Here's a sample from Part 4 of the series which covers Soy Sauce:

Then comes the more advanced part of dealing soy sauce: How sushi touches the soy sauce before it goes into your mouth. This may be surprising, but you should dip the fish side, not the rice, into the soy sauce. Many people in Japan do not know this, either, but this is critical in good sushi eating. How you do it, however, is a little tricky.

When you pick up a sushi piece using your thumb, first and second fingers, you just turn it over so that the sushi piece is almost upside down. If you are using chopsticks, you might first want to knock down the sushi piece sideways, so that you only need to twist your wrist 90 degrees to dip the sushi upside down into soy sauce..

When you touch the soy sauce with sushi, it has to be quick, almost like a fish jumping up from between the waves. You do not soak sushi in soy sauce. Sushi is after all about freshness and liveliness. So, remember, if your sushi piece resembles yourself in a bathtub on weekends, resting comfortably in liquid very long, you have done it incorrectly.

Many sushi connoisseurs say that it is the fish side, again, that first touches your tongue when eating. This way, you can enjoy the taste and the texture of the fish well with some flavors of soy.

¶ 12/29/2006

Behavioural Targeting (or BT) is simply the ability to show you ads based on what you do (your behaviour) on the web.

The Wall St. Journal has an article discussing Microsofts attempt to use behavioral targeting in combination with search ads:

If someone types in "compare car prices" on Live Search, Microsoft's computers note that the person is probably considering buying a vehicle. The computers then check with the list of Hotmail accounts to see if they have any information on the person. ... Microsoft says that in testing in the U.S., behavioral targeting increased clicks on ads by as much as 76%.

Microsoft doesn't really matter much, especially in search. The biggest player in BT is Yahoo!, since they have the most email users and a broad network of content sites. Google could start to do it more as well, mostly based off of your search history, though they've mentioned they won't use it with AdSense. Tacoda is the best known, and probably biggest 3rd party ad network which touts BT more than anyone else.

Behavioural targeting is important because it works and when you have a huge database of peoples interests from their click stream and email, you can really improve ad response rate. In 2007, you should see a lot more offerings, a lot more proof that it works, and the price should go up a lot too!

Labels: 2007, advertising, predictions

¶ 12/28/2006

Here's an excerpt from a recent post that shows that he understands Google's grand strategy as well as anyone:

[Google's] actions in the user generated video space defines that I think will come to be their approach to capturing content markets. Pay a lot of money in a land grab for the best properties you can find and aggregate them in a manner that can pre-empt or displace the existing competition, and then give the content you just paid a boatload to license away for free, hoping to make your money back along with a return by selling advertising around it.

Simple:

• Buy premium content with GoogleCash.
• Give it away.
• Sell ads.

In an earlier post about Google, Cuban summarizes why Google can afford to do this. In fact, he calls it arbitrage:

I think their ability to excel at monetization dramatically above and beyond their competition has put them in a unique position to arbitrage the financial expectations they have for a page view vs the expectations of everyone else on the net. Myspace, AOL and many others are on a list of partnerships that appear to be made on a simple principle of "We can pay you more for your traffic than you can earn for yourself".

Or as Cuban summarizes:

Get big, subsidize and monetize.

Google alone can leverage the 60%+ search share to send people to their content.

Furthermore - and this is key - Search is the most intentional behaviour on the internet. That's why they can afford to pay more for the content than others. Advertisers are happier getting leads from search activity than anywhere else. ( That's also why they invest so much in infrastructure.)

¶ 12/28/2006

If you don't already have a blog, here's my best advice to you: Don't use blogger. It will cost you visitors, readers and cause you headaches.

Here are the reasons not to use blogger:

• It's not good at trackbacks - so it costs you links and traffic.
• It's slow.
• They've "upgraded" and have 2 separate systems which aren't compatible.
• It's hard to set up sitemaps with all your posts on a page.

What you should do:

• Get your own domain name.
• Use wordpress.

I know you can use blogger on your own domain, but I still would use a different package to run the blog.

¶ 12/28/2006

2007 will see mass consumer adoption of smart phones (i.e. More than 5M units sold). These are cheap blackberry clones - phones with mini keyboards.

In the US, this trend will be the thing that puts mobile apps on the map.

Why has it taken so long? The main barriers have been cost, screen resolutions and 3G networks. All those are now solved.

These trend-creating devices cost about $200, and will enable mass adoption of texting, mobile email and creative mobile applications. Examples of these phones include:

• Samsung Blackjack
• Motorola Q
• Nokia E62
• Blackberry Pearl
• Treo 680

Paul Kedrosky recently posted some stats that 2M BlackBerrys out of 7M total sold are in the hands of consumers, rather than business people.

This is my most obvious and sure-fire prediction for 2007, and it will have a huge impact on the way the whole internet is used over the next 3 years.

You are hereby behooved to get one of these phones or risk missing out on the next big phase of consumer web, social networking and search development.

Labels: 2007, mobile, predictions

¶ 12/22/2006

Here's what I said last year in a post where I was trying to predict the major event for each of the upcoming 6 months:

January: Google Calendar releases. Integrates with Gmail by sucking dates out of your mail and into your calendar.

Good. 1 for 1.

February: Google Wallet or "Cash by Google" gets pushed out. Designed mainly to let you buy stuff from Froogle vendors.

OK. 2 for 2 - Google Checkout.

March: Google buys a music startup like Pandora.

Totally Wrong. They bought You Tube instead. 2 for 3.

April: Yahoo releases revamped Y! SMS (aka Overture) so the auction model works like Google, based on relevance. MSN takes away a bunch of inventory from Yahoo (i.e. MSN itself).

Totally wrong on Yahoo's time frame. MSN of course did take the inventory, and managed to decrease it as well. 2 for 4.

May: Ask Jeeves is re-branded / renamed. Still remains irrelevant, however.

I think this is right. I really haven't looked at Ask all year. 3 for 5.

June: MSN and Yahoo fight for more inventory, buying up properties, but they struggle and Google is more dominant in PPC than ever.

Good. MSN and Yahoo did do some deals for inventory (i.e. Facebook and Ebay). 4 for 6.

Overall, the best prediction I made last year was that Google would be MORE DOMINANT than ever. That's been true in spades.

Labels: 2006, predictions

¶ 12/21/2006

Google is giving away $10 on $30 purchases for customers who use Google Checkout, and this is causing big uptake from merchants. Ebay expert Scot Wingo agrees with the effectiveness of this promotion (Google Checkout's transactions are up over %1200 in Q4 from Q3) - it's definitely attracting eBay sellers to use Google Checkout and eventually AdWords. It's also attracting major sites like Toys R Us, which is promoting Google Checkout as it's top payment method.

I feel a bit vindicated by the positive reception to Google's strategy, because I thought Google Checkout was going to be a big deal from the beginning, especially relating to EBay's community.

Here's what the NYT article misses, and I still think most people don't see: Google will build a distributed "trust" system, much like eBays. The trust system will work across any website, and allow consumers to review which sites are worth buying from. Combined with the knowledge Google will have of people's purchase history, it will make Google Checkout a very strategic long term initiative, and a big problem for eBay and potentially Amazon.

Checkout is really a brilliantly simple move by Google:

1. Change the credit card processing game by making it free.
2. Reap the benefits of getting more merchants into AdWords.
3. Be the source / arbiter of trust in internet commerce.
4. Own the purchase data for huge numbers of internet buyers.

I think Google Checkout is a key initiative which opens up growth on the needed scale to make an impact on Google's future value.

Labels: ebay, google, google checkout, predictions, trust

¶ 12/20/2006

More importantly, this is a harbinger of the death knell for the SOAP web services protocol (which was originally spear-headed by Microsoft). Along with ex-Googler Nelson Minar saying that he wouldn't use SOAP if he had to re-design the AdWords API, it's more evidence that the far simpler REST approach has won.

¶ 12/19/2006

How do I know it's over? A couple of things. One was when I built Texsy.com and started to promote it to bloggers. Unfortunately I found that long tail sites have no traffic. Basically, to make an ad network succeed you need BIG sites. One large site is worth, oh, two million small ones...

Second, readwriteweb has a post that uses data from Compete.com to show that contrary to Fred Wilson's intuition, the big sites are getting bigger percentages of internet attention.

It was a good story though...

¶ 12/19/2006

Overall theme of this show is growth in affiliate activity, and Jeremy had his buddy Todd (aka Stuntdubl) collect a bunch of affiliate questions. I transcribed / paraphrased his Q&A below.

Q: What do you look for in affiliate programs? Terms?

A: Word-of-mouth from other people. Payout is important, determining conversion ratios depends on testing over time.

Shawn said: "Trust" - wants to trust the program and know the people - how do they treat other affiliates, how do they convert, do they pay on time.

Contact the merchant / affiliate broker - if you are doing something shady / 'innovative' / in the gray area. Check it with them, so you don't get banned and lose commissions. Stay in communication with the affiliate manager.

Q: As a publisher is it sufficient to just create landing pages for particular products or better to create a unique site or domain?

A: It really depends on the product - on ringtones, you want a high quality landing page. Doing more work and more research makes it easier to develop something unique. Use tools and track. Mentions crazyegg.com as a service that collects good stats. Owning your own domain lets you do tracking better.

Q: How do you negotiate a better deal or find "private" affiliate program?

A: Every company has leeway. They always can pay you more, if you get what's advertised you need to press. (He encourages you to be aggressive - especially if you have the volume). "Never show all your cards". He kinda emphasizes "the squeaky wheel gets the oil."

But he gives great advice - don't ask a question which is an easy no. Don't ask: "Can I get a pay raise?". Easy no. Phrase as: "What can you do for me, I need to get bumped for this volume."

Most of the time when Shoemoney talks about "affiliate" companies, he's talking about CPA places like AzoogleAds, CPAEmpire, XY7. He talks about Commission Junction as well.

Q: Do you sell leads to multiple programs?

A: Shawn answers: make sure it's ok with the partner companies if you do it. He's partly talking about advertising multiple offers on the same page, versus say mortgage leads, which can be sold to multiple places.

(Ad Break - there's a funny spoof ad in here with Shoemoney rapping/beat boxing some X-mas songs )

Q: Affiliate links and commission hijacking

A: Shoe brings up cookie blocking in Spyware Search & Destroy. Which could prevent DoubleClick and AzoogleAds leads from getting credited. Shawn adds that the toolbar / spyware world can over-write cookies - taking credit for affiliate sales. Stuff to watch out for... more awareness is needed... Flash Games, DRM video, Zango, certain toolbars do this, and the affiliate companies aren't doing much about it.

Best way to protect yourself now: Watch your stats / ROI.

Q: What do you use for tracking?

A: Shoe uses Webalizer (log analysis) and in-house stuff reporting on ROI. Also crazyegg to improve landing pages. Mentions click fraud tracking from enquisite.com. Clicktracks gets a vote from Shawn.

Q: How does he target typos via ppc ( strategies etc.. )? How does his typo strategies differ from engine to engine ?

A: Shoe doesn't do typos on MSN, since traffic is so cheap. Mentions the MSN limit of 40k keywords per MSN account. On Google he does almost nothing but typos - mentions Syndk8 forum has tools for generating "one-off" keyboard typos. On Yahoo - quality control makes it painful to do too many typos. He brute forces it on Yahoo.

Q: Can you explain what incentivized offers are? How do you make money and keep fraud down?

A: Incentivized offer: you can give people an incentive for completing the offer(s). Like a free iPod, or a free TMX Elmo, or a gift card.

There's a ton of fraud since people can put in worthless lead info. Leads get rejected by the affiliate manager.

Shoe says you have to maintain a database of email farms and he filters out those bad leads.

It's a differentiator if you can keep the quality of your leads up. Shoe advises communication with the affiliate managers, again.

Q: If I get a hot chick to write your name on her chest and take a pic, will you link to me?

A: Yeah. He'll totally link to it. Maybe with a link condom... (I KNEW the answer to this one... :)

Q: Besides PPC, what are some of the best way to drive & increase traffic to a real site for affiliate links?

A: Could use Myspace, Digg, social sites - "Ten ways to make money online" and you add your affiliate links. Digg users will click on certain affiliate links, but not adsense.

Use forums. Start a thread, and drop your link in...

3rd tier PPC like AdBrite, Kanoodle - take a lot of experimenting.

Cloaking is OK by Shoe in some cases. I.e give Digg users other stuff than AdSense, because you know they aren't gonna click on that. Show them something else!

Q: If he lost everything (his websites, partnerships, etc) today what would his actions be tomorrow?

A: He pretends like this happens everyday! What to do now? Keeps him hungry. Come up with new ideas.

If he could go back 5 years, he would have done more social media, more ringtones, more community driven sites around music.

Personally, I would have bought a lot more good domain names if I could go back :)

Also Shoe would've started his own affiliate company. He would've pushed the gray areas oriented. Wouldn't have listened to the nay-sayers.

He thinks it's a great time to be in this business. Place to be.

Q: What does he feel is the single biggest mistake most people make that keep them from making more money?

A: He's emphatic here: People think of reasons why stuff won't work, they talk themselves out of it. Jeremy, says "JUST START WORKING ON IT". You'll learn. Don't sit on it. Just do it.

Basically: it's Yoda: "There is no try, there is only do..."

(For more info, check the forums that Shoe checks on: he reads ABestWeb.com, DigitalPoint, and he's got his own forums)

¶ 12/14/2006

Normally, you have to wait a year or more to vest and then exercise stock options. But Google is setting up a market that will let people sell their options basically anytime they like.

What does this mean?

• Google just got a lot more appealing for people who dread the commitment of 5 years of employment before getting a payout.
• Google is now that much more attractive than say Yahoo or Microsoft, or any company which doesn't have this program.
• Current Google employees are gonna be spending a lot of their money sooner!

Basically, as those two Irish guys in the Guiness Beer ads say: BRILLIANT!

Once again, Google out-innovates the competition. And this time it's in the human-capital arena. It'll be fun to watch as the rest of the world figures out what this means.

A side-benefit: we'll be able to see which other companies are even close to awake by how fast they announce a copy of this program. It doesn't hurt to have a stock that keeps going up, though.

Update: Some blogs seem to think that this program is only valuable to those whose options are underwater - i.e. the option strike price is higher than the current share price.

While that is true, it overlooks the time value of the option. It's also possible that an employee option that's in the money (say with a basis %10 less than the current stock price) is worth MORE than the price differential of the option and shares.

I.e. Potentially even for in-the-money options you could make more by selling the option than by exercising it and pocketing the difference.

So the program really does two things:

• Makes explicit the time value of the option.
• Gives employees with vested options more opportunities for monetization

It's simplistic to opine that this program is some sign that the Google financial people think the stock isn't going up anymore...

¶ 12/13/2006

However, the first comment on Josh's recent homage to his beloved HP-42S calculator proves the old (Homer Simpson) maxim: No matter how good you are at something, there's always about a million people better than you.

In this case, the commenter trumps anything Josh could ever dream of, with a stunningly simple opening:

Bill Starr said...

Hi, Josh. My google alert on "HP-42S" articles just pointed me to your post.

As they say in Chess: !!

¶ 12/11/2006

Berkowitz does not defend the brand choice he inherited.

"I don't know if Live is the right name," he said, adding that he had not decided what to do about it. But before he gets around to deciding whether to change the brand, he wants to make Microsoft's search engine itself more appealing to consumers.

This is like any new Hollywood studio head, or new athletic director at a major university coming in. He's looking at the current regime (in this case, the "Live.com" branding) and getting ready to kill it - cause it's not his.

The fact that Berkowitz is the head of the something called "Online Services Group", pretty much tells you everything about Microsoft's problem here: bureaucracy. Here's a quote of a man in over his head:

"I'm used to being in companies where I am in a rowboat, and I stick an oar in the water to change direction," said Berkowitz, who ran the Ask Jeeves search engine until Microsoft hired him away in April to run its online services unit. "Now I'm in a cruise ship, and I have to call down, 'Hello, engine room!' Sometimes the connections to the engine room aren't there."

And you're just figuring this out now?

Executive recruiting - it's all about romance, when it should be an audit.

I mean, Senior VP-level guys get enthralled with the recruitment dance, the big $$$ packages, the private jet flights, meeting Gates, etc. I guess the papaya scented towels in the executive washroom distracted him from noticing that Microsoft has over 70,000 employees. This is not even remotely like a row boat OR a cruise ship.

The quotes in the article seem to subconciously suggest that what Berkowitz is looking for is more like a lifeboat...

¶ 12/10/2006

• Scoreboard of predictions for 2006
• Fresh Predictions for next year
• Goals / Resolutions for 2007
• Best of (Something) from last year

Stuff like that... I'll try and fill these canonical posts in pretty soon. Let me know if I'm forgetting an important topic.

¶ 12/10/2006

Having only one charger wouldn't do, so I went to Amazon and found the Nokia AC-3U, which works with the N73, N91, N80, etc. It was about $9 including shipping.

Works just fine. So if you need a US Charger for the Nokia N-Series, get an AC-3U.

¶ 12/08/2006

They had clear boxes of $100 bills in the lobbies, replete with guards. That's a lot of petty cash, about $5M worth.

And yes, Google calculated the appropriate taxes, so everyone's actual bonus was more like $1343 or something. This is all in addition to the regular bonuses that Google employees get (usually between $20k - $30k a year, and up to $800k for founders awards).

I find this a bit too cheesy game-show-like but a bit cool. I'm not a big proponent of the idea that we are in Bubble 2.0, but Google handing out cash is pretty hard to ignore.

¶ 12/08/2006

Yahoo reorgs into three groups. CFO Susan Decker gets operational control of the "Advertiser and Publisher" group. Farzad Nazem, who's been there FOREVER, gets all of technology. And they're gonna bring in someone to run the content side (Audience) group. (Good luck with that, btw).

Who's not on that list? One missing name is Semel golden boy Jeff Weiner, who was responsible for search when Yahoo! Search didn't get fixed. The infamous Lloyd Braun, another Semel guy, has been non-grata for a year and is now gone. And Brad Garlinghouse - who cares?

My initial reaction to the reorg is: Close, but no cigar. Semel isn't the answer. My main reason for this is that he's been too patient with the problem child: search advertising, even though the Overture team is in his backyard down in Pasadena. Fundamentally Terry Semel doesn't have the technical chops to kick the right asses at Yahoo.

Decker is a good choice to run the ad business. But 'Zod in charge of technology isn't really a change of any kind.

Losing Rosensweig is probably not a benefit, though.

Here's what Yahoo should do:

• Leverage the mail user base, revolutionize and market the display ad business on a Google-esque scale. This means opening up the behavioural targeting to a much larger swath of advertisers. Do for behavioural targeting what Amazon did for recommendations. Make it THE CORE and make it huge. Yahoo's the only one that has this.
• Put in an open marketplace. Create an AdWords-like auction market for ad space that knows who's looking at cars or mortgages. I know they have this technology, but they need to make it a $10B revenue stream, up from the $4B it is today.
• Fix search monetization. The Panama rollout is underway. Once that happens, clean out Overture of the people who failed to deliver it 2 years ago. Then make an all out run at Google - develop a way for advertisers to dump entire Google campaigns right into Yahoo. Respond to advertisers better than Google.
• Monetize all the APIs to get the publishers using them. Yahoo innovates in APIs but they are mostly for geeks. Attach monetization to them, so that publishers will build businesses around them, like is done today with AdSense, the Amazon Associates program, the eBay affiliate program and the Shopping.com Partner program. This means: ADD A WAY FOR EVERYONE TO MAKE MONEY when they build with the APIs.
• Go Wide with YPN. NOW! DAMMIT! Before MSN launches their contextual program. This is another inexcusable Overture mess.
• Get serious about distribution deals. Win the next 5 from Google. Focus on entertainment content, news, mobile, games. GET MORE INVENTORY! Areas you already are in. You can't change that Hollywood personality with Semel at the top in any case.
• Companies Yahoo should buy: RightMedia, DoubleClick, AzoogleAds & EfficientFrontier. These are not content companies, but advertising marketplaces they could centralize and leverage.

Can Yahoo do it? I doubt it. I think Semel is the wrong quaterback, and lacks the confidence of the team. I don't think the re-org is radical enough.

¶ 12/05/2006

Here's the billing for November on an AdWords account of mine. Can you spot the funny / cruel joke Google is playing on me?

Is it:

• A) They are crediting me all of Twelve Cents for "Invalid Clicks" on a spend of over $11,000?
• B) They are charging me an extra $0.12 as a click quality adjustment?

Date	Description	Amount(USD)
2-Nov-06	Payment	($582.23)
5-Nov-06	Payment	($560.56)
8-Nov-06	Payment	($601.15)
10-Nov-06	Payment	($547.26)
12-Nov-06	Payment	($551.95)
13-Nov-06	Payment	($557.76)
15-Nov-06	Payment	($574.87)
16-Nov-06	Payment	($548.61)
17-Nov-06	Payment	($567.67)
19-Nov-06	Payment	($637.42)
20-Nov-06	Payment	($766.06)
21-Nov-06	Payment	($555.25)
22-Nov-06	Payment	($567.40)
23-Nov-06	Payment	($660.80)
24-Nov-06	Payment	($657.03)
25-Nov-06	Payment	($672.92)
26-Nov-06	Payment	($668.19)
27-Nov-06	Payment	($654.82)
27-Nov-06	Payment	($502.91)
29-Nov-06	Adjustment - Click Quality	($0.12)
30-Nov-06	Payment	($519.29)

I hope someday Google sends me one of those boxes of schwag to make up for this.

¶ 12/05/2006