Many people talk, including myself, about how to build good campaigns that have great quality scores and a solid ROI. But what exactly does this mean in the real world? I have said before, we are all working off essentially the same keywords. So given the same list of say 10,000 keywords, how does one person’s campaign completely fail, while another person makes a killing? It’s all about campaign structure. The best way I can explain this is with an example.
A couple months ago I began a quick and dirty test of a new niche and affiliate offer. To do this I got 10,129 keywords together into 1 campaign with about 35 adgroups. I wrote 2 ads for each adgroup, bought a domain, slapped together a landing page, and began the test. This hastily made campaign actually showed some promise with a slight profit. The potential was there for a real money maker, so I decided to push forward full force on the campaign. This was a heavily contested niche, probably one of the most competitive out there, so I knew I had to start from scratch and build a proper campaign. Here are the steps I took to build the winning campaign.
First of all after taking a hard look at the keywords, I knew I had to logically break up the campaign into 6 separate campaigns. This was necessary for the adgroup structure I had planned to work. To me, adgroup structure is the key to PPC marketing. Keep in mind some of the hard limits of certain services like Yahoo (max 1000 adgroups per campaign).
Next I was able to pare down the list of 10k keywords down to 8,257 using some of the keyword scrubbing techniques described here. The more non- converting keywords you can remove from your lists the better.
Then the hardest and most important part – adgroups. I took the original 35 adgroups, and broke them out to 1,523 adgroups spread out over the 6 campaigns. This was absolutely necessary to get the highest possible relevancy of the keywords to ads. How relevant? Well, about 300 of the adgroups had just 1, short tail keywords in them. Overall, my keyword to adgroup ratio was a tight 5.42 kws/adgroup.
After building out the adgroups, ads had to be written. Recently I have found some evidence that writing non-dynamic ads lead to better quality score than using dynamic insertion, so all ads were written manually. I did use a template to speed up the process, but no keyword insertion tricks were used. I had at least 2 ads per adgroup, to begin the process of A/B testing.
Finally the landing pages. To continue to reflect the relevancy of the keyword, from the ad to the landing page, I needed to build 1,142 landing pages. Much like the ad writing, a template was used, but each page differs enough to be totally relevant to each keyword and ad.
Landing Pages: 1
Landing Pages: 1,142
So you can see, it takes a ton of work to compete in PPC to affiliate marketing. Gone are the days you can slap up a campaign in an hour and have it make money over the long term. The work I describe probably took around 35 hours to complete, and that was just the initial build out. That may sound like a lot of time, but if you consider this will keep making money for months, maybe even years, it’s not anything. Optimizing and improving the campaign is a never ending battle. This campaign is really starting to make money now though, and the time investment has already paid off. The great thing about PPC is that it’s a level playing field. Anyone can compete in any niche, it’s just a matter of how hard do you want to work at it. Work harder than your competitors by building a better campaign, and you will succeed.