The thing that pisses me off about Mahalo is that Calacanis goes around claiming it's some type of innovative idea/technology and that everything thing they do is for the benefit of the users. It's a straight SEO play, along the lines of About.com. Everything they do is to increase Mahalo's search traffic! Just look at the widget mentioned in that article. Does anyone really want that set of links on their blog? Does it really help anyone? And all of Jason's friends on the blog circuit still give Mahalo good press...
Actually, search traffic is not really a good thing to build a business on. It's nice to get started, but you really want something like Mahalo Answers that brings back a core group of folks over and over again to do something virtuous--like helping each other.
Syndicating tech headlines to a tech blog does actually help people. Also, syndicating the questions you've asked/answered with our Mahalo Answers widget also helps folks.
Also, I've never claimed we are so much innovative as practical. We really just try to do the simple thing: provide folks with the content, links and Q&A they want. The big win of Mahalo will be if we save a lot of people a LOT of time and keep them safe.
That being said the system is fairly advanced running on things like hadoop, nginx, memecache, etc.
Also, the idea of putting a virtual currency with knowledge exchange has never been done. If it works we will have done something fairly historic: created a currency for knowledge. That really hasn't been done before. I've been around the world in the last two years looking at virtual currency and knowledge exchange in places like Korean, China and Japan and the truth is folks are using it for games 99% of the tim e (the other 1% is virtual gifts).
We're the ONLY people to have ever put a currency on a knowledge exchange. That's fairly innovative from what folks have told me.... in fact some fairly intelligent folks have called it very innovative to brilliant.
.... time will tell I guess. :-)
all the best, Jason
ps - if anyone wants to try Mahalo Answers signup and send me your account name (or post the URL here) and I'll send you M$5 to spend.
Add rel="nofollow" to the widget links and I'll believe your goal is to help people and not game the search results like everyone else. There is nothing wrong with optimizing for search traffic. Competition is all well and good. But your site is nothing more than an ad-laden Wikipedia clone with a Yahoo Answers clone tacked on. Actually, a much better business model than most startups...
didn't google answers let you put money on questions years ago. i don't know why they dropped that feature. they are the first to do it as far as i know.
i don't think you can do it any longer, but there are still plenty of posts on the internet where people talk about it.
All this really shows is that there should be several completely independent ways of ranking search results which alternate so that it will never pay off to try to scam them all because the ROI is too small.
Gaming the google ranking system is only effective because of the monoculture.
Google's ranking blends together hundreds of variables, just one of which is pagerank. If it is so prominent, there is a reason for this: it is effective in delivering relevant results.
I disagree with that to some extent. Yes, it is still effective, but it is much less effective now than it was when it was launched.
Pagerank turned out to be the death knell to the value of linking to someone, it created an opportunity to be gamed so large that it changed the nature of the object (the internet) that it was supposed to be measuring.
A good measuring device does not 'load' the circuit to the point that it stops to function...
If you introduce a new system, it has to be at least as good as what we have now: no one would want to sacrifice relevant results.
It is wrong to think about pagerank as a single algorithm: it is a family, the common things being that these algorithms analyze the global link graph (yes, I lumped the HITS algorithm in the same family).
The problem is, if you don't use the link graph, there is not much left. There is the content, but this is not enough: before Google you could often steal someone's high ranking by just copying his content verbatim or maybe tweaking it a little. So, pagerank is actually a big improvement over what was out there before.
Pagerank can be thought as an estimation of the traffic a website gets. Suppose you can measure the actual traffic (e.g. by forcing all sites to install google analytics). That would be a pretty well gaming-proofed system, but you would introduce so much inertia - every new site would have a huge chicken-and-egg problem to overcome.
Actually, I've talked to search engine companies about syndicating content between blogs and they have said over and over again that there is no issue with this.
Everyone has a blog roll, and most blogs have a "WIN Grid" at the bottom (a feature I created at Weblogs, Inc. where at the bottom of engadget/autoblog/joystiq/etc you see the headlines from the other sites).
From what I've seen said at conference the search engines only pass a small amount of page rank from these areas so it's not a major deal to syndicate your content around.
HackADay is now owned by Mahalo (we didn't sell it to AOL in the WEblogs, Inc. sale for obvious reasons... like hacking TimeWarner cable boxes :-). There is nothing wrong with syndicating headlines between mahalo and HackAday. If we're told there is we will stop.
This is just the SEO community upset at me for saying that "seo is BS" back in 1995!
"This is just the SEO community upset at me for saying that "seo is BS" back in 1995!"
I think it's more of a case where they are upset at you for gaming them for knowledge and advice, then beating them at their own game, which, by the way, has resulted in me having a tremendous amount of admiration for you...
You've 'scaled' SEO and kept the product quality. Kudos, J.
Edit: Oh, and it was 2005, not 1995 - since you seem to appreciate people correcting you ;^)
not a conspiracy -- if your entire startup is focused on essentially ensuring your pages are top of google results, and you're a killer networker like the Mahalo CEO, you'll ensure you have someone you can call and whine at if you suddenly get blacklisted.
You honestly believe Google would ever blacklist a prominent startup like Mahalo?
I wasn't offering a counter point or opinion. I don't "honestly believe" anything. I just wanted you to explain your reasoning, adding some more information to your comment. After first read, it sounded like a conspiracy theory. (I think if someone at Mahalo can call someone at Google and get special treatment, that qualifies as a conspiracy by definition.)
I don't understand why Calacanis or Mahalo are important enough to warrant such special treatment, but I don't think I'm Mahalo's target audience. I've never visited the site. That's why I was interested in more information. Thanks!
While this seems like shady tactics from Mahalo, it also seems to suggest the need for a better (read: less game-able) ranking system than PageRank. There's now tons more data about which sites users visit for a given query, whether they return to the results page, how much time they spend on a site, etc. If you can sift through the noise, it seems like this would be an even more accurate indicator of site quality than hyperlinks to a page.
Don't get me wrong, sometimes 'copying' is great when somebody truly believes they can solve a problem better, but that doesn't seem the case here — the only new feature I see is that you can add a bounty to your question; other than that, I see no advantage over Yahoo! Answers.
Calacanis' playbook is to add features early and often. Their 'answers' feature is an example of this. Their long term strategy is all about content, or pages with aggregated content/links that help people find the content they need. It's actually a great idea, instead of relying on Google's algorithm 100% to serve the right results, Mahalo is employing people to aggregate the content.
Yahoo's 1994 playbook is to add features early and often. Their 'answers' feature is an example of this. Their long term strategy is all about content, or pages with aggregated content/links that help people find the content they need.
It's actually a great idea, instead of relying on Google's algorithm 100% to serve the right results, Yahoo in 1994 is employing people to aggregate the content.
That's right... we try a LOT of different things, put them in the lab and test them in the real world. We're getting over 5m uniques a month and the largest growing part of the site is the direct traffic to Mahalo Answers.
IF anyone wants to try Mahalo Answers create an account and email me at jason at Mahalo.com and I'll put M$5 in your account to play with. 60 days in where getting fairly good answers to a range of questions.
its interesting that that is how search engines began -- directories, we've got toward algorithms, and now we're pushing back to "manual" content assignment.
Is this because no one can do it as well as google, so we're having to provide different methods?
By the way, I was chatting to Jason the other day about the tipping/bounty feature:
"So far the tipping results in 2-10x the amount of response... so clearly it works. Give it time! :-)"
Maybe it was worth re-creating Yahoo's service just for that one extra feature, we'll see! (Though it won't have been if search engines penalise them for cheap tactics like this.)
Calacanis once mentioned that they moderate these heavily, and even fix spelling to try and avoid the "How is Babby formed" type posts you find on Yahoo! Answers.
Yes, this is true. We spell and grammar check the popular questions on the service in order to make it a more pleasant experience... I know if I misspell something on the web I would appreciate someone fixing it for me, as opposed to 10 folks attacking me.
In fact, that's the other big secret of why Mahalo Answers is having a fairly strong start: we don't allow obnoxious behavior. If someone responds with "why don't you google it n00b?!!?" we delete the response and warn the person once before closing their account.
We have a "no a-hole" rule basically. I'm going to do an email newsletter about the things we've learned with community building. My email newsletter is at http://www.tinyurl.com/jasonslist if you're interested.
I don't think the downvotes were for the mistake that was made (forgetting about nofollow), but rather the perceived intent behind the post in the first place.
mahalo is the first result, there are other very popular queries where mahalo is on the first page. So they get a bunch of relevant clicks and then they put ads on those pages so surely they must be making decent money.
I doubt they make enough money to cover all of their employees but eventually they can just fire all of their employees and live off of the content revenue for a few years or sell the site to somebody who will.
That's a great how to article. Our mission at Mahalo, since not everyone here has spent 20 minutes having coffee with me (yet) is the following: "to help people find information they can trust."
We are testing different ways of doing this including:
1. hand crafter search results with content
2. a knowledge exchange (Mahalo Answers)
3. how to articles
4. (coming april 10th)
5. (coming September 15th)
The real goal of Mahalo is to build a service that combines the three most important services on the internet: content, search and knowledge exchange.
This page has nice wikipedia/About.com style content on the left, search results and Q&A on the right.
If a normal person (i.e. not a ycombinator/Techcrunch50/Scoble type) comes to a page like this they have a really nice spam-free experience.
We know this because we actually have a lab at Mahalo where we study human behavior on the web. If any of you would like to see this here is a video from before we launched:
http://www.youtube.com/watch?v=OoehBwTAH5Y
In these studies we found that folks LOVES our model.
I like that service - I've used Mahalo several times - and generally the idea of aggregation is one that I think can serve a better purpose. However, I've noticed that Mahalo has been shifting farther and farther away from its original beautiful aesthetic. Now your pages look crammed rather than pleasing.
I think that's a slightly cynical view. While I can certainly appreciate and relate to the distaste for SEM/SEO gaming, their article for how to play the guitar is a very newb-friendly summary and a lot better than 90% of the sites out there. Look at the top results, it speaks for itself.
A successful startup does not always rely on unique and innovative technology, the execution of a consumer "brand experience" can be as equally important. In this case, their IP is human-driven. Google's algorithm is nice but it's not AI yet.
To me, Mahalo completes the yin-yang of search. The concept is not new, human-filtering was around in the 90-00's. Popular algorithms and directories have come and gone. Mahalo is doing the best job of it today.
Its kinda funny, if some guy did the same exact thing to increase page rank for his blog, Google would come down on them hard the second they found out. Mahalo on the other hand gets a free pass
