Mahalo 4.0: All about learning | Hacker News Reader

Mahalo 4.0: All about learning

thesethings.posterous.com

31 pointsthesethings46 comments

I think they see the "content mill penalty" writing on the wall and are pivoting to spamming YouTube in advance.

Hadn't considered that. Can you expand a bit? I'm really into YouTube and would love to hear your theories.

I think it is a little harder to spam YouTube (with content) than the web. On the web, you can just parse feeds/spin text. On Youtube, you have to upload things. According to the presentation, they built out studios and are recording videos at a fast pace (and they do have thousands of videos on their channel, all with a signature white negative space backdrop). It is totally possible to make bad videos, not denying that. But it's a bit harder to trick systems into thinking there's something there that's not/misdirect.

Not saying there aren't other places in the system to abuse stuff (On YouTube you can friend spam, Fave-spam, etc), but it's hard to do with content.

Were you thinking content spam?

Instead of low-quality articles meant to get search traffic and AdSense revenue, these are low-quality videos meant to get search traffic and AdSense revenues. YouTube is the second largest search engine and also has a lot of space in Google SERPs.

It's also almost identical to Videojug, if you want to see how something like this plays out.

Our videos cost hundreds of dollars to make, are hosted by experts, shot in 1080p HD in a studio, have perfect audio and are on super clean white backgrounds (typically).

They are 10-50x better than most of the videos on YT in terms of quality and production value.

Most videos on YT don't have an expert and are shot on flipcams or iPhones with an open microphone, no lighting and background noise.

Our goal is to make HIGH-QUALITY videos that help people learn. If we do that we will be one of the top 25 sites in the US..... inside of three years.

You claim to be shooting for 500,000 videos in the next couple of years. There's no way that each cost hundreds of dollars (just $100 would make it $50M) and are all expertly made / edited.

http://www.sfgate.com/cgi-bin/article.cgi?f=/g/a/2011/01/25/...

"The goal this year is to start creating (from scratch) 2,500 original videos per week in 4 studios at Mahalo's L.A. headquarters. And Calacanis says the goal over the next couple of years is to reach 500,000 original videos."

That's 625 "professional" videos per studio per week. Sorry Jason, I don't buy it.

The ads seem to be way down, maybe they are trying not to look like a content mill while there is a heap of Google attention on the issue.

You are totally right about the ads being down. That's why it's so unfortunate/confusing that the how-to pages still feel cluttered. Visual clutter on sites like this are often just to make places for ads that "blend" in with content. But the new Mahalo has few ads.

As I wrote in my post: there are multiple sections of interesting information... presented in a disorienting manner.

I feel like the new learning mission is sincere, but am waiting for more tweaking.

You are right, Google is gearing to change its algorithm to start penalize pages instead of sites. Good sites with few bad pages will see that not all their pages are ranked. Whereas bad sites with few good pages will see that only their good pages are bringing traffic. So maybe this could be a nice motivation for bad sites to add more good pages.

Well, we never made low-quality content like Demand Media, but you are right that making high-quality videos with EXPERTS in them keeps us out of trouble with search engines.

Google will solve the ehow/demand media problem this year--or consumers will by blocking specific sites.

When that happens Mahalo's traffic should actually go up, because there won't be people with 11-year old spamming domains standing between us and teh top 10 results!

good observation!

If Mahalo has been such a loved product with versions 1.0, 2.0 and 3.0 why are they basically scrapping all of it and releasing a new product?

It must be hilarious being an investor and seeing entrepreneurs pitch so confidently with ideas and then the next week see them something almost entirely different and totally confident in that.

Oh right if you pivot it's okay.

Why do they cling to that brand? I've only heard about it here, and the opinions were unanimously negative.

I'm sure I'm not in their target demographics, though. Who do they appeal to?

The 99% of the population who is less internet savvy than you.

They haven't scraped it at all, they've just evolved the product a bit. I think the new designation is more showmanship then "pivot"

I wondered about this exact same thing and wrote in length about it this morning: http://robisit.com/mahalo-version-4/

"If you remove the video revenue sharing scheme from Mahalo 4, you’re left with a ballooning group of in-house staff members that are creating content. Mahalo version 1 started exactly the same way. Large groups of people were hired in-house to generate written articles and cull popular hyperlinks. Shortly thereafter, Mahalo 2 relied on offshore outsourcing and then inexpensive remote American labor to mill content.

When thought of in this light with a view of Mahalo’s history, it starts to become clear that Mahalo 4 is an iteration that has already occurred. Mahalo is iterating over itself. Partly due to the fact that they have enough money to do so, partly because the revolving door of staff members hasn’t experienced previous iterations and partly because online video is simply making money for content companies at the moment."

Incorrect, but I wouldn't expect you to understand the complexity of the business without being inside it.

1. We had inhouse people, people at home and people working in Manila (on data, like fast facts--not full- on writing) of articles and that worked well. In fact, you can almost make the at home and in the office model work if you keep the costs down to $10 an hour/$10 a page. To have folks in the office you would have to do it as $400-500 a week and no benefits and no chef (which kind of sucks), so work from home wins this model (as it did on Demand Media, Weblogs Inc, etc).

2. Mahalo is 40 months old now, which means we have better SEO, better ability to move folks around the site and a better ability to know which pages make money and don't. So, we can make 3x as much now per page as we did back then (or more). So, we've learned a lot.

3. Video is VERY different than writing articles. I won't get into the metrics, but you can't compare it to articles. also, the traffic to each is very different.

So, Mahalo will continue to grow and folks will continue to misunderstand how it works... and that's just fine. We're not here to explain how to run a business to folks who don't know how to. We're here to build an amazing place where people can "learn anything" and to get a return for our investors.

We are the 160th largest site in the USA now.... and if you counted our video traffic to our domain traffic we would be in the top 100 (maybe #75). Quantcast doesn't do this (grrrrr).

Next year we will be in the top 100, then top 50 and then in the top 25.

at that point you'll write a long post and explain to me how i don't know what I'm doing, and i will write a long post back to you explaining how business works..... again! :-)

You are a cynical one pclark!

keep in mind that Flickr, Groupon, Nintendo and Nokia are all pivots. Your first idea is rarely your best, and pivoting is what great entrepreneurs do while B-level entrepreneurs run their companies into the ground.

Now, factually you're a little off... so please let me get you up to speed:

1. We've been doing articles and Q&A for 3 and 2 years--so that's not being scrapped. 2. A year ago we stopped pursuing human-powered search since it wasn't growing 3. We examined what was working and it was helping people learn--so we branded the company today "learn anything." This is a great big mission and it will make us a top 50 site in the next 2-3 years.

We currently have 105 team members and are adding 100--and we hit profitability in 2010. We are the 160th largest site in the United States not counting our YouTube traffic (1M+ views per day). If you added that together we would be top 100 today in fact.

In fact, I just left a meeting with one of our investors and they are thrilled with the fact that they invested in Mahalo and that we tried hard to make human-powered search work. They love the fact that we iterated on our idea. So much so they asked if we would take MORE of their money.

You must be a young entrepreneur, because success in life is not about confidence--it's about humility and the ability to try things.

Good luck to you!

Thanks for the clarification I'll be sure to pivot my comment to reflect what the audience wants!

Well, if commenting is your job then perhaps you should. That is the point being made here I believe.

If anyone has ever paid you to be correct or insightful with your comments, then perhaps you should consider it.

Thanks to Jason for dealing so well with such snark.

I think it's a really smart idea. How-to content is huge. Amateur videos of esoteric hobbies can get a quarter million page views in a couple months. I think the biggest risk they are running is trying to produce in super high volumes. I think this will lead to less compelling content. At first they can make it up in volume, but if they have success it will be relatively easy for domain experts to copy their approach and produce relatively fewer videos, tailored to the field, which will lead to higher CPMs.

The other big opportunity is using video as a gateway to ecommerce. QVC is the 2nd highest grossing tv station. There is definitely an opportunity to take this model to the web.

Does anybody here use Mahalo? What are some use-cases where it makes sense?

EDIT:

So I checked out the new Mahalo's advertised feature, "How to Convert Fractions":

http://www.mahalo.com/how-to-convert-fractions/

They have a video which is pretty good. (I wonder if creating hi-quality videos scales to satisfy investors?) Then they have the steps in the video in text format with words such as "Fraction" hyperlinked to a spammy Mahalo page:

http://mahalo.com/fraction

Below this content they have "Related" stuff like "How to convert fractions - XXX" where XXX = {Videos, Images, Newsfeed, Twitter}. These are {Google, Bing, Twitter, etc} auto-generated spam/noise and contain no useful information. (You can't auto-generate a sensible feed about converting fractions.)

Overall I'd rate the new mahalo.com as spammy. The videos + transcribed text may be good, but there's too much surrounding crap.

I was going to ask the same question because the only times I ever hear about Mahalo is from people who complain about how they are spamming. Haven't heard from a single user of the service yet.

The spamming claim on HN is one done by a secret kabbel of SEO's who I've pissed off over the years.

These guys basically obsesses over everything I do and follow me around the social web. They vote my stuff down, flame me and generally try to bust my chops.

They are known as "jayters" because they hate on Jason. I love my jayters, because they make me appreciate how lucky I've been in this life. They have every reason to hate on me: I drive a tesla, I have not one but TWO bulldogs and I get more press they anyone deserves.

Hate on!!

Hi. I've never seen mahalo.com before nor have I commented on it. What I wrote down was my first impression as a user.

Also note that I pointed out the positive: "They have a video which is pretty good." But all the auto-generated stuff surrounding it is noise to me.

Best of luck with mahalo.com.

You're referring, of course, to the Mahalumnati?

FACT CHECK!!! FACT CHECK!!! FACT CHECK!!!

BTW: The page you are looking at -- http://www.mahalo.com/search/?q=fraction -- is a page we haven't built yet.

When we don't have a page built we show you Google/Bing search results.

These pages are NOT indexed in Google.

So, they can't be google spam. :-)

I didn't say it is FOR GOOGLE. I said that the page contains numerous boxes which are search results from {Google, Bing, Twitter}, which IMO don't add much value to the topic.

For example:

  http://sp.life123.com/bm.pix/fractions2.s600x600.jpg

or:

"@wyrdgirl_aj gurl ima call u after this 930 class... We learnin how to convert fractions to % -_- #bored from @QueenofKings16"

Best of luck with the new mahalo.com.

A fun side-story: There was a question on Quora recently - "Why did Duck Duck Go Block Mahalo.com?" and Jason Calacanis, of course, felt it necessary to weigh in. I had to read this response twice, assuming there was some intended irony:

"if they did that is a big mistake... we are not close to spam and a we have the best articles on many things like how to play guitar, speak french, etc."

http://www.quora.com/Why-did-DuckDuckGo-block-Mahalo-com

I know the HN community has mixed feelings (at best) about Mahalo, but definitely watch the video!

When I saw Mahalo in the title I almost skipped the article completely, but Jason provides a solid dissection of his company's evolution and the strategic mistakes they've made so far. (Plus it's pretty short.)

(Skip the Mahalo promo at the end.)

Thank you for actually listening. :-)

I would like to offer you a complimentary ticket to the LAUNCH Conference as a thank you. Feb. 23/24th in San Francisco. Ping me at jason@launch.is and i will arrange.

I was ready to say mahalo was the biggest spam website on the internet last week after the whole google search debate but now I think mahalo may have a killer product. right now it's like wikipedia 2.0 .... so far I like the 4.0 version ...maybe I'm wrong but I like this pivot

Thanks for a considered response.

May I offer you a complimentary ticket to the LAUNCH Conference as a thank you? Feb. 23/24th in San Francisco. Ping me at jason@launch.is and i will arrange.

it's going to be a killer event (see www.launch.is for details... 40 companies launching killer products over two days!).

Mahalo 4.0 is great! Please send my ticket to ... ;-)

4.0 as in four steps ahead of Google in the art of spam.

lame.

cynical.

Looking forward to Mahalo 5.0 and their next pivot.

This should be code-named Hawaii Five-O .

cynical

Since you are focused mainly on video you could check Impossible Software , their demo in DLD 2011 was quite impressive . http://www.dld-conference.com/articles/article/dld11-live_ai...

Instead of How To Pages they're going to concentrate on How To Videos. Not all that much of a pivot IMHO.

Pros:

+ It should be more popular than user-generated content of the same type (overall).

+ Very few other companies would be willing to invest as much time and money into competing. How To pages are much much cheaper.

+ High CPMs if they can build a trusted brand with quality viewership.

+ There's less competition for video content than text content, and it's harder to steal (no synonym substitutions, etc).

Cons:

+ Hugely expensive. Low profit margins.

+ Very slow to generate.

+ Far less video views than page views (less ad impressions). Many on YouTube itself.

+ Hard to manage and keep quality at large scale.

+ Very small buyer market for high CPM video ads. Low click through.

I guess this is in direct response to Matt Cutts openly pointing Mahalo on HN. We should make the Aaron Wall the real SEO police. Just read the seobook blog and you will see all the sneaky things sites like Mahalo and ehow does.

Lame.

Slander.

Please read HN community standards.

Could you please list a few things they do?

The problem with all these pivots is that Jason never addresses the real problem looking for a solution:

The Problem: How do you couple an ad based funding model to user generated content without having that content model destroyed by ads displayed not in content of content and ads and ad frequency not fitting with user generated content.

Thee is the problem Jason and you still have not solved go back to the drawing board!

And to be fair I do not know if there is solution to the problem with a non-volunteer human infrastructure...maybe there is not one.

Make lessons 1-10 free (20-30mins each), charge for lessons 11 onwards. Most people that make it to lesson 11 would be willing to pay going forward, after seeing the quality of lessons 1-10.

Thank you for a great insight.

Would love to have your intelligence at the @LAUNCH conference. please ping me jason@launch.is and i will arrange.

Reply on news.ycombinator.com