Wednesday, December 16, 2009

Investing in Assets to Boost Organic SEO

Frequent reader of this blog (Hi Mom!) will know that I place a lot of value on domain names. In Decembers past, I have made end-of-tax-year "big-ticket" (ie low 4-figure) domain name purchases. This year is different. That's partly because 3 sales I pursued fell through (1 person wanted $200k for a domain that I was willing to pay $1500 for; Sedo had another domain listed, but it turned out that the person listing it had already to sold it to someone else; and Rick Latona advertised a name in "Daily Domains", but it turns out that they didn't bother to check whether the person selling it actually owned it.)

But there's a bigger trend going on. Despite some really large domain sales this year, domains as an asset class have substantially declined in value. The biggest factor is Google's decision to dramatically cut payments for parked traffic. Perhaps that was a strategic decision on Google's part -- they certainly have no interest in propping up the business models of (competing) type-in traffic or spammy "made-for-adsense" companies. Perhaps it simply reflects advertisers taking advantage of the ability to segment parked traffic out from search traffic, and deciding that search traffic is more valuable. Either way, it's harder to justify buying domains when the opportunity to earn parking revenue has evaporated.

A second factor is that (at least from the tests that I've run) "mini-sites" are a complete and utter failure. I purchased 3 mini-sites from AEIOU in 2008 (, and, promoted them for several months in 2009, and then watched to see what happened. The answer, unfortunately, was "nothing". When AEIOU announced that they were no longer supporting this business, I gave the domain to, moved to my own server, and I still haven't figured out what to do with

Where does that leave domain investing? It seems to me that there are really only two options left -- "buying to flip" (at which I have failed repeatedly and miserably) and building domains into real businesses. While I am a big fan of the latter strategy, it doesn't scale! Buying a new domain name makes your job harder instead of easier -- you now have to develop content, functionality and links for more sites! So, at least in the short term, you're worse off than you were before you bought the latest domain.

Domain owners: wake up! Unless you have a truly premium domain, it lost half of its value in 2009, and will lose more in 2010. End users have less money to spend, and fewer ways to make money on the domains that they do buy.

Domain buyers: wake up! There are higher value assets to buy for organic SEO than domains. Invest your money in real content, real functionality, better business relationships, and assets that make it easier to attract editorially-chosen links. Although these assets are more difficult to manage than domains, they are (much!) more valuable.

Thursday, December 03, 2009

More signals from Google that Page Load time will affect SERPs

Just a quick note -- there are a couple of recent signals that Google will start to use site response times to affect how well you rank in their search results in 2010.

The first is found in a summary of a recent PubCon session with Matt Cutts:

He [Matt Cutts] hints that people at Google really want to use site speed as a factor in rankings. They’re not using it right now, but they want to be. They want search to be like a magazine. Google wants to make the Web fast. 2010 is a great time to pay attention to speeding up your site. HINT. HINT.

The second is that Google Webmaster Tools now offers a page called "Site performance" which gives you feedback on how fast your site loads relative to other sites on the web. (Announcement is here.)

Have you thought about how quickly your site loads? Probably a good time to take it seriously.

Thursday, November 19, 2009

Bad experience with 99Designs -- my takeaways

I'm coming off an unhappy website redesign with -- only 2 or 3 of the 13 designers submitted designs that seemed worthy of iterating on. This contrasted sharply with my previous experience having a logo designed at 99Designs when I had nearly 350 designs submitted (compared to 37), and 4 of them were very good. This time around I have 1 design that I'm pretty happy with (and the overall hit to my wallet is twice as high).

Part of the explanation appears to be that many more designers submit designs for logos -- they're probably much less work -- and payouts tend to be about half the payout for a website design, so designers opt for the logo projects.

One approach to designing a successful project:
If you sort the webpage designs by the number of designs that were submitted, you can find projects that received more designs than projects that offered to pay more. Sometimes these high-marketing / low-cost projects beat out projects offering to pay nearly 3x as much! These disproportionately successful projects suggest a few takeaways:

  • Tell designers that there's a possibility of ongoing work in the 2nd line of the project description

  • Projects that get lots of designs often offer a high payout but don't guarantee the project. Some of the projects that do guarantee a payout didn't get many designs. There doesn't seem to be a clear association between guaranteeing a project payout and the number of designs submitted.

  • Making the project "blind" -- ie designers can't see what other designers have submitted -- doesn't appear to be correlated with the number of design submissions.

Here's what I'm going to do in the future:

  1. Make an effort to "sell" the project, both by describing how tricky a project it is, and by approaching designers directly and asking them to take a crack at it (see the tips in the links at the bottom of this post).

  2. Don't buy separate logo and webpage design projects for the same website.

  3. Give feedback to everyone, even if you only give personalized feedback to the folks whose designs you like. My guess is that designers look at whether or not a contest-buyer gives feedback to everyone before they decide which sites to design

  4. Avoid the "blind contest" option like the plague. I understand that designers love it, because it prevents less talented designers from stealing their best ideas. However, as the buyer, I kind of like designers riffing on other peoples designs. You probably get more designs overall, and more of them are going to be appealing. Plus, (granted with only two data points) I had about 1/10th the number of designs submitted in my "blind" contest.

Other tips for 99Designs Contests:

  1. 10 Tips for Obtaining a Stellar Graphic Design via

  2. tips: crowdsourcing a design project

  3. 4 tips to hosting a fun design competition

Monday, November 02, 2009

Using "Outposts" in your SEO Strategy

I attended the Cambridge SEO Meetup again tonight, and was once again reminded why I go: there are some great people that attend. Props to Derek Edmond for giving me a great link-building idea, and thanks to for bringing food!

Tonight's speaker was Stuart Foster. He has an interesting resume and it sounds like he's done some genuine legwork to create a meaningful following on Twitter. However, I'm iffy on his SEO recommendations, and I disagree pretty strongly with one of them. He recommended that people follow a strategy that he credited to Chris Brogan: that you "syndicate" the articles that you write to 5 or 6 "outposts", like Facebook, a blog, etc, and that you then funnel links from those "outposts" back to your main money-making site.

Now, I first heard the name "Chris Brogan" about two or three years ago when he came to speak at one of the Cambridge SEO meetups. I later heard him speak at Affiliate Summit in Boston after his star had ascended a bit, and I have to say that he came across as a genuinely decent human being. And further, he does appear to have written a post titled: "Using Outposts in Your Media Strategy" which uses a lot of the same words and concepts that Stuart used. But, for Chris' sake, I'm going to assume that this strategy makes more sense for personal branding than it does for SEO -- because it's a really bad idea for SEO.

Here's the idea behind using "outposts": if you set up 5 or 6 mini-sites on different hosts (maybe a Facebook page, a MySpace page, your LinkedIn account, etc) then you can point all of those sites to your main (money-making) site, and Presto! -- instant inbound links | PageRank | link juice. You'll have the magic of Facebook's PageRank 11 site to push your money-making site up in Google's rankings.

Here's the problem: If you want this to actually work, you now have 5 mini-sites plus your money-making site that you need to update. Let me say that differently: instead of simplifying the work it takes to get inbound links, you have multiplied it six-fold!

Let's say that your brand-spanking new Facebook page has exactly one article on it, plus a link to your money-making site. Guess how many visitors it will get per month if you don't promote it? 3,000? 1,200? 8? Nope. Zero. The only way to get people to visit your Facebook page is to go out and promote the page.

Make no mistake -- getting editorially chosen links is really, really hard. Why in the world would you want to increase your link-building workload by a factor of six? Especially when the link juice passed from 5 of those 6 sites doesn't flow directly to your money-making site? Insanity!

Take all that hard-earned content and publish it on your money-making site. Spend all the time you would tell your friends and acquaintances about your MySpace page, and tell them about your money-making site instead. Stop tweeting. Stop retweeting. Write more content; promote that content. Lather. Rinse. Repeat.

Tuesday, October 20, 2009

The NOLO Consultant & Independent Contractor Agreements book rocks

Just a quick note to folks out there that hire independent contractors. You should run, not walk, to your nearest web browser and order a copy of Nolo's Consultant and Independent Contractor Agreements book. It comes with a CD with stock contracts, and excellent explanations of each of the clauses within the contracts. I use it at least monthly.

(Those of you who know me know that I had a situation back in the Spring where I thought an independent contractor was trying to get their contracting work for me re-classified as an employee to get benefits after losing his full-time job. In the end, it just turned out to a mistake on his part, and he cleaned everything up. However, I now use this book with every new contractor I hire to be sure that I have good contracts in place up-front. $34.99 well spent!

Tuesday, September 01, 2009

What Great Long-Tail Organic Pages Look Like

I attended tonight's Cambridge SEO Meetup, and left the meeting somewhat frustrated. The speaker, Chris Baggott presented blogging software from his company, Compendium Blogware. Chris was articulate and well-informed, and dealt gracefully and with humor while being challenged by several people who seemed intent on proving that he isn't a Linux sysadmin -- which he's not.

That aside, here's what frustrated me: Chris was presenting a solution that really bothered me. He was recommending that clients use his company's blogging software to build web pages that target long tail keywords. That in-and-of-itself is laudable. It's relatively easy to rank for long-tail keywords, and a lot of high-volume "fat-head" keywords don't convert that well. His software has some proprietary sauce that helps to categorize those posts in several different categories, and post them with keyword-enhanced titles and using keyword-enhanced directory names. So, for example, the page might be:

And the resulting page would have the title "Four Slice Cuisinart Toaster". Aside from these basic SEO optimization techniques, he encourages his clients to write a 100-word blog post on what they're doing that day that relates to four-slice cuisinart toasters. I'm imagining posts that say something like: "My family ate four slices of sourdough toast this morning from our new Cuisinart toaster."

Compendium will then also helpfully categorize that post under: and

The result is made for Google psuedo-spam that takes very little effort by the end user, is easy to update on a regular basis, and probably ranks pretty well for queries involving these keywords. (Indeed, just several weeks after registering a $10 domain and pushing 150 posts live, he claims that they have sold 20 toasters.)

Why does psuedo-spam like this work?

It seems to me that when Google visualizes the web, it sees a vast topographical map. At the center of the map are commercial terms like "mortgage", "gambling", "viagra" and "insurance". And surrounding these keywords are vast mountains of pages that are attempting to optimize for those keywords. The 10 pages that make up the very peak of this mountain are displayed to the end user when they type the query "mortgage". Surrounding that peak are lesser peaks, such as "online mortgage", "second mortgage", "quick mortgage", and foothills such as "san diego mortgage", "interest-only mortgage new york". Within each of those lesser peaks and foothills there are 10 pages that fit the Google algorythm well enough to become local maxima.

Compendium's brand of psuedo-spam works because out beyond the mountains of high-volume, highly-commercial keywords, there are vast plains, with 4 and 5 word keywords that very few people search for. Because so few people search for them, few marketers optimize for them. And because few marketers optimize for them, it takes little work to create the sub-surface geology that gives these pages a moderately higher peak than the plains around them. By choosing a domain name with the "toasters" keyword, highly-targeted sub-directory names and titles, and a bit of text that hasn't been translated from English to Russian and back again, Compendium's clients gain enough altitude to rank for their long-tail keywords.

So, what's wrong with that? I guess nothing. It's certainly better than the true spam that's out there -- stuff scraped from RSS feeds, and then churned through an automated content re-writer and spewed onto the web. Stuff that almost reads as English, until you realized that the last few sentences you processed don't actually make sense. I mean, from one perspective, I'm glad that Compendium's software works. Certainly reading about what toaster people used for their breakfast is better than reading true spam. However, the problem is that it doesn't take that much additional work to create truly useful long-tail organic pages. In fact (in volume), it may even be easier.

Start by asking yourself: When someone types 'Four Slice Cuisinart Toaster' into Google, what are they hoping to find? What are the implicit questions that they're asking?

  1. They probably want to know what models are available, how much they cost, and where they can buy them

  2. They probably want to know what product features they have

  3. They probably want to see pictures

  4. They might want to know where the user manual is, or troubleshooting tips

  5. They might want to see user reviews

Why not create that page for these users? If you're going to create 250 pages (which Chris recommended), it's probably faster to collect the data in a spreadsheet, and then have someone spit out that data into a web page. Include a link to a "Consumer Review" page, and then roll those reviews back into the page. The upside of this approach is that you have a page that genuinely ought to rank #1 in Google. The better it ranks, the more people will come and submit reviews, which creates a virtuous cycle.

Now, over time perhaps Chris's clients will create these higher-value pages that offer more substance to their readers. Or maybe domainers will adopt Compendium (or one of the the other similar CMSs) to monetize their middling domains. And maybe that shouldn't bother me -- I've certainly got one or two projects out there that I'm not really proud of. But my personal bias is that two-or-three years out you'll reap the rewards of building a platform that results in great pages, rather than pseudo-spam.

Update: Here's the danger of writing blog posts at 2:00 am. I slept on this and decided that I was being an idiot. Compendium is a much better solution than just ignoring your long-tail keywords, and clients certainly have the option to make those pages genuinely useful. The "pseudo-spam" aspect of it has nothing to do with Compendium's product -- just what I imagine most people would enter as blog posts.

Wednesday, August 26, 2009

PoliticalCalculations just did an excellent Cavalcade of Risk

Just a quick pointer. Ironman from Political Calculations just did an excellent job rating and ranking the 30 submissions to this week's Cavalcade of Risk. (Of course, I may be biased -- my submission on home insurance and hurricanes was tied for the highest rating that he handed out).

Friday, August 14, 2009

Favorite Techniques for Buying Domain Names

I've written in the past about the value of domain names. I thought I'd write a brief summary of some of the methods of buying domain names that I've found to be most successful.

1. Approaching domain owners directly

The best names I've ever bought I found by approaching the owners directly. and are 2 examples. Here's the story of a 3rd:

I recently decided that I wanted to buy a better domain name for my life insurance website. The old name was, and I felt like it detracted from the site's credibility. I generated a list of what I thought were the 20 best generic keyword names related to life insurance. (ie:,,,,, etc.) I then checked each domain, and contacted all of the owners that hadn't built actual sites on their domain. Contacting the owner is often as simple as checking the record domain name at and sending an email to the administrative contact.

Given that life insurance is such a competitive niche, many of the names were unavailable. Prices on some of the available names were as high as $100,000. However, the owner of was willing to talk in a couple of months, if the deal he was working on at the time fell through. Luckily for me, it did, and we were able to agree on a price.


NameJet is a drop-catching service. That is, when a domain owner allows a domain to expire the name "drops", and a number of services (such as NameJet, and compete to register the name and then resell it. However, NameJet also has exclusive relationships with a number of registrars that guarantees that Pool and SnapNames can't compete to catch names. That means that many of the best names are only available through NameJet. I've purchased 8 names through NameJet in the last 2 years, including, and a third excellent name that I'm not going to mention here.

The trick with NameJet, I think, is to be patient and selective. Spend 10 minutes every day checking the service, and don't buy names of marginal value. (For example, one of the 8 names I bought was -- what in the world was I thinking?) If you want to buy names of marginal value, I'd recommend a different method.

3. Buying names for Reg Fee

"Reg fee" means that you pay only the registration cost of a name -- somewhere around $7 - $8. The trick for buying names for reg fee is simply to generate a large number of names (either via scripts or just through keyword munging in Excel). Then run 10,000 names through a Bulk Registration tool (like GoDaddy's bulk reg tool).

The advantage of this approach is that you can develop a relatively large portfolio at low cost. I've sold three "reg fee" names in the past few months for between $300 - $450 apiece.

However, the disadvantage is that you may be paying reg fee on 200 names a year to sell 3 or 4 names. If you don't have a repeatable method of selling names, then you may have a hard time making a profit, even if you bought the names for cheap.

The trick for all three of these approaches is not to fall in love with a particular domain name. If you must own a particular name, you're going to pay a premium for it. If you simply need a good name in a particular vertical, you can generally find one that represents good value. (If you simply want a good commercial name, and you don't care what niche it's in you can probably get a steal!)

Friday, July 17, 2009

New Look and Feel for

Just launched a new look and feel for I'd love to hear feedback or suggestions!

Thursday, July 16, 2009

Renaming to

I have been developing a life insurance site called since 2006. I'm excited to announce that I have now renamed the site to more accurately reflect the site's focus on helping consumers choose the most appropriate term life insurance policy.

We've done some really interesting consumer-related research on life insurance, including topics such as:

  • Should you request multiple life insurance quotes? (link)

  • Why do insurance quotes from different companies vary so dramatically? (link)

  • What to do if you have given inaccurate information to your Life Insurance Company (link)

The site also profiles more than 60 life insurance companies, policies in each state and more. If you're in the market for life insurance, please take a look (and let me know if we're missing anything!) -- great end-user domain

I always like to see great domain names put to use by end users (as opposed to domainers), so I wanted to give a shout out to John Sowden. John works for (an online appliance parts retailer) as their VP of something, and he's been really helpful with a couple of articles that we've published.

He pinged me last month to say that he had launched a site for his appliance repair show called (appropriately enough): John's been working in the appliance industry for more than 2 decades and he really knows his stuff. If you're interested in radio along the lines of "Car Talk for Appliances" check it out.

By the way, from a domainer's point of view is a fantastic name. My stats show that people are much more likely to type-in a domain name that's "" than one that's "" (ie, and my guess is that that gives "NounVerb" domains a bit of a boost in search engine results as well (the domain name is one of the few bits of data that the user gets to see before they click).

Wednesday, June 17, 2009

Affiliate Sites in Highly Competitive Niches

One of the things that I've "learned" in the past couple of years of building websites is that some niches have a hell of a lot more competition than others. Life Insurance, for example, is just a tough niche to break into. My takeaway from that is that it probably makes more sense to be a big fish in a small pond that a small fish in a big pond.

However, this article makes a couple of interesting points while encouraging people to just into "big niches":

  • More offers to split-test
  • Higher payout due to more competition
  • Longevity -- these niches aren't going anywhere

Interesting thoughts.

Friday, June 12, 2009

Domain Names as Force Multipliers for Websites

My day job involves developing and promoting my own small stable of online businesses. However, on a daily basis I devote time to domaining -- finding good domain names to buy. One thing that I have banged my head against in the domaining community on a regular basis is (what I believe to be) a substantial misunderstanding of what makes domain names valuable to online businesses. A domain name's true value lies in its value as a force multiplier. Independent of a domain owner applying substantial force to the domain, it has relatively little value at all.

First, I should make a few caveats:

  • I buy a handful of domain names each month, so I believe domain names are both useful and valuable.

  • There are a couple of classes of domain names that are valuable independent of their value as force multipliers. For example, misspellings of popular generic domain names aren't really useful to develop a site on, but they may generate substantial and valuable traffic. Alternately, short commercial .com names (, and various adult-themed names have been in the news this year) attract substantial easy-to-convert type-in traffic. And there are all sorts of stories about domainers building 5-6 figure a month revenue streams on these sorts of domains.

Here's an example of the misunderstanding that I see often:
"The domain name itself takes care of something like 50% of the SEO work" (from a comment on I've had other conversations with domainers who have argued that "the domain is the biggest piece of the puzzle" when it comes to building sites.

I don't buy it -- and more importantly, I've never seen a result remotely similar to that. Building websites into effective businesses takes a tremendous amount of work, experimentation, time, patience and risk. Good domains are a useful starting point, but it takes a lot of work to grow a site from "registration fee" money (which the vast majority of domains that people buy and sell every day never make), to "car payment" money to "work for yourself" money. In fact, what strikes me most about the domains that I own is how little type-in traffic there is out there relative to the traffic available from search engines.

(As one data point, I own a single-word commercial .com domain that I have been developing for a bit less than 1 year. Type-in traffic represents less than 5% of the site's traffic; search engine traffic > 90%. And true type-in traffic -- navigation to the site from folks that have never heard of it -- will continue to decline as a percentage of total traffic.)

Domains as Force Multipliers

A "force multiplier" in military speak is a factor that makes a group of soldiers dramatically more effective: for example, the ability of precision-guided bombs to dramatically change the outcome of a war. (thanks Wikipedia) I think that's a good analogy for the utility of most premium domain names.

A domain name is useful for several things:
1. You'll get a small amount of type-in traffic -- people that think "Hey, I want to buy some candy -- I bet there's a website at that sells candy."
2. Search Engines give a bump to sites that have content related to the generic terms in their domain name. I don’t think they do this because it makes it easier for search engines to categorize the site — I think it’s because end users are more likely to trust the “brand”.
3. If you’re building a site, it’s easier to get on the phone and explain to someone that you own “” than “”. The person on the other end of the line is more likely to believe that you’re serious with the former domain than with the latter, whether you’re trying to get a link, become an affiliate, sell them something, etc.
4. The domain name — along with your title tag and some text the search engine scraped from your site — appears in the search results. End users trust good generic domain names, so all else held constant, they’re more likely to click on your link.

But, with the exception of type-in traffic, none of these benefits help you unless you help yourself. That is, if you do a lot of work, a good domain name can amplify your efforts. But if you don't do the work, don't expect the domain name to do it for you.

How big of a force multiplier is a good domain?

Tough question, and one that I can't do more than speculate on. However, I think there's a good case to be made that the more competitive the niche you're trying to break into the less useful premium domain names are.

1. In competitive areas, where much of your competition has a good domain name (perhaps a branded domain name), don't expect to see much of a bump because of your domain name. (One industry that comes to mind is life insurance.) There's lots of great content out there and people have been consistently building links for a decade. There's a lot of offline advertising and people trust brands that they know more than generic names. In this scenario, the search engines have lots and lots of trusted content for the most profitable queries. My guess is that the multiplier effect is minimal -- low single-digit percentile. Given that domain names in these niches are expensive, you may be better off coming up with a brandable $7 domain name (ie

2. In areas with middling competition, I think a good generic name can be quite useful. I'm guessing that it can give your efforts as much as a 20% bump. Maybe it becomes easier to attract links than it otherwise would have been; Maybe it's easier to find partners, or to negotiate a better payout from those partners because they took you more seriously; Maybe a higher percentage of people clicked on your links in the search results because they assumed that such a great domain name would be coupled with useful content and features.

Other Force Multipliers that You Can Buy

There are, of course, other force multipliers for your business that you could buy instead of a domain name.

  • Editorially-chose inbound links are probably the single most powerful force multiplier. While I don't buy (or recommend buying) links directly, I certainly invest in techniques and assets that increase my chances of acquiring such links. It's hard to overestimate the degree to which a single good link can multiply the rest of your work.

  • A technology platform that makes it easier for you to develop content and features that help your end users.

  • Analytics tools and consulting that help you understand what your users want.

  • Increased traffic can itself be used as a force multiplier if it helps you better understand how to convert your users.

Names matter, not least for the simple reason that humans seem to be wired to remember them (present company excluded). But before you go out and spend 5 figures for a domain name do some deep thinking about whether that name will pay back your investment within a reasonable time horizon. And whether it will multiply your efforts more than some of the other assets mentioned above.

Wednesday, May 27, 2009

Almost caught by a clever domaining scam

Kudos to the folks at -- they just helped me avoid getting caught in a clever domaining scam. Someone wrote me (email reproduced below) and said that they wanted to buy, and that they would pay 65% of the appraised value of the domain name. Then they recommended 3 companies whose appraisals they would accept: Sedo, Podzz or

I forked over $29 to Sedo to do an appraisal of the name, and wrote the folks back saying "Sure, I'm happy to try it".

Sedo contacted me the next day saying "We believe the email is an attempt to get people to order an appraisal from After purchasing an appraisal, the domain holder will not receive a response from the person again."

Anyway, Sedo then offered to refund my appraisal price, which I think is tremendously good customer service. Thanks Sedo!

Here's the original email expressing interest in the domain name:

Dear Sir or Madam,

We are interested to buy your domain name BATH-CABINET.COM and offer to buy it from you for 65% of the appraised market value.

As of now we accept appraisals from either one of the following leading appraisal companies:

If you already have an appraisal please forward it to us.

As soon as we have received your appraisal we will send you our payment (we use Paypal for amounts less than $2,000 and for amounts above $2,000) as well as further instructions on how to complete the transfer of the domain name.

We appreciate your business,

Yours truly,

J. Harris

Tuesday, March 31, 2009

Can Google Analytics and Google Website Optimizer play nicely together? (Part I)

Over the past couple of years, I've run both Google Analytics (GA) and Google Website Optimizer (GWO) on the same site. The problem is that out of the box, if you add a GWO experiment to a set of pages, GA will stop reporting correctly about those pages. (The situation that I run into is that I add a GWO experiment to one of my Goals -- say a lead gen funnel -- and then the GWO stats are correct. But GA stops reporting on my goal funnel. Quite an annoying situation.) Given that I'm fundamentally lazy (in a bad way) I haven't really tried to sort this out (until now).

First things first -- does Google realize that there's a problem, and do they think it can be solved? Turn out the answer is "Yes" and "Yes".

How to fix it. Google gives instructions here. Those instructions look like they're a bit outdated, as they include a call to secondTracker._initData() -- and GA / GWO doesn't seem to use that function call anymore.

But, it looks like the change is pretty straightforward -- all your GA code is going to be using:

var pageTracker = _gat._getTracker("UA-xxxx-xx");

While all the GWO code is going to be using:

var secondTracker = _gat._getTracker("UA-yyyy-yy");
-- or --

So, I've made the changes -- I'll have to check in tomorrow to see if they actually worked!

Update 6/20/09: By the way, this solution works perfectly.

Tuesday, March 17, 2009

I need some Google Analytics Help!

A few weeks ago, one of the folks that I drive affiliate traffic for mentioned that another affiliate had seen improved revenue by running a 300x250 ad for their affiliate offer in place of an AdSense ad. I said I'd test it, and within a few days I was running a bright and shiny new clickable image. Now I must admit that I was lazy, and I ran the ad without really thinking through how I would tell which ad performed better. I was just hoping that the results would be so dramatic and obvious that I wouldn't be required to actually, you know, think. However, I did go so far as to add a JavaScript call (pageTracker._trackPageview('/affiliate/click1')) to the image that would record the click with Google Analytics each time someone clicked the image.

Now that I've dug around on the web a bit, I have a strategy for comparing the revenue generated by the two, and I'm looking for some feedback on this approach.

Basic Methodology

I'm going to write a cron job that switches the ads every 5 minutes. So, the AdSense ad will be displayed half the time, and the affiliate ad will be displayed half the time. (I could display the ads randomly on every pageview, but I can't think of any bias that this simpler approach introduces.)

Calculating AdSense Revenue

The AdSense ad is its own channel within AdSense, so I can slice and dice the data by channel and date.

Calculating Affiliate Revenue

This is a bit trickier. Basically, I'm setting up an "advanced segment" within Google Analytics for everyone that views the page '/affiliate/click1'. Then, if I segment "Goals -> Total Conversions" for the date range that I'm comparing to AdSense, I can get the total number of conversions. By multiplying the conversions by the revenue per conversion, I can get a dollar value for the affiliate ad.

If the two ads are alternated for a week, at the end of that week I should have a dollar amount for AdSense and a dollar amount for the Affiliate revenue. Whichever is higher, wins, no?

Potential problems

I'm assuming that by alternating the ads every 5 minutes, I will serve about as many of one ad as another. If it turns out that in fact, I've served 10x the number of affiliate ads as Adsense ads, then the results would be biased. Is there a straightforward way within GAnalytics to show the number of times an image is displayed? (An onLoad event for the image perhaps?)

One of the issues that comes to mind is that people may be more likely (or perhaps less likely!) to click on an ad if they see it multiple times. So it might be that the very act of alternating between AdSense and an image ad increases (or decreases) the number of conversions. So, I think ideally this would be further segmented: A group that sees only AdSense ads, a group that sees only Affiliate ads, and a group that sees both (if they request multiple pages from the site). But I can't think of a straightforward way to implement this (absent a cookie infrastructure that I'm not currently willing to invest in). Any thoughts?

This feels sort of kludgy -- is there a more elegant way to do this from within Google Analytics?

Thursday, February 26, 2009

Why Geeks Suck at Organic SEO

Or, Why Google Loves Links

I have a t-shirt that has #!/usr/bin/perl on the front, and a bunch of obfuscated perl code on the back. If you were to type that code into a file, and submit it to a perl interpreter, it would print: "The three chief virtues of a programmer are: Laziness, Impatience and Hubris". Now I'm as much of a fan of automation and a willingness to boil reluctant oceans as the next person, but such virtues don't apply to everything.

Success with organic SEO depends on two things: (1) Good long-tail content; and (2) Good inbound links. People say other things matter, but they're wrong.

Geeks excel at the first task: generating long-tail content. That's because content generation scales. You can create a technology platform that makes it relatively easy to published edited articles; you can integrate databases with templates to create endless variations on subtly different keyword combinations. If you're a fan of generating textual crap you can scrape RSS feeds, use a rewriting script to munge the text and publish limitless text with the push of a button.

But, Geeks don't excel at building good inbound links: acquiring inbound links doesn't scale, and hubris is rarely your friend. It requires repetitive work that isn't easy to automate, such as making phone calls and developing relationships.

This, of course, makes the quality and quantity of inbound links an ideal criteria to keep the world's most effusive content generators from spamming Google's index. Look at how Google reacts to link acquisition strategies that do scale: link-buying (removing pagerank of participating sites, requesting that people report purchased links), links in comments / forum posts / wikis (devaluing IBL, promoting the nofollow tag) not to mention older link-building techniques such as massive link farms.

Geeks suck at organic SEO because Google has to make organic SEO depend on something Geeks suck at. Or, said differently, if Geeks were great at building good organic inbound links, Google would have to put its algorithmic weight behind factors that Geeks suck at.

Wednesday, February 18, 2009

Charles & Hudson: Best Home Improvement Blog

I just thought I'd give props to my favorite home improvement blog: Timothy Dahl's Charles and Hudson. As you know if you read my blog regularly, I work in the home improvement industry, and Timothy's blog is the best I've found.

He has a long history in writing about home improvement issues, and you can expect to read posts that cover tools (here's my favorite), DIY news and new products. If you use a RSS reader, you subscribe to his RSS feed here.

Thursday, January 29, 2009

New experiment in selling domain names

I've written in the past (here and here) about my attempts to sell keyword-rich, but typein-traffic-poor domains. I'm going to try again.

I'm going to sell the domain name via a reverse auction, using Twitter as the platform. Starting at 8:00 am CST tomorrow, I'll twitter the initial price of the domain on my Twitter account (you can follow me here: The initial price will be less than $100. If you want to buy the domain, just reply (publicly) to my tweet with the word "SOLD", and we'll work out the payment details (via PayPal). The first reply with the word "SOLD" will win the domain.

If the domain doesn't sell at the initial price, I'll drop the price a few hours later -- and continue dropping the price during the day.

The name I'm going to sell is: Google's AdWords keyword tool show 246,000 queries for the term "curtain rod" per month, with a clicks costing $1.33 apiece. Given that consumers are holding off on larger home improvement purchases, my guess is that they're shifting their spending to less expensive repairs and improvements around the house. (Indeed, Google's December traffic numbers for "curtain rod" show 301,000 queries, rather than the average of 246k queries.)

Generic keyword domains are great tools for ranking on the keyword in question, so you could easily tie this domain name with content from a service like and potentially develop a valuable site for a very small investment.

If you have questions about the auction or the domain, please feel free to leave a comment, or email me at: madbury at gmail.

Monday, January 26, 2009

What does a "natural" distribution of links look like?

From time to time I see articles claiming that link-building campaigns should try to acquire a "natural" distribution of inbound links. Sometimes people think you should have links from sites of different page ranks ( used to claim this). Other times people talk about the distribution of anchor text for your inbound links. (Here's a SearchEngineWatch article that claims this.)

If you're out there doing weird artificial stuff with inbound links (like buying them, or link farms) then I agree -- you should worry about having a "natural" link distribution.

However, the problem is, you and I have no idea what a "natural" link distribution looks like.

When I used to run, people would often link to the site with the link-text "Talkr" or "".

When I linked to the SearchEngineWatch a couple of paragraphs ago, it was much more natural for the anchor text to say "article" than "anchor text distribution article".

For the last few great feature articles that I've published, I've gotten a few links from very popular high-pagerank sites that made the editorial decision that it was a good story. But no links from little sites.

So what does a "natural" link distribution look like? You and I don't know. And I doubt that Google knows.

How would Google build up a corpus of websites that only had editorially chosen links, against which they could train their algorithm? (Perhaps it would be easier to build a corpus of websites that had spammy inbound links -- but the problem there is that good sides with spammy inbound links would also attract natural links, because they would tend to rise in the search results.)

Probably, there's a distribution of distributions. And it may be quite possible for Google to spot dramatically-manufactured link distributions.

But the irony is that there's a good chance that manually trying to tweak your anchor text or the PR of your inbound links results in easy-to-spot unnatural distributions like:
  • Buying links based on pagerank formulas (ie buy x PR 5 links, 2x PR4 links, 3x PR3 links, etc.)
  • Systematically varied anchor text (ie 20% says "mortgage leads", 20% says "mortgage quotes", 20% says "mortgages", 20% says "buy mortage", 20% says "")
My guess is that the easiest way to make your link distribution stand out as manufactured is to manually try to vary it.

What do you think? Am I an idiot, or is this spot-on?

Tuesday, January 13, 2009

One reason fighting paid links is so difficult

I contacted a website today with a pointer to a great new article that I just published. The writer that wrote the article probably spent more than 20 hours on it. I spent another 10-15 hours, and the editor that worked on it spend 3-5. It's a 1600+ word article that nails an important commercial topic, and it's a very strong piece of work.

As part of promoting the article, I contacted a number of websites that are interested in the vertical that this article covers. One of the replies that I received said:

"We normally charge $25 to include links, but I've waived it in this case and added a link to the article to"

Has Google flagged in the past for selling links?

In my opinion, this link should pass PageRank. I think Google should think that it should pass PageRank. Yet there's simply no way (unless they're parsing Gmail messages with some very sophisticated algorithms) that they could attain a sophisticated enough understanding of my site's relationship with to correctly pass PageRank on this post, but not the ones that are sold for $25.

The reality is that there's a chance that this link will pass PageRank (ie Google hasn't penalized them for selling links in the past). Or, as I understand how Google addresses paid links, the link simply won't hurt me. I suppose the net takeaway is that I'm glad I don't have Google's job!

PPC is a Dead-End Strategy for Affiliates

Affiliates should exit the PPC market as quickly and efficiently as they can. Here's why:
  1. Barriers to entry in PPC are incredibly low. If you're working in a space that has decent payouts you either (a) have lots of competition -- which means small margins; (b) are in a really small niche -- which means a relatively small upside; and/or (c) are about to get lots of competition as other PPC affiliates test your niche and discover that it has good margins.
  2. As an affiliate, every conversion has a relatively small lifetime value. (I see this all the time in life insurance -- a single conversion might be worth hundreds of dollars over its lifetime to MetLife, but only $13 to me. There's simply no way I can compete with the 300+ life insurers (not to mention potentially thousands of insurance agents and the networks that funnel leads to them) in the PPC space.
  3. You have no long-term competitive advantage and no assets. On the technology side, let me describe the ultimate technology to you: you know on a keyword-by-keywords basis how much a click costs you, how much it's worth to you, and a pretty good estimate of how much more traffic you could get if you increased the amount you spent per click. You're able to tie those three pieces of information together programatically, so that you maximize your profits. Here's the problem -- lots of companies figured this out 3 or 4 years ago, and they're busy expanding this formula to every niche they can find. Their technology is more robust, more scalable and better designed than whatever you will throw together in the next couple of years. They've already tested and wrung advantage from the next 10 niches you think of, and they've negotiated better payouts on their conversions.
Mini strategies that might buy you a bit more time:
  • Get some of your PPC traffic to subscribe to your email list. Good idea, but it still won't save your bacon -- you'll probably lose them as a conversion in the immediate term, thereby driving up the average cost of your traffic, and running a successful email list takes actual work.
  • Hope that some of your PPC traffic will link to your site, thereby sending you more organic traffic. Yeah, not likely on the a/b tested, limited content landing page you just sent them to. Either (a) your landing page solves their problem so that you make no money from them but 1 in 1000 might link to it; or (b) your landing page convinces them that they need to buy a product to solve their problem, so they leave your page and link to the wonderful product or service that you just pointed them to.
Of course, I don't mean to sound like an idiot -- PPC is great for lots of folks. For example, the folks you're acting as an affiliate for, for example. They make a lifetime value from each conversion 10x what you make. They're reinvesting in an actual business that solves people's problems. They're getting better and better at converting their visitors, so their cost per lead probably stays relatively stable as increasing PPC competition drives up their prices. They reinvesting the money you've earned them into building a bigger affiliate network, meaning that you are less and less important to them, and are subsidizing their attempts to reduce your margins. You're just the sucker that sends them cheap conversions between now and when you're run out of business.