Web X.0

An alternative Plan B for Microsoft

Yahoo remind me of the famous saying about the Palestinians:

"They never miss an opportunity to miss an opportunity"

After playing a hand of cards they didn't have with Microsoft, they have now essentially committed harkiri by becoming a node within Google's business. Why would any advertiser now bid on Yahoo's platform? I can't find a single reason, and I doubt advertisers will either (not that they had many reasons before...). I agree with Michael Arrington - Yahoo totally screwed this up, and their search business is pretty much dead.

Now everyone seems obsessed with Microsoft's next acquisition target. I'd like to offer Ballmer my 2c on a potential alternative to the $50B acquisitions they're looking to make. It's called the 200% rev-share program.

But first, a little background:
Online advertising is a strange business. While the advertisers are the ones paying for everything, acquiring advertisers is a secondary concern for an ad network. A distant second. The #1 key to making an ad network work is the publisher side. Even though the publishers are being paid, it's much more difficult to win publishers than it is to win paying advertisers. The reason is pretty simple: Ad space is binary, while advertising budgets are not. A publisher has to make a binary decision on who gets to sell certain ad space. Whoever they choose becomes the de-facto exclusive "owner" of that ad space. Publishers cannot take risks on that kind of exclusive deal, and therefore they all choose the leader who has proven to monetize best - Google in the case of text/search ads.

Advertisers, even though they are the ones paying the bills, are much easier to obtain because their choice on how to distribute their budget is not binary by nature. They can put some money on Google, some on MSN, some on Quigo, etc. Advertisers will generally follow the distribution. He who has publisher real estate will eventually get the advertisers. The other direction is far from guaranteed (see Miva, LookSmart, etc, etc).

A much better plan B for spending those $50B is by seriously upping the rev-share paid out to publishers. And I'm not talking about upping it from 60/40 (about what Google pays out), to 70/30. My suggestion would be to go for a 200 / -100 rev-share with publishers. Take those $50B and use them subsidize the publisher earn-outs for the next couple of years. 

Publishers are dying to have an alternative to Google for monetization. The trouble is that no one has been able to naturally monetize better than Google. And that gap is widening by the minute. This game cannot be won in a dog chase.

The only way to get back into the game is by locking publisher distribution, and the only way to do that is by out-paying Google, even if that means doing it artificially via subsidizing the rev-share and not though higher yield.

Microsoft should offer a 200% rev-share to all publishers (and search sites, etc) for next couple of years until every publisher in the world is talking about how much better monetization is with MSN than it is with Google. When that happens, critical mass of distribution will occur, attracting massive advertising $$'s, allowing Microsoft to *slowly* throttle down the revenue split to under the 100% mark.

This isn't too crazy... If anything, I'd go for a 300% or 400% rev-share and completely nail this down. Google can obviously react to this and raise rev-share splits as well. But for Google this is the *only* revenue source and it would be awfully painful to turn that into a money loser. Microsot can still afford to do this while Windows and Office are still the cash cows that they are. In addition, some of these publisher subsidies will be offset by the improved monetization of MSN's own properties.

This is one of those unique moments where it can really be said that it's now or never.

Breaching Google's 'moat'

Fascinating analysis of the Quigo-AOL deal by Avner Mandelman over at The Globe & Mail.

I wasn't tracking the data, but Avner points out some interesting numbers:

• November 5th, 2007 - the Quigo/AOL deal was first announced
• November 6th, 2007 - Google shares hit their all-time high - $741
• Since then Google shares are down ~37% to $471

Coincidince? Probably for the most part yes.... there are so many big factors in play here and I doubt Quigo is the most significant one. But it's still a pretty cool coincidince nevertheless...

The chess game continues...

So it looks like the predicted Microsoft-Yahoo deal will happen finally. This is waaaaay overdue - it should have happened back when MSN decided to leave Overture and pursue its own performance-based ad platform. The day that happened (sometime in 2003 as far as I remember) created 2 losing platforms.

I love the timing of this announcement. Looks like Microsoft were sitting on this waiting for Google's stock to dissapoint, and then hand it a beautifully timed 1-2 blow. Cool.
(on a side note - it's really great to see Google's stock crash from $750 to $525 within 3 months... healthy reminder that Google does not have exclusive rights to the world's wisdom and shareholder value)

Seems like the next natural move will be for Time-Warner to spin off AOL (disclosure: I co-founded Quigo which is now part of AOL), and sell it to Google (~$25B?...). The game is advertising, and the key is distribution. There are more ad $'s flowing to the internet every day, and not enough high quality distribution to meet that demand. With Yahoo out of the game, AOL will now be the #1 biggest player with available ad inventory (both on its own properties as well as on its huge platform-A). In a 2-player race, Google will not be able to afford losing AOL to the MS/Yahoo combo and will have to make this move (not to mention that it already owns ~5% of the company). Interesting days...

In case you want to get rich and/or work for Yahoo

Regardless of what happens on November 5th, and regardless of whether we end up calling social networks a "graph" or not... -
It's fairly clear that Yahoo will soon be scrambling to find a company that takes your Flickr friends, del.icio.us friends, Yahoo Mail contacts, IM contacts, etc, etc and merges them all into something that makes sense. A "social graph" perhaps.

This probably won't be a multi-billion-dollar acquisition, but it will almost certainly happen.

I don't recommend building-for-flipping, but if you're an entrepreneur and are eager to work for Yahoo with a nice signing package - here's a nice opportunity for you.

If you do use this tip, please send me a small comission in 6-8 months... :-)

{image captured here}

Now that's not nice Google...

Google has apparently made a change to their ad placement formula. John Battelle has the details here:

"Under the new system, Google will change its eCPM calculation by using a CPC that is equal to, or less than, the advertiser’s max bid CPC. This change will result in increases to advertisers’ actual CPC paid when a CPC that is equal to or less than the advertiser’s max bid CPC generates an eCPM that exceeds the threshold eCPM required for north promotion...."

The language around this change is extremely washed out and even misleading, so the change sounds quite benign... But here's what's really going on...:

Google's auction system sorts ads by eCPM (or - effective cost per thousand ad impressions).  Yahoo's Panama and Quigo's AdSonar (disclosure: I'm founder of Quigo) do essentially the same.

These auction systems charge the advertiser 1c over the price needed to secure the highest possible placement (but of course, never over the advertiser's bid). This means that the advertiser can rest assured that the only thing affecting their actual cost-per-clicks are the real market forces - advertisers changing bids, other advertisers getting more clicks, etc.

What Google is now doing is sneaking secret reserve prices into their auction and artificially pushing advertiser bids up. Essentially this is rigging parts of the auction to Google's favor, without disclosing what the reserve prices are (I assume they are different per keyword).

This is a very slippery slope... there are a hundred hidden corners where artificial reserve prices can be snuck into the auction system and Google should be called out on shenanigans like this...

[UPDATE] - Google puts a great spin on this scheme, flipping it all into a wonderful improvement for the advertisers. Google advertisers, believe me - all this does is take more money from your pockets and puts it into Google's already overflowing pockets.

Mahalo

Mahalo is probably the most interesting company I've seen in a while. I'm betting this is going to be a huge success - much more than most people realize.

Last week I said exactly that to a couple of VC's that asked me what I'd invest in personally these days. By the WTF?! looks on their faces and the questions that followed (below) I understood that I've been eternally disqualified from giving them any investment advice... ;-)

• "Why invest in a company with zero technology that relies entirely on expensive human labor?"
• "How will they ever compete with Google with all their PhD's, and algorithms, and servers, and $$$'s?!?!"
• "Isn't this the failed Ask Jeeves all over again?"
• Etc, etc.

Disclosure first: I have no connection to Mahalo whatsoever. I don't know Jason Calacanis personally, have never met him, and don't know anyone else involved in this company. I have no insight beyond seeing what's up on Mahalo.com

Here's why I think this is going to be big big big:
The best thing that ever happened in search is obviously Google. And for a while (2000-2003ish), Google's search results were way better than anything else - AltaVista, LookSmart, AskJeeves, etc. But it's growth was also it's biggest curse - the whole world became obsessed with manipulating Google results using SEO[1]. 

Today Google's results are very reasonable on average. C+ or even a B- . On the more commercial terms there's a ton of crappy link-farm, AdSense-infested pages, while on the longer tail of queries the results are usually pretty good.

Google, doing what it does best (=develop huge-scale infrastructure and algorithms), keeps innovating at a furious pace and constantly improves its search algorithms and spam filters. In classic Innovator's Dilemma fashion, they're doing an incredible job on improving the stuff that made them so successful to begin with...   

Mahalo in contrast is doing something seemingly stupid - paying human beings to create the perfect search results for the handful of queries that matter (=the ones that get ~70% of the search traffic). The most common mis-perception is that they are competing head-to-head with Google, and that doing this manually is not scalable. The truth is, I don't think Mahalo is in any sort of arms race with Google... Google is already taking care of that piece itself by spending huge capex on constantly growingly sophisticated algorithms and infrastructure. Mahalo, like Quigo (my company), is playing in the same field as Google is, but they're playing a completely different game. And that's the beauty of Mahalo.

Mahalo's assumptions are simple:

• 70%+ of the search queries are on a tiny # of terms (10's of thousands at most).
• An intelligent human being, given enough time, can *always* create a better search result than any algorithm currently can.

Google's incredible infrastructure is the world's greatest machine for providing a fairly good response to *any* query that anyone submits to it. Algorithms are great for scale, but they are a game of compromises. They're blankets that when pulled to cover one problem, expose different problems in completely different places.

That is the big disruption opportunity that Mahalo is betting on. They're never going to compete with Google on processing power, or algorithmic sophistication, or the handling of long tail of sites or long tail of queries. If I were Jason, I'd happily leave those nasty headaches to the Google (/Yahoo/Microsoft) geniuses to solve. Those issues aren't getting any simpler, so they are guaranteed to be a huge resource sucker. Arms races are never a very good business model for a startup.

Mahalo has identified the huge pile of low hanging fruit and is going to pick it up. If Google's pitch to the users is "We'll get a fairly good response to *any query* you hit us with", Mahalo's pitch will be "We'll guarantee you a *great* response for the majority of your queries. For your other queries we hole-heartedly recommend using the long-tail experts - Google/Yahoo/MS/etc"

As with the Innovator's Dilemma, Google & Co will have a difficult time emulating this because it would essentially mean taking a 180 degree turn from what has served them so well up until now - convince the whole world that their search algorithms are superior to everything else. My guess is that they'll counter Mahalo with improved search algorithms, better personalization algorithms, better spam filters, etc, etc. This may work, and it might not. After all - at the end of the day these are all just blankets.

Jason Calacanis and Mahalo have the opportunity to be the first true disruptors in the search business since Google came along and ate everyone's lunch. The fact that many folks, like those VC's I met, don't get this just makes it an even bigger opportunity IMHO because it means people (like Microsoft and PowerSet) are still going to try to one-up Google which is stupid to do against the best one-upper in the world today. The real disruption will come from "one-downing" Google like Mahalo is doing.

Should be fascinating to see this happen.

If I were a VC - I'd be on my knees begging Mahalo/Sequoia/Jason/whoever to get in on this deal.

If I were Google/MS/Yahoo - I'd try to buy this while it's young, or at least put some money in it now. Don't let the NIH choir pooh-pooh this....

If I were looking for a job at one of the search engines - I would definitely go to work for Mahalo[2] and not one of the dinosaurs.

------------------
[1] Funny thing is that the biggest driver of  all the SEO crap is Google's AdSense which actually let all these people profit from the Google traffic... but that's a story for a different post.

[2] ...or Quigo. The interesting stuff and the biggest potential upside is with disruptors like Quigo and Mahalo (or Google in '99). We currently have 19 open positions listed. Come change the world!

 

Google, Quigo and ad transparency

A couple of months ago The New York Times published a story about Quigo (disclosure: a company I co-founded). A couple of highlights:

What Quigo offers is transparency and control in what can often be an opaque business: advertisers pay Yahoo and Google for contextual ad placement on a wide variety of Web pages, but get little say over where those ads run or even a list of sites where they do appear...

...In response to further questions about Quigo, though, Google said it was prepared to make changes to its AdSense service that mimicked Quigo’s approach, an unusual step for a company accustomed to mapping the terrain in every aspect of its business.

Looks like the NYT nailed it. Today Google started following Quigo's lead on becoming a more transparent network. More about this by John Battelle, Barry Schwartz, SEW, and Mashable.

From what I can tell, the Google implementation is more lip service than a real way for advertisers to buy placements on specific publishers. That is to be expected. AdSense would not be successful if it weren't fundamentally a blind network. Google takes a small number of loss leader sites like Ask.com and AOL on which it makes little or no money. Those are thrown into the blind mix to keep the overall blended-average quality of traffic reasonable. But Google makes its real AdSense money on the very long tail of crappy/fraudulent/parked-domain/self-clicking/link-farm/etc websites. Those are the sites that advertisers would never ever bid for if they had the choice. Those are also the sites that Google can take whatever % of the revenue they see fit (which I estimate at 50% at least) because they never tell long tail publishers how much they pay out.

That's where Google's true money pot is, and if they remove their network's opacity and truly allow advertisers to bid transparently for specific sites - all that revenue will go away.

This new report is definitely a welcome change for Google advertisers. Even lip service is a form of service, I guess... But don't hold your breath for any genuine effort from Google on making its network truly transparent as long as it makes so much money by having advertisers bid blindly on sites they'd never want to be placed on. For true transparency your only choice is still Quigo's AdSonar.

Microsoft finally understands - dog chasing is dead

With today's news re Microsoft looking to acquire Yahoo, it seems like they finally get it -
The performance-based advertising world is like no other space that Microsoft has ever played in before.

In spreadsheets, browsers, servers, databases, etc, etc - the Microsoftian way of doing business works like a charm. In all those categories, dog chasing proved to be a great way for Microsoft to win the market because it could out-price and out-patient its competition. Oh - and tying products into the OS monopoly didn't hurt too much either... ;-)

With performance-based advertising[1], dog chasing ceased to be an option. You can't undercut competition, because competition ain't pricing their product (it's the customers/advertisers who do, and they price it upwards, not downwards). You can't out-patient your competition because performance-based advertising has this wonderful virtuous cycle about it:

• The more clicks you have, the smarter your yield-optimization algorithms get...
• ...as the algorithms grow smarter, you can better predict your revenue per page...
• ...as your revenue-prediction power grows, you can better price new distribution deals with the confidence of not losing money...
• ...the better you can price deals, the more distribution you get...
• ...and the more distribution you get, the more clicks (and $$'s...) you have.

This virtuous cycle means that with every *second* that Microsoft was spending figuring out its world-domination dog-chase plan, Google (and others, Quigo included) were opening the gap making it even closer to impossible for Microsoft to become a real player in this space. And every purchase of a DoubleClick, RightMedia, etc just moved way more click data out of Microsoft's hands, opening the gap even more. Chasing a dog that's only gaining momentum faster than you, well - that's not a great chase to be in... 

I've had many conversations with good friends over at Microsoft over the past couple of years. And while the notion was that "we changed", "there's no more NIH Syndrome", "we're ready to make bold acquisitions needed to win this market", etc, etc - It's clear now that none of that really registered there and the plan all along was to dog-chase Google to victory. Looks like the price for realizing this mistake is going to be pretty big.

A ton more coverage on TechCrunch, Henry Blodget... oh what the heck - and everyone else on the planet.

[1] Disclosure: I'm co-founder of Quigo, a player in this space.

GooglePoint, part 2

Some due housekeeping. Ugh - I was hoping to avoid this post...:

A few months back I predicted that Google would not develop anytime soon a presentation app to compete with PowerPoint.

D'oh! - I was so wrong. Note to self: Never underestimate the Google'rs eagerness to inflict pain on Microsoft (and vice versa...)

"Prediction is very difficult, especially when it's about the future." --Nils Bohr, Nobel laureate in Physics

Thanks for the catch, Barak.

One reason why Hotmail sucks

I used to be a big Hotmail user. Now I use only Gmail. There are probably 100 reasons why Gmail is better, and probably a few in favor of Hotmail too.

But if I had to put my finger on a single reason why Hotmail sucks, it would have to be that damn leaderboard ad on the top of the page (screenshot below).

The traditional 3-pane layout in email apps causes the most precious asset to be the vertical axis. These vertical pixels have to accommodate both the list of emails, and the email itself (not to mention all the regular stuff like menus, logo, etc).

The leaderboard ad kills 90 vertical pixels. On my 1024x768 screen resolution, the vertical viewable browser area (after deducting browser menu bars, Google toolbar, tabs, windows start bar, etc) is 518 pixels. Which means that one stupid ad reduces my precious vertical real estate by nearly 20%. As soon as the Hotmail window loads, it's apparent how annoying this ad is.

I'll bet you that the designer of Hotmail (er - Windows Live) is using Gmail.