A client who sells a consumer product online and doesn't want to reveal its identity was spending about $35,000/month on AdWords. The client was tracking conversions but not the actual dollar amount of each sale. When we connected the shopping cart data to AdWords conversion tracking, we could now see the exact ROI of every keyword, site placement, and ad in the client's account.

We began collecting data on February 7, 2008. I started split-testing ads at that point. After 3 weeks, I ran a keyword performance report and paused negative ROI keywords. I ran a placement performance ad and excluded the negative ROI placements. I ran an ad performance report and deleted the inferior ads.

As you can see from Figure 18-1, the client's AdWords spend dropped significantly, but their total sales remained the same. Before, they spent $35 to make $100. After, they spent only $20. In their case, that amounted to a monthly savings of almost $14,000.

Figure 18-1. By deleting ROI-negative items, spending dropped dramatically without affecting sales.

Here's the really cool thing about this case study. Go back to the first paragraph and read carefully what I did. Notice: There was no creativity involved. No new ads. No clever keyword variations. No insightful market analysis. Just looking at each element and asking, "Is this paying for itself?" And when the answer was no, out came the axe.

Shane Keller of www.gcflearnfree.org, a non-profit organization that offers free online computer, technology, and life skills training to help people improve their lives, visited my office on May 8, 2008. We talked for a while, I looked at his AdWords campaigns, made a few suggestions, and he left.

Shane shared his actions and results on August 15, 2008. Here are the relevant excerpts:

After I met with you, I came back and we adjusted our campaign setting to target a worldwide market instead of just the US. The AdWord campaigns exploded into action. We started maxing out our budget by noon every day. Now, finally I could start to see how each test and adjustment affected the results.

I worked the system — outlined in your book — each and every day and saw immediate improvements week after week. This led to a run of seven consecutive record-breaking weeks of stats for our website, the adoption of Google Analytics to track our conversion rate.

We added Analytics the week of July 7th – July 13th. Now it was all about the conversions. We now have a conversion rate of 15.42% on our AdWords campaign. We are averaging around 6,600 new unique signups each week and we're now heading into what's typically the busiest period for us. Here is a screen shot of our website signup growth. (See Figure 18-2.)

Figure 18-2. Account optimization led to a surge of new traffic for this client.

And here's what it did for our Spanish side of things! (See Figure 18-3.)

Figure 18-3. The increase in traffic was even more pronounced for the Spanish language campaign.

We now are trying to take everything we are learning from AdWords and utilize it in order for us to increase our organic search positions.

David Rothwell of www.adwordsanswers.com took over the AdWords account of a client, www.perforce.com, which provides versioning and collaboration tools and systems for software development companies. When David took over the account, the main challenge was the lack of traffic. Originally, the account had been set up with thousands of long-tail (three- to five-word) keywords to take advantage of the low bid prices. The trouble was, keywords like "embedded software development tool" and "software development management tool" were getting almost no searches.

David relaxed the keyword specificity; began using more common, shorter-tail keywords, such as version control, software configuration management, and software configuration tools; and put them into lots of tightly focused ad groups.

Because they were now bidding on common keywords in broad match format, they had to pay attention to the searches they no longer wanted. David did so by running the search query report on a regular basis, and adding negative keywords to the appropriate ad groups. For example, "kitchen design software" and "website design software" both showed up, so -kitchen and -website were both added to the software design group.

David reports that long-tail keywords, once a best practice, appear to be losing their effectiveness because Google prefers to show ads for keywords with significant search volume. After all, if nobody's searching, then nobody's clicking, and Google isn't making money.

Results: Perforce had been frustrated by their inability to spend more money on AdWords. As they offer a fully functional two-user free trial software download, the targeted search traffic they get converts quite well.

Once David added conversion tracking and removed the negative-ROI keywords and ads, not only did the impressions and CTR increase, but also cost per lead went down. As an added bonus, the extra traffic now allows for more robust split-testing of landing pages, including new ones with video demos.

Originally, the AdWords account had targeted only the United Kingdom. Perforce is building on the success of the UK-based campaign to roll out campaigns in Germany and Sweden.

Christian Bedard of www.calimacil.com sells high-end and high-quality form swords for live action role-playing aficionados. Christian thought he had tapped out all the potential keywords in the swords, shields, and live action weapons department, but recently, he had a brainstorm and bid on the keyword Calimacil and headlined the ad "Calimacil Official Site." (See Figure 18-4.)

Figure 18-4. An effective ad that capitalizes on searches for their own brand name.

He got this idea not from a competitor in his industry, but from copying the AdWords strategy of Dell Computers. (See Figure 18-5.) One of the smartest things you can do as an AdWords advertiser is to pay attention to what other advertisers are doing that catches your eye when you're searching. Most innovation is just borrowing a good idea from somewhere else and applying it where it has never been applied before.

Figure 18-5. Calimacil borrowed the idea for its ad from this Dell ad.

Even though the Calimacil listing comes up at the top of the organic listings for the keyword Calimacil, the ad and organic listing combined generate more traffic and sales than just the top organic listing by itself. Rather than cannibalize the free clicks, the sponsored link reinforces Calimacil's domination of the page. The ad's CTR is 26%, with a very high sales conversion rate, and very low bid price because there are no competitors.

Ken Evoy, president of www.sitesell.com (a turnkey Web hosting and e-commerce business-building system), thought that he had fully optimized his Web site conversion process through years of comprehensive testing. Nothing he tried could beat his control site. But he found that adding a short, friendly Welcome to My Website video to his home page dramatically increased sales for his Site Build It! service. The video helped increase sales by 30 percent by explaining the product and building an emotional connection with visitors.

With the help of Web video consultant Joe Chapuis of www.webvideozone.com, Ken created a 3-minute video shortcut for the SiteSell homepage that walks prospects through a quick tour of his site and service. Ken explained, "Site Build It! is a big product that takes a lot of words to explain. Video enables us to get so much more information across so much more efficiently."

Ken added the video to the top of his home page, using the Flash video player available at www.webvideozone.com. Joe Chapuis explained the importance of video placement: "If you want someone to watch your video, you need to have it at the top of the page where it will get noticed, as well as on a Web page that prospects are likely to visit."

Ken reported that sales increased by 30 percent since he added video to the home page: "Video is incredibly powerful, especially for a product like SBI.... We have never been able to communicate so precisely, effectively, nor with such emotion."

Kelly Conway kindly saved me several hours of work by providing the following case study. Notice the simple steps that cumulatively produced stellar results.

I met David O'Hara, a product-development expert, when he was looking for help in selling one of his products, the Breatheasy blood-pressure reduction system, via PPC advertising. David's goal for this ad campaign is to drive visitors to a landing page where he makes a one-time sale of either a CD or downloadable product. Two landing-page examples include www.highbloodpressurehq.com and www.highbloodpressurehq.com/about_me.html

The campaign includes 30–40 ads, which we continuously split-test. Examples include:

15 Minutes to Lower Blood
Pressure - 6 Weeks to a Better Life
Free from High Blood Pressure
HighBloodPressureHQ.com
2,070 Clicks|3.49% CTR|$0.12 CPC

These Breathing Exercises
Lower High Blood Pressure
Naturally - Just 15 Minutes a Day
HighBloodPressureHQ.com
220 Clicks|4.90% CTR|$0.24 CPC

How To Lower Blood
Pressure - 15 Minutes/Day - 6 Weeks
to Freedom from High Blood Pressure
HighBloodPressureHQ.com
137 Clicks|6.82% CTR|$0.23 CPC

Lower Blood Pressure
Start Immediately - 15 Minutes/Day
Simple, Practical & Affordable!
HighBloodPressureHQ.com
101 Clicks|3.63% CTR|$0.16 CPC

The state of the ad campaign, at the time that David contacted me, was similar to many other campaigns I've seen. He was bidding on 600–800 keywords, all of which were in a single ad group. None of the keyword phrases was making use of anything other than Google's broad-match option. David's initial request was to get the minimum cost-per-click of many of his keywords below the $1 and $5 Google was requesting.

Mike Stewart of www.internetvideoguy.com doesn't spend hours creating new ads or designing amazing Web sites. He doesn't research hungry markets or spend hours creating long e-mail follow-up sequences. He buys cheap clicks on AdWords on keywords related to recording teleseminars and phone calls. He split-tests his ads but doesn't generate enough impressions to make significant changes very often.

Mike's landing page (at www.teleseminartools.com) features a prominent 2-minute video commercial for a home talk show recording studio priced at $1,695. The video cost almost nothing to make, compared to tens or hundreds of thousands of dollars for television commercial production. But the decreased quality doesn't matter on this site, because the AdWords traffic is so highly qualified. A commercial airing during a break in The Office is shown to everyone watching the program; only a tiny percentage of the viewers will be interested in a given product at a given time. That's why commercials on mass media need to be repeated so often. But a commercial on an AdWords landing page can specifically respond to the itch represented by the keyword.

For the low-traffic keyword teleseminars and its variations, Mike generated 134 clicks at $0.15 CPC over a 3-month period, for a total advertising spend of around $20.00. These clicks generated 13 sales totaling more than $22,000 over the same period. Mike cautions that these numbers are possible for two important reasons:

Mike teaches others how to create inexpensive Internet commercials at www.internetvideoguy.com.

One of Mike Stewart's clients is www.carpetdepotdecatur.com, a local carpet store in Decatur, Georgia. The owner, Brad Flack, bids on about 30 keywords in the Atlanta market only, and drives traffic to a Web site that uses video to introduce the store and answer frequently asked questions. The ads are simple, and include the call to action, watch online video.

The video on the home page features Brad introducing himself and explaining the benefits and dramatic differences of his store. He guides visitors to click the "Deal of the Week" video, view customer testimonials, and meet the staff. He also invites them to call and shows them where to find the phone number, and demonstrates how to shop on the site. The most significant call to action is for the visitor to view the "Measure Your Home" how-to video.

Customers now come to the store feeling like they already know Brad from viewing several minutes of video. He has stopped running ads in the Yellow Pages because most customers use Google first and some rely on online search exclusively. Because none of his competitors is advertising with Google at this point (and he hopes they don't read this book!), his ad is the only one that prospects see. With no competition, clicks are cheap and his number-one position is guaranteed.

Joe DiSorbo of www.webgistix.com uses the Web to generate targeted business-to-business leads for e-commerce companies who want to automate or outsource the packing and shipping of orders. Because his service is a complex business-to-business sale, he uses the Web strictly to generate leads. Joe explains how he went from AdWords zero to hero in 12 months by employing very basic strategies:

Prior to using AdWords, we engaged in the painfully slow process of surfing the Internet, locating a potential target (an existing e-commerce company), finding their contact information on their Web site, and then cold-calling them to pitch our services. We opened an AdWords account in May 2005, made a list of all possible keywords, and got started.

Our first 12 months using the system can be broken into 3 distinct 4-month periods with significant jumps in CTR.

Jaco Bolle of www.savefuel.ca sells a small generator that can be installed in cars and trucks to add hydrogen gas to the engine for better gas mileage and lower emissions. When he came to me for help in improving his Web site conversion, he faced several marketing challenges:

Through informal market research (talking about the product with everyone we could find and reading posts in online forums dedicated to hydrogen technology and fuel savings), we identified these and other objections that were preventing sales. Then we put together a nine-day e-mail course, titled "Supplemental Hydrogen Secrets Revealed." Each day addresses a different objection head-on — educating and entertaining the reader, building a bond of trust, and moving them closer to buying.

The first e-mail has the subject line, "Hydro-Gen, Iran, and Poison Ivy." It lists the big benefits of hydrogen supplementation, includes several testimonials, and is chatty and engaging. The references to Iran and Poison Ivy in the subject line, which were included to pique curiosity and get the e-mail read, are echoed in this text from the e-mail:

For most of us, saving money and time (fewer fill-ups) is a pretty good motivation all by itself. But in this case, there [are] a couple of very strong "non-selfish" reasons to want to consume less fuel.

Everyone has a different take on world politics, but I've never met anyone who thought that the U.S. should become more dependent on Middle Eastern oil. It's a little crazy to think that we're worried about Iran's nuclear program and that we've been funding that program at the pumps for years.

And I don't know about you, but I'm starting to believe the signs of global warming. There [have] been reports on the news about poison ivy plants growing faster and being itchier as the carbon dioxide levels in the atmosphere rise.

On the third day, we tackled the big objection: safety. An excerpt from the e-mail, titled "The Hindenburg, Die Hard, and Sad Toddlers: Is Hydrogen Safe?" follows:

Of course, we get a lot of questions at SaveFuel.ca about the Hindenburg. People want to know, "Is my car going to burst into flames over central New Jersey if I install a Hydro-Gen unit?"

It's a good question. It's the first thing I'd ask before installing anything in my car: Is it safe?

First of all, nothing is without its hazards. Forks, staplers, bicycles, cars, toasters — they all can be misused in dangerous ways. So the real question is, Is the Hydro-Gen safe if used properly?

Here's why the Hydro-Gen is safer than the gasoline or diesel you currently use:

+ + + Hydrogen is less of a fire hazard than gasoline + + +

The Hydro-Gen uses a small amount of electricity to turn water into its two component elements, hydrogen and oxygen. Hydrogen is much less flammable than gasoline. Gasoline bursts into flames at anywhere from 228–501 degrees Celsius (442–933 degrees Fahrenheit), while hydrogen doesn't ignite until 550 degrees Celsius (1,022 degrees Fahrenheit).

All those scenes in the Die Hard movies where the car flips over and bursts into flames — that's ordinary gasoline, not hydrogen. Now, cars don't actually burst into flames very often in real life, but still — gasoline is far more likely to ignite than hydrogen gas.

Here's another thing: If you were ever forced to memorize the first row of the periodic table, you'll recall that hydrogen is the first — and therefore lightest — element. How light? Fifteen times lighter than air. Twice as light as helium.

Did you ever give a helium balloon to a small child at the amusement park? And they let go? That balloon really took off, didn't it? Next time you comfort a sad toddler about the loss of their helium balloon, imagine a balloon twice as buoyant in air, filled with hydrogen.

Because it's so light, hydrogen disperses upwards. Quickly. It doesn't stick around and burn, like gasoline does. If you watch a video of the Hindenburg fire, you'll notice something amazing: That giant ship burns hydrogen only for 30 seconds. Rescuers run to the scene almost immediately.

(Check it out yourself in this video: http://tinyurl.com/lue73)

Here's a startling fact: 62 out of 97 passengers on the Hindenburg survived the disaster on May 6, 1937, many of them relatively unharmed. The ones who died either fell, or were burned by dripping diesel fuel.

Note the chatty yet authoritative tone, the use of story to make the point, and the non-salesy nature of the e-mail. This e-mail ends with a by-the-way call to action: " ...if this e-mail has just put your mind at ease regarding safety, here's the link to order a Hydro-Gen and get started saving gas and money right away: www.savefuel.ca/oxy-hydrogen."

Another e-mail addresses concerns about the unit voiding a vehicle's warranty or harming the engine. Day 5's installment answers the objection, "This sounds too good to be true. If hydrogen supplementation works, then the big car companies would be using it."

Slowly, the autoresponder sequence answers objections, provides more and more testimonials, explains the technology, and culminates with an ironclad risk-reversal guarantee. Borrowing language from a Robert Collier sales letter of the 1930s, we write, "By ordering, you obligate me, not yourself. If it doesn't work the way I say it will, I insist that you return the Hydro-Gen for a full and prompt refund."

The result of this e-mail sequence, coupled with several other improvements to the site, including tested headlines, the addition of live chat, and live phone operators available 20 hours per day, was a doubling of the site conversion, from roughly 1.75% to 3.5%. Jaco relates that a competitor approached him in despair, asking whether he would consider selling the competitor's product on the www.savefuel.ca Web site.

David Bullock of www.davidbullock.com believes so strongly in rigorous testing of the sales process that he decided to put his methodology to the test in a nearly impossible situation: buying AdWords clicks for $0.25 and sending the traffic to a Web page promoting an affiliate product that produces $4.25 per unit sale for David. In order to break even in this scenario, David's site needed to produce one sale for every 17 clicks. In other words, if the affiliate site — over which he had no control — could convert a very healthy 5% of its visitors to paying customers, David could not have made money on the front end, even if he had been able to send every single visitor to that site!

His goal for the campaign was a net profit of $30/day, stable and predictable. He gave himself one advantage: He chose an inexpensive product with built-in continuity — an impulse purchase that would lead customers to buy more over time.

During the intense testing phase, David limited his Web site traffic to 100 clicks per day. His conversion at the start of the experiment was 2%: one sale for every 50 visitors to his site.