the pages they create to use Google to their advantage. You see, Google changed some of their rankings so that new posts appear at the top even with no inbound links. Mahalo knows this and so if they continue to create pages for the hot terms, then they will rank well. Of course Jason will say that they are creating the pages because his visitors care about that content, but c’mon, let’s speak like they do on 86th street and Bay Parkway Jason.

And to be clear, there’s nothing wrong with being a SEO play – Wikipedia’s been doing it for years.

Otis over at Simpy asked for someone to pull some charts from Compete so I did. Here are a few samples of Mahalo’s traffic:

Search for Vanessa Hudgens Photos:

Search for Halo Skulls:

And here is a list of the top 25 search engine keywords that drive traffic to Mahalo:

It’s fine to be an SEO play – just apologize for calling it shit and dead. Come out of the SEO closet Jason, we will all love ya a bit more.

As I have thought about human search over the past days, for me it boils down to two main factors as to why it won't be able to compete with a machine. I thought it might be beneficial to share the reasons and what the feedback is like. I will use the latest released human search, Mahalo, in my examples.

Reason #1 – Speed

The simple truth is that a human cannot do something as fast as a computer. We may be able to do "something" better than a machine, but certainly not as a fast. I just read an article about the NYSE and how more and more of the traders on the floor have been fired. Why? Because while they believe that they can get a better deal for their customers, the machines just process the information quicker. And in today's stock market, quicker is better.

Let's look at the topic from the human search POV using the latest human search engine Mahalo. I checked Google's hot trends page for June 1 as checking today provides little on Mahalo as the humans have not caught up with it yet. I picked Katherine Heigl from the list as she had a page on Mahalo. Here is her page on Google Hot Trends and on Mahalo.

The Mahalo page is a directory for links about KH not something that will share current news. The Google page is actually very well crafted allowing the viewer to see what's hot with Katherine across the web from blogs, to news, etc. And just like on Mahalo, you can discuss the actress on Google Groups.

Now to be fair, Mahalo did create a page quickly when the NYC JFK terrorist scare was in the news last week. But how quickly can they do this worldwide? Will Jason be able to afford paying people to sit at a terminal looking for the hot topic and continuously updating it the same way Google, Technorati, etc do?

When I post content on Cn, it immediately is available via a search on Technorati. Is it like that with any human search? Even if I post something in the "top 10k" terms, will it be there? I doubt it.

Competing with Google, Yahoo, Ask and the others on speed won't work.

Reason #2 – Bias

This is a big issue to me. When Jason sits in an interview and says that Mahalo will index the "Top 10,000 searches" that is a fib. What he means to say is, "We will create pages for the Top 10,000 searches, anything else we feel like it and our friends."

This is not just with Mahalo but with anything that is 1-many from a human standpoint. When I ask a friend what is the best flavor of ice cream at Carvel, I expect bias. But I don't expect it from a search service. I expect everyone to be on a level playing field.

I think eventually bias might hurt Jason's site if it gets larger in popularity. How does one determine who gets a page. Will people claim issues of race, ethnic, religion etc? How does someone who runs a human-powered site protect against these type of issues?

Let me illustrate an example: Check the technology page on Mahalo. Look at the names that are there that you can clearly say are outside of the top 10k searches. People such as: C.K. Sample III, Om Malik, Nick Denton, Jason Calacanis, Doc Searls, Jeff Jarvis. Are these people really in the top 10k? No. I doubt it. But they have a page. Why? Because they are in the "know"? So what does a site such as CenterNetworks do to get a page? There is no way to request a page be created. And since I am not friends with Jason, the odds of ever having a page is unlikely.

If Jason asked me for my suggestion, it would be simple. Stick to the Top 10k searches and nothing more. If your name is not in there, don't create a page. Otherwise eventually you will be called out for playing favorites which will make a huge dent in your credibility.

Machines don't have bias (unless they are programmed to). Everyone has an equal chance of making it to the top.

Please note that while I discuss Mahalo in the examples above, there are other human search engines as well. I applaud Jason's efforts in promoting the heck out of Mahalo.

So it's a rainy night here in NYC and Carvel is out of my favorite flavor. As I walked back to my apartment, I thought to myself, let's see what the domain history is for Mahalo. Surely, Mr. Calacanis couldn't have been the first registrant in 1995, right? Alright, so let's pop over to the Internet Archive Way Back Machine, spin the wheel and see what comes up!

Actually, there isn't much which was disappointing. Only nine links over four years. So I started clicking, and six of the pages are your standard "your webserver is ready" crap. So I click "December 3, 1998" and what do I find? A PORN SITE. That's right folks, the new "Super great human awesome going to beat Google search engine" used to be a porn site. :)

Here is a screengrab (images didn't load):

Anyway, just a bit of fun on a hot, hazy and humid night. And I guess this post is better than eating a 2,000 calorie Flying Saucer.

I reviewed the launch of the new human directory called Mahalo yesterday. In it, I discuss four of my concerns with this new directory. I posted my review on the Wikia search mailing list and (much to my shock) Jason replied. I thought ya'all might be interested in what he had to say.

In response to my review, he replied with the following (my emphasis):

only critical review?!?! you're kidding right? the SEO slimebuckets are going crazy saying this will never work (I wonder why… :-)

if you want to do a real review I suggest taking our search pages and compare them to Google, Yahoo, and Ask. Your review doesn't really do that.

That being said we only have 4,000 pages and it's like Wikipedia in year one… we've got a long way to go. In a year when there are 10,000+ it will make more sense. also, you have nothing to lose if you use mahalo: we show google if we dont have a response–so nothing
lost, and a huge gain if we have a hit!

j

Since he did not address any of my concerns, I submitted the following to the mailing list:

Jason,

It would be great for you to provide responses to the 4 points I mention in my review. And you should know that calling people names behind their backs would never be allowed on the handball court in bay ridge. Especially since your site could become the next "experts-exchange" and since you appear to want to own the seo space, the same space you claim is dead.

I understand how your service works and whether you have 4k or 10k, it will work the same. Perhaps you can provide this mailing list with the RTB and other benefits that you sold the investors for Mahalo.

– Allen

Jason promptly replied with the following (my emphasis):

My point about seo is that it is gaming the system and done by weak people who have sites that shouldn't rank high. We are not trying to seo–we are trying to help people avoid bad sites and find good ones.

We hope seo dies–quickly. I'm afraid it will be slow however.

Seo means making your sites better for search engines. Mahalo fights against that–we reward people who optimize their sites for humans.  :)

Seo will be a footnote in the history of searxh in five years. Jimmy wales and his team, and the team at mahalo.com will end this nonsense in short order.

Mahalo and wikia FTW!!!

J

Unfortunately he did not answer three of my concerns nor share the information that he shared with his investors on why this is such a great play. Jason has a pattern of calling SEO people names and while there are a ton of horribly unethical SEO people out there, there are an equally good amount of people who are doing good things in the space. He claims that SEO is not for Mahalo but won't directly answer my question. And I have no idea what he means by "optimize by humans".

Update: Jason has sorta replied to the SEO question… check out his reply

So let’s get right to it. Earlier today, TechCrunch posted an exclusive about a new search engine founded by his conference partner, the Silicon Valley Superstar Jason Calacanis. I have spent the last 30 minutes using it and here are my findings.

Called Mahalo, meaning “Thank you”, this is a “human-powered search engine” says Arrington. Mahalo states their objective as, “Mahalo’s goal is to hand-write the top 10,000 search terms.” It seems most of the “top tech bloggers” made it in, they must be somewhere in the top 10k, I learned something new today. Actually, their technology page shows a lot more completed pages. So from what I can tell, it might be top 10k plus other “hot” names. Which got me thinking… see below.

So I guess CenterNetworks will never have a page! And neither will the other 24 billion web sites and search terms. We just get the same-ole Google results. Though there are no ads (I guess for now?) so that’s a good thing.

It appears to use a Wiki engine for the technology which will help to keep costs down along with (what I assume) a network of very low paid offshore staff to make the pages.

Item #1 – Human Search or Human Directory? 

I am not sure I get why this is a “human search” when it really is more of a “human directory” – right? Or am I missing something? Either way, maintaining those pages will be an ultra-bitch. What happens when new “gossip” needs to be added about someone, how will those pages stay up-to-date?

Item #2 – Mahalo vs. AboutUs

Now let’s get back to my title. First, what’s the difference between Mahalo and AboutUs.com? Both Wikis. Mahalo looks better. AboutUs allows me to edit my page. I don’t see much else that’s different honestly. As much as I bashed AboutUs when it launched, Ray will never have the pull that Jason will so eventually Mahalo will overtake AboutUs even though I prefer the ability to edit my own page.

Item #3 – Sustainability

Will it be possible to create a “live” search engine from this? I doubt it. Maintaining 10,000 (2008-25k) will be nearly impossible without hiring every available content person in most of the world. Another reason this is a directory, not a search engine.

Item #4 - Gaming the real search engines

More importantly, I have the same concern with Mahalo as I did with AboutUs. This is the ultimate way to “game” Google/Yahoo/etc. Create a page that highlights x person, get Mahalo up to a high page rank (with links like the one I just provided) and now traffic is sent to Mahalo instead of to the person/company web site. These type of web sites should not have a pagerank nor should be listed within the search engines. It is interesting to me that after all of Jason’s talk about SEO being dead, he launches the master SEO play.

Summary

Will this get some play? Sure, it’s like if Apple put out a half-baked iPod. Would it still sell millions? Yep. I just don’t see the value here. And while this might get to be the “answer” for the Silicon Valley/San Francisco crowd, I can’t picture my mother using this with her mahjong friends over Googling-it.

WatchMojo has a good writeup about the service and why it may or may not work.