“Todd” was keen to compete in one of the most notoriously difficult categories – work-at-home. He was smart, earnest, hard-working and keen to make affiliate marketing more than just a part-time endeavor. He had dreams of living on the beach in Mexico and surfing between sessions at the computer.

He had put together a sweet little landing site. He had 20 keyword groups running about 10,000 unique visitors a month. He was spending $5,000 per month on pay per click on sales of $6,000. He was collecting email addresses at a rate of 100/day through an opt-in form. He had placed his Google tracking data on the confirmation page of his subscription form. He had great data he collected daily over the past 4 months. He had calculated his visitor value based on commissions divided by clicks and was bidding accordingly. He was a dream client.

Could I help him?

I was drooling over this one. In this case I was 100% sure I could at least double his profits, and likely triple or quadruple. Why was I so sure?

1) He was tracking sign-ups, not conversions 2) He was selling 22 different items off of one landing page 3) He was calculating his visitor value incorrectly

The first thing I did was turn off his keyword groups one at a time. When I got to his top keyword group, the one with the highest conversion rate, we prepared for a massive drop in commissions. There was only a slight dip.

He was stunned. He had made a simple assumption – that there would be a correlation between signups and sales. That is, a keyword that produces twice as many signups would likewise produce twice as many sales. But that wasn’t true. We had seen vividly that there were keywords that produced lots of signups but few sales, and visa versa. He was bidding based on wrong data. His finely-tuned Ferrari of a website was, in fact, a peddler’s cart of clanking pots and pans strapped on with little thought to efficiency.

I had him write letters to each of his advertisers, requesting they place his Google tracking code on their Thank You page. That way he could track sales directly. (Important: Place the code in an attachment; email programs can scramble the code.) For a sample letter, click here.

Most didn’t answer. Normal. After three failed attempts, I advised him to drop that particular advertiser. Others refused, to which I reply, “Are you nuts?” No one stands to profit more from an efficient affiliate than the advertiser. I recommended that he drop them, too. He did, all but one, which was a big money-maker.

Now that he was actually tracking actual sales, not subscriptions, the guesswork was gone. We began eliminating keywords – lots of keywords. Advertising costs plummeted. Sales sailed along at the same pace.

But we were not done.

His landing site was very broad, selling over 20 different products, which was find for keyword groups like work-at-home. Yet about half of his Google ads were related to specific product. I had him create mini-landing pages to receive these ads. For example, a keyword group on medical transcription was no longer directed to the main landing page, but to a mini-landing page with info about medical transcription only. This slashed his bounce backs and upped his sales.

Finally, I had him recalculate visitor value for each keyword group. A visitor who came from a search on “free home-business ideas” is worth far less than one who types, “medical transcription.” We adjusted his bids accordingly. The results? His sales stayed at $6,000 per month, while his advertising costs plummeted to $2,500 per month. His monthly profit had increased 350%, from $1,000 to $3,500. It took about a month to complete the project, and, yes, it was a lot of work, but it was certainly worth it. Best of all, he now had the information he needed to keep the campaign on track as sectors go hot and cold, and keywords get overbid. Whereas his competitors were working in the dark, he understood why and how his site was performing, and why and how it was not.

So let’s review the steps 1)Turn off one keyword group at a time to test assumptions 2)Add tracking to advertiser welcome pages 3)Remove poorly-converting keywords 4)Create mini-landing pages for certain keyword groups 5)Recalculate bids based on visitor value by keyword.

We had taken a campaign that was producing a modest income and transformed it into a juggernaut, without any SEO, all built around pay per click advertising, a hot website and a killer auto-responder sequence. But none of that would have worked without first tracking the keywords that allowed us to find not just any customer, but the right customers.

About the author: Michael Hetzer, The Artful Affiliate, is President of Webster Henrietta Publishing. He is an author, consultant and sought-after speaker. He is the founder of a number of services for Internet affiliates, including “My 1st Google.” You can learn more about Michael Hetzer at: http://www.michaelhetzer.com. Email: Michael@michaelhetzer.com