My gift to the world, Guy Kawasaki minus the Alltop.

I’m not the only one who is frustrated by how often Guy posts about alltop. Guy’s response to the criticism? “Unfriend me.

It’s nice that he’s honest about not changing, but it flies in the face of everything he says about listening to the customer. With “do as he says, not as he does” in mind, I offer to put his customer’s back in control.

Announcing a twitter bot that retweets everything Guy Kawasaki tweets, as long as it doesn’t contain “alltop.”

Meet the real Guy Kawasaki: http://twitter.com/gk_minusalltop

Caveats:

  1. It’s updated once a minute, so occasionally you’ll be 1 minute behind the over 60,000 other people who want to know what’s going on with Guy.
  2. If Guy complains, Twitter might choose to remove this account. I doubt that will happen since Gal Kawasaki seems to be allowed.
  3. I have no intention of checking any replies to gk_minusalltop. It’s a bot, you can contact me at my twitter account at @whoisgregg or leave a comment below.

Update #1 I fixed the bug where @replies weren’t being posted. Apparently PHP’s implementation of cURL thinks that anything that starts with an “@” is a path to a file. All working now. Amazingly, over 20 followers and Guy noticed it right away. As I told him, I like him and I like Alltop, but too much of the two together is not so good.

Update #2 A bit more explanation of how this came to be… I saw this tweet from Guy the other day where he explains that he sifts through 60K+ followers updates by simply ignoring anything that doesn’t contain “Alltop.” My immediate thought was, “What a great idea! If only everyone who followed Guy could do the exact opposite for his tweets.” Today at lunch I realized I had just recently coded up a twitter bot and with a few modifications could have exactly that feature available. Amazingly people are already following it and at a rate much faster than can be explained by twitter spam bots. ;)

I’m glad I could help. :)

Update #3 @gk_minusalltop has been running 13 days and has attracted 131 followers. The bot updates the Bio with how many alltop tweets it has *not* retweeted. In 13 days, 265 of Guy’s 810 updates have included “Alltop” in them. So, basically a third less tweets and at Guy’s rate, that’s 20 less tweets every day. Definitely a worthwhile project. :)

Update #4 @gk_minusalltop has been running 84 days and has attracted 589 followers. Guy linked to a page about Star Trek four times about a week ago. He was being sneaky by using an adjix link to point to Alltop instead of being upfront about his advertising. Thanks to Leo Notenboom and Ed Palumbo for catching it. The good news? The bot is smart enough now to check if Guy is posting an Alltop link without being honest about it. Should I add a counter for how many times he posts sneaky links? ;)

Update #5 Guy has switched from using Adjix to a new custom short URL service: trkk.us. Unfortunately, this service uses javascript redirects to circumvent the method @gk_minusalltop *was* using to check for sneaky links. I’m working on it, but I can see where this is headed. What a shame. :/

ZOMG 102px tall!
ZOMG 102px tall!
Update #6 It feels like whack-a-mole sometimes. Now Guy is using the URL shortening service om.ly which adds an incredible 102px tall neon gold and orange header around above other site’s content. And, of course, now there’s no hint of Alltop advertising in the tweet’s themselves. So, this is what I’ve done:
  • Automatically removes all ghosttwitter posts (they always advertise for Alltop)
  • Do extra testing for any weird URL shorteners that pop up
  • Hired an army of rabbits trained to recognize the particular shade of orange/gold that Alltop uses to manually check each tweet

Of course, if the rabbits get tired or run out of food, the occasional tweet may slip through. The trick is to tell me as soon as you recognize an Alltop spamvertisement so I can train the rabbits to recognize the new trick. BTW, that screenshot links to an article with this great quote:

The most important lesson that I’ve learned about business through all these years is that the great companies are founded on the desire to make meaning—and not necessarily to make money.

– Guy Kawasaki

Thanks for that, Guy.

Twitter Bots

After reading about Dave Child’s Trivibot (and playing for awhile) I thought, “Wow this whole Twitter Bot thing could be huge! I should do something neat!”

Of course, my disappointment at learning how many Twitter bots are already in existence just underscores how far behind I have fallen after my one year hiatus from tracking industry developments.

The good news? I built and deployed my first Twitter bot within just a few hours of discovering the possibility. Basically it retweets the blog posts from HalfEagle.com. (Of course, I’m doing that because HalfEagle.com is a project of mine… I’m not scraping someone else’s site.)

Next step? Writing an interactive twitter bot. :D

…I just need an idea that isn’t already been done. :/

The joy of having a site go live.

After a 12 hour debug session last Friday, a new site I’ve been working on finally went live.

Working in-house on web projects mostly means I am fixing sites that are already online. Although I often build entirely new sections and features for the site, I rarely get the satisfaction of launching an entirely new site.

I’d almost forgotten how it felt.

Especially considering the short amount of time this site took to get going (compared to my last site which took a year and a half). Although, I have my coworker to thank for that as he took the lead on the more difficult aspects of this new site.

Ecom sites aren’t the coolest projects to work on, but this one was pretty fun, particularly the image zoom feature for individual products. For example, check out this Eagle Scout coin. You can still get the traditional “show larger image” experience by clicking the thumbnails, but if you simply mouse over the main product image you are treated to a much smoother experience.

I don’t know if I could have made this post unless the site had already made a sale so I wasn’t expecting to have the chance until later this week. However, thanks to the miracle of Adwords I had an ad running in about five minutes on Friday night and arrived to work this morning to see the first sale had already been made and shipped.

It’s a good day. :)

You too can have a web crawler and index!

With all the debate raging over SEOmoz’s newest tool Linkscape, the most ridiculous is whether or not SEOmoz has their own crawler/spider. All the word play is confusing. If you haven’t yet, this explanation of Crawling vs. Indexing clarifies the terms involved.

I realized something interesting yesterday, and believe I have found a helpful way of explaining at least one of the communication issues involved. Forgive me for the metaphor, but I am both a huge fan of Law & Order and the son of a man who explained everything through stories.

Here’s the basic plot synopsis of quite a few Law & Order episodes:

John Doe hires a professional hit man to shoot Jane Doe.

Discussion questions:

  1. Who owns the gun?
  2. Who is responsible for Jane Doe’s murder?

The District Attorney doesn’t care who owned the gun or who pulled the trigger. They always go after John Doe, the person who initiated the murder.

Reading Rand’s comments on Sphinn, one thing is pretty clear: He makes no distinction between owning the gun and pulling the trigger himself or hiring a hit man to pull the trigger.

For him, they are one and the same.

What’s entertainingly flawed about this logic is that, if you pay SEOmoz to use Linkscape and you download some of those reports to your computer, you can claim the same accomplishment they have claimed. Because you have paid them for access to their data, and you store your own copy of what they sold you, you too can advertise that you have your own web crawler that has indexed 30 billion pages.

This may be hard to pull of if you do in-house SEO (unless your boss thinks a spare hard drive can fit 30B pages) but if you have SEO clients, it’s all in how you spin it. Just add a page to your site with a friendly picture of a robot:

I CAN HAS LINKZ?

And, you know, keep saying over and over again that you have your own crawler. Good luck! :)