This is the subhead for the blog post

2% – for milk anyway – is pretty good. It’s rich, wholesome and makes for a very nice milk mustache. In the world of online marketing, a lot of people also think 2% is really good when it comes to conversion rate. In other words, if two out of 100 people who visit your site end up buying from you, that’s considered success for a lot of Internet companies.

But should it be? I think not. Another way at looking at 2% – or even 5% or 10%, is that 98, 95, or 90 percent of the people who come to your site leave without paying you a red penny. That’s pretty embarrassing.

In the offline world, not only would this be embarrassing, it would be bankrupting. What would happen to Walmart if only 10% of the people who entered one of their stores actually bought something. Even high-ticket retailers like Best Buy or IKEA probably get 30 or 40% of walk-in customers to walk away buying something (anyone who has data on the actual numbers, please let me know). And what about car dealerships? I actually found a study online suggesting that 41% of people who test drive a car end up buying it from that dealer.

Now consider the world of Internet marketing. A company buys a targeted keyword on Google or Yahoo, for example “Canon inkjet printer.” The company creates targeted ad copy to encourage qualified consumers to click through to their site: “Canon InkJet Printers – Six Models – Starting at $59.95 with free shipping!” The user clicks through to a targeted landing page with the six promised models. And the result – a 2% conversion rate? Something isn’t right here.

And in fact, something isn’t right, starting with the fact that most online marketing campaigns don’t even follow the simple process stated above. Here’s a short list of the common mistakes I see over and over again:

1. Generic keywords: Generic keywords are generally good for one thing – getting you lots of traffic. But since eyeballs aren’t valued at $300 a pop like they were in the bubble, that doesn’t do you much good. Buying the word “electronics” is going to result in a lot of unfocused consumers coming to your site. Unless you are eBay, a shopping comparison site, or Amazon, this is very bad for business.

2. Generic or misleading ad copy: Ads that don’t properly set user expectations doom your site to failure. For example, an ad that invites users to “compare prices from top merchants” only to send users to a site with one price from one merchant is creating a level of distrust from the get-go that is going to reduce conversion rates. Even worse, lazy SEM campaigns often send users to product pages where the product no longer exists. Generic ads are no better – if you create one ad for 500 products, you are going to end up with “browsers” instead of “shoppers” and crush your conversion rates.

3. Non-targeted landing pages: A surprisingly large number of merchants still haven’t figured out the simply art of targeted landing pages; in other words, rather than sending a consumer to your home page, send the consumer to the precise product page. I once heard that HR folks spend an average of 22 seconds on every resume they receive. Thus, if you don’t get their attention immediately, you will immediately be sent to the “rejection” pile. The same applies to your site. Send someone to a non-targeted page and expect the user to go right back to Google and click on a competitor’s ad.

4. No focus on usability: This is probably the #1 mistake Internet retailers make. Unlike the Ronco rotisserie oven, ecommerce Web pages are not “set it and forget it” experiences; you have to constantly test and refine user experience to find out what works and doesn’t work. And this is where most retailers go wrong. They create a Web site and get to the magical 2% conversion rate, then go out for a celebratory beer. Wrong, wrong, wrong.

Consider the impact of just 1% more conversion – from 2% to 3%. If you are paying $1 per click for a keyword and you have a 2% conversion rate on a $50 product, you are making exactly zero dollars ($1 CPC and $1 revenue per click or RPC). At a 3% conversion rate, however, you are now making $.50 on every click. This means one of two things – you either make a nice profit on every click, or you can bid more per click and get even more visitors.

As the paid search market becomes more and more competitive, these little differences in conversion rate will mean life and death for companies in tough industries, like affiliate marketing, electronics retailing, and travel.

My advice is to do the following.

1. Make usability as important to your company as SEM. Human Factors calls this the “institutionalization of usability” – in other words, from the top of the company down, make usability a top priority, in terms of resources, budget, and priority.

2. Measure everything. You already measure click-through-rate, CPC and RPC, but do you measure conversion rate, time spent on site, frequency of visits, number of pages viewed, exit links, conversion funnel, and so on. If you don’t (or if you don’t even know what some of these mean), you need to start right away. Check out Web analytics companies like WebSideStory, Omniture, and CoreMetrics, or even Urchin from Google. Trust me, every major etailer or lead gen company is either using one of these companies or building their own internal competency around analytics.

3. Test, test, test. The best way to improve conversion rates is to test different ad copy, keywords, and ladning pages. You can do this on a very basic level with classic A/B testing. A lot of savvy marketers these days, however, are embracing the concept of multivariate testing or the Taguchi method.

4. Keep up to date. There are a lot of amazing resources online to learn about usability. My favorites are Jakob Nielsen’s, Jared Spool’s and Bryan Eisenberg’s

I could talk about this until I’m blue in the face. In the end, some people will jump on the usability bandwagon, and despite worse products or getting into the online retailing space later than their competitors, these companies may eventually win the day.