Alex Hochberger

So, back in September, I mused about how AOL could have become the open platform that twitter is, and mused about why they didn’t.  Back in the day, AOL was THE game in town for communication, especially after purchasing the ICQ network.  Failure to integrate the network and failure to open it up, ultimately, propelled competition to their IM Client.  Concerned about losing advertising revenue, AOL wouldn’t allow third party clients with real access to the OSCAR Protocol, instead just a barely supported Talk to OSCAR Protocol.

(NOTE: Back when I was in school, the Unix machines lacked an AOL Client, and the weird TAC Client that ran under Motif.)

Had AOL simply provided a few APIs, perhaps just sending messages (for automation), as well as login and authentication, they’d be where the Open ID type plays are.

I still use AOL IM, but I use it via a LibGAIM Client, Pidgen at the office, AdiumX at home.  I have multiple AIM Names, plus the other networks for those on the others.  Interestingly, I was using IM less and less, but the support for Facebook in Adium recently brought me back.  I think that this is a great move on AOL’s part, as it will help bring AIM back to relevancy as a great communication medium, especially as casual Internet users are all on Facebook more than a computer with an IM client.

Social networking is amongst the hottest topics of the past few years, but while our US Centric media has focused on Twitter (popular in urban areas) and Facebook (popular throughout the western world), the growth of social media is universal.  Here is a country-by-country map of the dominate Interestingly, Google’s Orkut has a small presence in Latin America, and it is likely that Facebook will overtake it in 2010.  This has the added benefit of destroying anyone who thinks that Google’s web dominance in anyway approaches Microsoft’s desktop dominance in the late 1990s.

Most interesting to me, the map looks increasingly like a Cold War map, with NATO/Facebook squaring off against Warsaw Pact/V Kontakte, although the division is less political and more language based.  Proper Cyrillic support is necessary in Russia and areas with high numbers of Russian speakers.  As communication morphs, increasingly email is the dominate business communication and social media is the dominate personal communication method, making sense of personal communications matching personal friendships while business needs the real time communication for which the academic system was developed for.

As communication increases, the world becomes increasingly similar, and a US-centric view of your potential user base may look myopic as more and more of the world starts to look similar.  If you can sell in the US via Facebook, you can sell across Europe via Facebook, as long as you adapt to the social and language conventions of your target country.

So again I open my email to an article about the creepy way of finding Jews to market to on Facebook, this time entitled, Kosher Ham’s Jewdar Tell-All.  A while back, I commented on how I would identify people by ethnicity on Facebook, Ethnic Targeting on Facebook, and I was amused to see an article almost half a year later laying out the things that popped into my head in 10 minutes.  However, there were two things that bothered me in this article:

1. Actually targetting Jews with an advertisement for a “Kosher Ham” business seems really bottom of the barrel, a proud Christian or Muslim wouldn’t target their fellow members of faith with an anti-Christian or anti-Muslim business, would they?
2. They outrageously list the social groups they target, listing a few historically Jewish fraternities and leave out AEPi (Alpha Epsilon Pi), the Jewish fraternity.

In all seriousness, the point they make is that when targeting a group, you aren’t looking for 100% accuracy, just decent accuracy.  The US Jewish population is about 2%, targeting everyone looking for Jewish customers is pointless.  However, online we can develop sub groups with decent Jewish accuracy, and similar options exist for other populations.

For example, approximately 12% of New York City is Jewish, if you target people in New York City, you’ll get more non-Jews than Jews, but you are reaching 6 times more Jews than reaching just the United States.  Further, looking for keywords associated with secular Jewish culture (obviously a business making fun of Jewish dietary law as part of its snark isn’t looking for religious affiliation), you can start to reach a targeted segment.

Now, many, perhaps most Jews on Facebook won’t have ANY interests related to Jewishness on their profile, and if your goal is to reach all Jews, this will fail.  If you were running a political campaign and targeting Jews with a message, only reaching 40%-60% of the Jews on Facebook might not be enough.  However, for running a CPC campaign, it doesn’t matter what percentage of the Jews on Facebook I reach, what matters is what percentage of the people that I reach are Jewish.  Running an online small business doesn’t require “market share,” it requires reaching profitable customers.

If you only target people affiliated with ethnically aligned groups, you might find some segments that are 80% – 95% your target demographic, which means that you’ll get higher click through rates and waste fewer clicks.  People affiliated with those aligned groups, even if not of that ethnicity, might be useful to target anyway.  Someone that frequented Hillel events, even though not Jewish, is probably either a prime target for Jewish-themed T-shirts, either for themselves or as a gift, making your targeting even better.  National ad campaigns are difficult, finding potential customers for a business is much easier.

Every year, more and more of retail takes place online.  The recession is pushing toward more direct response advertising, causing an uptick in online advertising between 3% and 5% in an otherwise down year.  Search is expected to perform similarly, and 2010 is predicting an increase in online advertising dollars between 10% and 15%.  For online publishers and advertisers, this means that despite sluggish growth overall, the Internet continues to grow at several times the overall economy.  Although search engine consolidation has caused bid prices to increases, the prevalence of content networks and alternative search networks creates new opportunities for online advertisers to reach customers in a cost effective manner.

Social Media appears to be one of the interesting potential story, as a source of advertising dollars and business, search continues to grow, display ads are relatively stagnant, and alternative campaigns in new medias inches upward.  Despite it’s status as media darling, Twitter remains a small niche play, with popular Facebook game FarmVille having more users than Twitter.  Email newsletters remain big business, but the movement from getting content pushed into your inbox toward networks and feeds continues, creating new opportunities for online advertisement.  2009 looks to have finished as a slight positive in a down economy, and 2010 looks quite promising.

One of the strangest things about technology is how it moves in circles.  The relational database isn’t new technology, and while many changes to the storage model and the performance of the system has changed, the underlying concept is the same.  The leading databases, except for Oracle, all bare SQL in the name, giving the impression that SQL was critical to the concept of the relational database, not merely a front end language for describing access to relational data.

Web sites fit nicely into a relational model.  They have categories, articles, products, etc., sets of data.  The idea of applying set theory to data is at the core of the relational database.  I can quickly and easily get all Articles in the Category of SEO, because those fields are tagged, and I simply pull the appropriate subset.  You can always get intersections (with JOINs), unions, set deletes (EXCEPT), and other set operations… if you are using sets of data.

Martin Kleppmann asks, on Carsonified, Should you go Beyond Relational Databases? That’s the wrong question to ask.  The question is, “Is your data relational?”  If you have groupings of like data, then you need a relational database.  If you are building an application with non-relational data, then storing it in the database to have a quick id look up is foolish, and you should be looking for persistent data storage that is optimized for that sort of data.

For temporary storage, a system like memcached is perfect, it gives you lightning fast references to data that may only exist temporarily.  For a long term storage, maybe a database is your answer, or maybe you need something more tied to your data structure.  We wouldn’t suggest Microsoft switch from it’s DOC format (and the Docx XML version) to relational databases, but I wouldn’t put relational data into something more object oriented.  You might use objects to represent it in memory for easier programming, but if the data is essentially relational, keep it in relations.

Data structures are at the core of computer science.  With all the free information out there, there is no excuse to be building a large scale system without knowing the basics.  The fact that Twitter built their operation without knowing what they were doing doesn’t mean that everyone can… Bill Gates dropped out of Harvard and made a fortune, not every Harvard drop-out is so successful.

Bing and Google announced deals with Twitter to access and utilize their data, and Search Insider is discussing the first impacts of this.  What’s most interesting is that this might be the first major change since “FreshBot” was added to Google (and later became the primary crawler).  Old hand SEOs remember the crazy update schedules of the early engines, but Google’s monthly “Google Dance” as their crawler finished and about a week later the PageRank was computed and the new index went up across their data centers.  Google started crawling and updating with “fresh” data (tagged with the data) with a guestimated PageRank for placement, and as they got faster at computing changes across the Internet, these Fresh results were no longer being inserted, they were the results.

Twitter has a disproportionate presence in media circles and other influential areas.  Twitter data, including trends, what people are talking about, etc., provides a view into what is new and what is going on.  While news feeds capture the mainstream coverage, Twitter will know what is news to the Internet.  This powerful medium helps determine if you are dealing with a “Google Bomb” or a bona fide story.  While Google originally assumed if people linked to you, like a citation in Academia, that made you authoritative, but only a select subset of the population had websites.  Blogs were more common than a full site, but Twitter is even more available to anyone.  Link and information trading on Twitter happens faster than someone writing a blog post, let along researching a news story, so Twitter gives a view into what is happening now.

This is an exciting time in search.  Twitter data will make it even more exciting.

Not a shocked to anyone that knows anything about search, but it’s nice to see it confirmed by Google, both on the official blog and on Matt Cutt’s blog.  I don’t think I’ve included Meta Keywords on a site in almost 6 years, but from time to time I hear people explain to me that they don’t need an SEO expert, they are doing keywords and descriptions… and I just wish them well.

In what might be huge news, Google is opening the DoubleClick Ad Exchange as an open marketplace.  The Internet has destroyed middlemen and process where ever it has moved, devastating travel agencies, newspapers, recruiters, and in some markets, realtors.  It has been market-by-market, but the value of exclusive information has been destroyed as information has become publicly available.

One exception, media buying.  For reasons that will perhaps never be clear, purchasing banner ads on websites doesn’t follow the open marketplace of the rest of the web, but the rather strange world of media buyers and sellers, with their high commission rates and huge selling costs.  I understand why $10 million ad campaigns might require some sales finesse and price breaks, I don’t understand why I can’t buy 30 second spots on a local independent station in the middle of the night without knowing a salesman and creating an account.  I really don’t know why buying banner ads also seems so convoluted.

When I have done business hosting ads and buying them, I’ve found the results pretty decent if you are targeted.  I believe that they are much cheaper than Text Ads for business reasons… a marketer with a credit card can buy Text Ads, it seems like you need to spend time talking to a salesman to do banners.  Given how many independent marketers are moonlighting on the web, it’s insane to have a medium only available from 9 AM EST to 5 PM PST, when the search engines will let me buy spots at 4 AM.

When my wife and I were discussing social media, and I mentioned that at Third Solutions we use Skype for IM, and at ASG Group we were using Twitter, she asked me if we had gotten old? When I arrived at MIT in 1997, I ran ICQ for friends from home + Zephyr for talking to classmates, the next year the freshmen showed up with big lists of AIM friends, and by the time I left school, ICQ and Zephyr were basically dead and AIM dominated communication.

Here is it, over 10 years later, and I still use AIM as a constant business communication tool, but it certainly lacks any hype or excitement. Gmail accounts form the basis of OpenID, yet AOL with 20 years of AOL accounts and 10-15 years of AIM accounts couldn’t make themselves the login option of choice for the community web or the Web 2.0 world.

An old AOL hand asks, “Could AIM Have Been Twitter?“  AOL fought third party integration, mostly because Microsoft was at the time masters of embrace and extend, and the only on-ramp was the weird open access AOL published for the Tik client that we ran on Unix, with limited access.  While AOL had the users, they didn’t have the culture of centrality.  Openness may have helped, but the open-IM groups pushed by Yahoo and MSN fizzled, Jabber went nowhere, and even though Google via Google Chat supports Jabber, Facebook chat seems more vigorous.

I think that AOL could have done a lot with their platform.  But the corporate culture, more than the business around technology, prevented them from being cool.  Everything AOL bought saw talent flee to start-ups and generally fall apart.  Other than picking up Time Warner for a steal, they weren’t able to use their early lead in the Internet, perhaps because of their Internet for the Masses reputation, they couldn’t be “cool” to the technologists, so even if the masses used AIM, nobody was building upon AIM.  That, more than AOL’s internal walled garden mentality, is why AIM didn’t become Twitter.

So something nice about a real database is cascading values, most commonly used to deletion, but you can use them for updates as well.  Let me give you a scenario: I track groups of data from my clients by a sub_id field.  As they add groups, their ids aren’t in a range.  If I wanted to consolidate them, I could update the sub_id (somewhere that won’t be stomped on by the sequence), but what about historical data.  Baring a cascade rule, PostgresSQL will stop the update.  A non-relational database like MySQL will just leave orphan foreign keys.  If you set Cascading, they come along for the ride.

View the following example:

CREATE TABLE A (a_id SERIAL PRIMARY KEY, letter char);
CREATE TABLE B (b_id SERIAL PRIMARY KEY,
a_id integer REFERENCES A(a_id) ON DELETE CASCADE ON UPDATE CASCADE,
letter char);

SELECT * FROM B;
b_id | a_id | letter
------+------+--------
1 | 4 | a
2 | 5 | b
3 | 6 | c


So we were able to renumber A, pass the values to B, without any updates to B. Pretty cool, huh?