It's not nice to fool MotherGoogle

In case you missed it, check out this Google Blog:

http://googleblog.blogspot.com/2008/11/searchwiki-make-search-your-own.html

It announces a new feature called SearchWiki and if you are thinking about spending money on SEO, make sure you know what you are doing because you could be throwing your cash into a black hole.

Let's step back for a minute to understand what Google is trying to do in producing their search results. Basically it is all about relevance and nothing else. To the degree that you hear an SEO consultant talk about back links, meta tags, formatting and any other gimmicks to achieve high ranking on Google, watch out. Google is trying to keep one step ahead of these tricks to make sure the search produces the most relevant results. And now with SearchWiki they are enlisting the help of users who can kill your listing or trash it. At this point, this only affects your searches although others can see your ratings. But...the writing is on the wall...I expect Google will make use this in a much more active way in the future. For example, it may allow searchers to rank results based on how others have ranked sites. So despite your best efforts to "trick" Google into a high ranking, if your site is rated low by users your rank could be set back many pages into a trivial position.

So, the future is here and the answer is to work hard at making your content relevant, unique and well organized. And stop trying to fool MotherGoogle.

Organic versus paid listings...not so simple

Much has been written recently about the superiority of organic listings versus paid. This is based on data that shows that 80% of the clicks on a search results page are organic. And that organic clicks produce 1.5 times the conversion rate as paid clicks. On the surface it would clearly say that your advertising money would be far more productive spent against organic search than paid search. But, why is it that 80% of search dollars are still spent against paid search? Is it that this is stupid money, or is something else going on here?

Here's my take...

They both are right but generally for my money, I'd usually opt for paid search, especially if the program is time sensitive like a sale or where speed-to-market is critical like a launch. Even in other cases, with paid search I know I can accurately measure my ROI. And I can use trial and error to build a strong bank of key words and combinations that work.

With organic search, it could take months to build a position of visibility that you can achieve in minutes with paid listings.

With organic listings, you are battling for first page position and not always winning. Search engines are constantly changing their algorithms to prevent "gaming" the system and make certain they are producing quality results. With paid listings, you are on a level playing field and can choose when and where you want to spend your money.

In other words, paid gives you more control, predictability and lower risk. And if there's one thing an advertiser cherishes, its lower risk. Thus 80% of spending is in paid.

This is not to say you shouldn't invest in organic. But the smart way is to make sure your site is optimized to give you the best chance of getting a high listing...but don't bet the farm on it.

Sober news for the ad model

In the Forbes issue of Dec 22, 2008 page 66 an article titled Mixed Messages provides some chilling news for those thinking they can make money on the web with a pure ad model. Forbes shows a chart comparing major Internet companies on an annual sales per unique visitors basis. No surprise, the leader is Google at $1.85. What is a surprise is how little the revenue per visitor number is. Yahoo is at $.91 and MSN at $.41. Here's what this means. Rule of thumb when planning a traffic acquisition program is to use 3X the annual revenue and be willing to spend 25% of that number. In Google's case, that would say they can spend $1.39 per unique visitor. In Yahoo's case, that number is $.68.

If you are starting up a new site with an ad driven model, let's see what the cost might be to use banners to drive traffic. Assuming you can buy banners at $3.00/M and you have a click rate of .04%, your cost per visitor would be $.75. Then you need to ask yourself how long will it take you to achieve anything close to the Yahoo ad sales revenue per visitor...if ever.

Perhaps you can come up with fantastic idea that grows virally on the web and brings millions of uniques for practically nothing...the next Facebook. But even these folks are having a hard time trying to build serious ad revenues.

The models that seem to work these days are still ecommerce and subscriptions. Unless I missed something.

There's always time to do it over...

Talk to most investors in web businesses and they tell you that they assume everything will cost twice as much and take twice as long as the worst case scenario. Even sites with the most brilliant and thorough planning often fail to perform as expected on the Internet. You can stress test sites, view on every different size monitor and resolution, optimize for all the major operating systems and yet when you run for real, there is at least a 50% chance that you will experience a problem serious enough to warrant crisis management.

Why is this?

The answer is that the technology behind the web is still very young and rapidly evolving. Just how many Windows patches are there? And each new version changes the playing field. So what you assumed and planned for yesterday, may not work today.

What can you do?

First, never think you are finished with a site. Consider it a constant work in progress. Second, you must have in place 24-7 monitoring and troubleshooting capabilities. Third, never rely only on in-house expertise to keep you up-to-speed on changes and development. And finally, ask everyone you know to use the site and watch them. As crude as it may seem, real life experience of even a few honest friends can give you as good data as spending thousands on sophisticated market research.